Meatqualityinspection-Review

See discussions, stats, and author profiles for this publication at: https://www.researchgate.
net/publication/286926399
Optical Methods and Techniques for Meat Quality Inspection
Article in Transactions of the ASABE (American Society of Agricultural and Biological Engineers) · July 2015
DOI: 10.13031/trans.58.11004
CITATIONS READS
29 4,721
2 authors, including:
Sagar Dhakal
United States Department of Agriculture
62 PUBLICATIONS 1,103 CITATIONS
SEE PROFILE
All content following this page was uploaded by Sagar Dhakal on 10 February 2016.
The user has requested enhancement of the downloaded file.

OPTICAL METHODS AND TECHNIQUES
FOR MEAT QUALITY INSPECTION
Y. Peng, S. Dhakal
ABSTRACT. Meat and meat products are closely associated with the daily eating habits of people around the world.
Quality monitoring of meat and meat products is essential to ensure public health. In recent years, the meat industry has
employed state-of-the art, high-speed processing technology, and meat processors need rapid, non-destructive, easy-to-
use technology to monitor the safety and quality of meats and meat products for economic benefit. Optical technology has
been gaining importance in research and industrial applications for real-time, non-destructive, accurate measurement of
the quality attributes of meat and meat products. Hyperspectral imaging, multispectral imaging, visible-near-infrared
(Vis/NIR) spectroscopy, and machine vision are used in research and industry for detection of the physical, chemical, sen-
sory, and microbiological attributes of meat and meat products. These optical technologies have shown potential for ac-
curate detection of individual attributes as well as multiple attributes simultaneously. This article reviews these optical
technologies for detecting the quality attributes of meat (especially beef, pork, lamb, and poultry). This article also dis-
cusses the prevailing challenges for practical application of optical technologies and future research prospects.
Keywords. Industrial application, Meat, Non-destructive, Optical technology, Rapid, Real-time.
E
conomic development has lifted the living stand- Chemical attributes are responsible for the nutritional value
ards of people around the world. Along with im- of meat. Microbiological organisms, such as bacterial con-
proved living standards, the meat industry has tamination, present a safety hazard in meat and are respon-
seen a massively increased demand for meat and sible for spoilage. Sensory attributes, such as color, smell,
meat products. Meat and meat products have become an flavor, and tenderness, directly influence the purchasing
important part of the daily diet for people around the world. decisions of consumers, as they provide the first impression
It has been reported that, over the past 50 years, global of meat quality.
meat consumption has quadrupled from 70 million tons in The above-mentioned attributes determine the overall
1961 to 283 million tons in 2011 (Tao and Peng, 2014). quality of meat. However, these attributes are mainly ap-
With growing consumption of meat and meat products, the plied to red meat and represent the quality of poultry meat
public has become more conscious of meat quality attrib- to a lesser extent. In poultry, fecal matter and ingesta are
utes. Meat quality is thus one of the important factors that the primary sources of carcass contamination by microbial
contribute to meat prices and to the purchasing decisions of pathogens. Physicochemical, biochemical, and microbio-
consumers. logical properties, including moisture content, water activi-
Meat quality can be defined as a measurement of the pa- ty, pH, total volatile basic nitrogen (TVB-N), and bacterial
rameters, attributes, and characteristics that determine the contamination, determine the freshness of chicken meat.
suitability for consumption of fresh or stored meat without Hence, it can be concluded that the quality attributes of
any deterioration for a certain period of time (ElMasry et meat can be broadly classified into the four categories re-
al., 2012b). Meat quality includes physical qualities, chem- gardless of the type of meat (fig. 1).
ical composition and nutritional content, microbiological Meat quality is influenced by many factors, from animal
organisms in the meat, and sensory attributes, as shown in breeding, growth, and slaughter, to packaging, storage, and
figure 1. Physical qualities include factors such as marbling transportation, and finally to cooking. This means that meat
and water-holding capacity (WHC). Chemical composition
and nutritional content, such as protein, fat, calcium, and Meat Quality
moisture content, represent the chemical attributes of meat.
Submitted for review in October 2014 as manuscript number PRS

11004; approved for publication by the Processing Systems Community of Physical Chemical and Microbiological Sensory
ASABE in July 2015. Attributes Nutrition Organisms Attributes
The authors are Yankun Peng, ASABE Member, Professor, and
Sagar Dhakal, Doctoral Student, College of Engineering, National R&D marbling, protein, alcium, bacterial color,
Center for Agro-Processing Equipment, China Agricultural University, water holding fat, moisture, contamination, smell, flavor,
Beijing, China. Corresponding author: Yankun Peng, 17 Qinghua East capacity, etc pH, etc etc tenderness, etc
Road, Haidian, Beijing 100083, China; phone: 86-01-62737703; e-mail:
ypeng@cau.edu.cn. Figure 1. Quality attributes of meat in general.
Transactions of the ASABE

Vol. 58(5): 1371-1386 © 2015 American Society of Agricultural and Biological Engineers ISSN 2151-0032 DOI 10.13031/trans.58.11004 1371
has a high risk of deterioration through the entire produc- compared to manual inspection, they require long pro-
tion chain, from breeding to consumption. The meat indus- cessing times, complex sample preparation, and trained
try is currently facing the challenge of obtaining reliable technicians, and they are destructive. Instrument methods
information on meat quality throughout the production pro- and chemical analysis cannot satisfy the meat industry’s
cess, to ensure the quality of meat products provided to need for fast, reliable, accurate, and non-destructive detec-
consumers (Tao and Peng, 2014). tion of meat quality.
Meat is also highly perishable due to its high nutritional Recent years have seen rapid advances in meat quality
content. The chemical composition and sensory attributes detection methods. Various methods have been reported in
of meat are highly influenced by pre-slaughter factors the literature for detection of the quality attributes of meat
(breed, sex, age, weight, and environment) and post- and meat products. Among these methods, biosensors have
mortem factors (storage time, temperature, etc.) (Andres et gained importance for detection of meat quality attributes
al., 2007; Lawrie, 1985). Pre-slaughter and post-mortem (Balasubramanian et al., 2009; Damez et al., 2008; Haugen
changes cause quality deterioration in meat, thus requiring et al., 2006; Ohk and Bhunia, 2013; Panigrahi et al., 2006;
reliable technology for meat quality determination before Zor et al., 2011). Although biosensors can provide rapid
consumption. Traditionally, trained inspectors were em- detection with satisfactory accuracy, they cannot satisfy the
ployed to detect meat quality attributes manually (Xiong et meat industry’s need because of their destructive nature.
al., 2014). Sensory attributes, such as tenderness, were as- Meanwhile, optical technologies have received wide atten-
sessed by pressing the meat, flavor was assessed by smell- tion in the research community because of their rapid, non-
ing the meat, and color and marbling were evaluated by destructive, real-time, and precise detection ability (Gomez
comparing the color of the rib eye muscle (musculus et al., 2006; Jamshidi et al., 2012; Lu and Peng, 2006; Pan
longissimus dorsi) or the proportion of intramuscular fat et al., 2015; Peng and Lu, 2007; Qin et al., 2013; Xue et al.,
within the m. longissimus dorsi against reference standards 2012; Zhang et al., 2014). Consequently, optical technolo-
for each meat species. However, manual inspection cannot gies are gaining importance in industrial automation (Li et
evaluate the internal quality attributes of meat, such as nu- al., 2011a; Mehta and Burks, 2014; Tong et al., 2013).
tritional content, pH, moisture content, presence of micro- Much effort has been made in the research and devel-
organism, and so on. Manual inspection is also tedious and opment of optical technologies for detection of the internal
subjective and depends strongly on the inspector’s alertness and external quality attributes of meat. Rapid, non-
and fatigue. Poultry carcasses are inspected visually for destructive, real-time, and precise detection of the physical,
surface contaminants such as ingesta and feces. The typical chemical, sensory, and microbiological quality attributes of
U.S. poultry processor handles 140 to 180 birds per minute meat and meat products from the start of the process to
on each processing line (Yoon et al., 2011). Although ex- consumption will protect public health, promote meat con-
ternal and sensory quality attributes can be evaluated by sumption, and provide an economic benefit for the meat
human inspectors, internal quality attributes, such as nutri- industry. This article reviews the optical technologies used
tion content, micro-organisms, pH level, and TVB-N con- for rapid and non-destructive detection of internal and ex-
tent, are difficult or impossible to detect by human inspec- ternal quality attributes of meat.
tion.
Instrument methods and chemical analysis have been
used for detection of internal quality attributes of meat for MAIN OPTICAL TECHNOLOGIES
decades. Such methods are effective and considered relia- Modern agricultural and food industries have seen rapid
ble for their consistent results. Instrument methods such as application of optical technologies for detection of the qual-
Warner-Bratzler shear force (WBSF) have been widely ity attributes of fruits, vegetables, meat and meat products,
used for evaluating meat tenderness, a pH meter inserted eggs and egg products, fish, processed products, etc. (Du
into a meat sample has long been used for pH measure- and Sun, 2008; Dowlati et al., 2012; ElMasry et al., 2012b;
ment, and meat color is evaluated using colorimeters. These Feng et al., 2008; Ji et al., 2012; Kimiya et al., 2013; Lee-
instruments are used to evaluate the texture or freshness of mans et al., 1998; Lin et al., 2011; Yao et al., 2014). Opti-
meat. Similarly, chemical methods, such as Kjeldahl, are cal technologies have proven their potential in a wide range
used for protein analysis (Xiong et al., 2014a), and other of applications in agriculture and food. The reflectance,
solvent methods are used for measuring fat content. The transmittance, and absorption of incident light on a sample
plate count method has long been used for detection of mi- depend on the sample’s chemical and physical properties,
cro-organisms in meat. Meat is prone to spoilage during and these relationships provide a detection mechanism for
storage due to microbial activity. Conventional laboratory optical technology. The interaction of light with the sample
analysis methods, such as high-performance liquid chroma- surface can be used to detect surface properties. Hence, by
tography (HPLC) and ultra-high performance liquid chro- illuminating the sample surface, we can capture a surface
matography tandem mass spectrometry, are used for detec- image with a camera. Scattering of light occurs due to in-
tion of meat quality based on spoilage status, chemical con- teraction of the incident light with the chemical and physi-
tent such as inulin, contaminant residue, foodborne patho- cal properties of the sample, causing changes in the direc-
gens, and other chemical constituents (Argyri et al., 2011; tion of light propagation. Scattering characteristics can be
Boselli et al., 2008; Clariana et al., 2010; Shalaby et al., quantified using mathematical models for nondestructive
2011; Vendrell-Pascuas et al., 2000; Xu et al., 2012). Alt- detection of different chemical and physical properties. The
hough these methods have proven consistent and reliable
1372 TRANSACTIONS OF THE ASABE

amount and pattern of light scattering depends on different
factors, such as the wavelength of the incident light, the Connection Cable
penetration of the incident light, the physical and chemical
CCD Camera
properties of the sample, and the sample’s refraction index,
to name a few.
Depending on the application, optical technologies can Imaging Spectrograph
be broadly classified as spectroscopy, machine vision, and
Computer
hyperspectral imaging. Spectroscopy is used to establish Light Source Lens Light Source
the relationship between the electromagnetic radiation and
the sample to develop a chemometrics model for detecting
different quality attributes of the sample. Spectral data col-
lected from a point or small area on the sample surface are
used to develop the chemometrics model. A spectroscopic Controller
system consists of a spectrometer, a light source, an optical Sample
x y
probe, a controller, and related software. Spectroscopy sys- Moving Platform
Speed
tems have been developed for different wavelength ranges,
such as Vis/NIR, infrared, UV, etc. Spectroscopy can be
used only for point scans, and it does not provide spatial Figure 2. Schematic of hyperspectral imaging system.
information. A spectroscopic system is generally used to
detect the chemical constituents of a sample for quality QUALITY INSPECTION BY HYPERSPECTRAL IMAGING
measurement, such as the freshness of meat (Prieto et al., Among the different types of meat, red meats such as
2009a, 2009b; Venel et al., 2001). beef, lamb, and pork are widely consumed due to their high
Machine vision (also called computer vision) is mostly nutritional value. With increasing consumption, technolo-
used for detecting external quality attributes of samples gies for quality detection of red meat are also advancing
through analysis of digital images. Machine vision analyzes rapidly. State-of-the-art hyperspectral imaging technology
the spatial information obtained from digital images of and the corresponding algorithms for non-destructive and
samples, such as color, size, and surface structure. As such, rapid detection of various quality attributes of red meat
the application of machine vision is mainly limited to sur- have been reported in the literature. Hyperspectral imaging
face detection. A machine vision system generally consists has been applied for detection of nutritional content, mois-
of a CCD camera, a light source, a computer, and related ture content, pH, tenderness, microbial spoilage, WHC,
software. Machine vision systems have been widely used in marbling, drip loss, and color.
agriculture and food for detection of surface quality charac-
Beef
teristics (i.e., color, size, shape, etc.) (Chmiel et al., 2011a,
Chemical and nutritional attributes are important factors
2011b; Penman, 2001; Zhang et al., 2015).
in determining the quality of red meat. The chemical con-
Hyperspectral imaging combines features of imaging
stituents in red meat, such as protein, fat, fatty acids, and
and spectroscopy, enabling it to detect the external charac-
moisture, affect the quality of red meat, and chemical reac-
teristics of a sample by imaging (such as shape, size, and
tions cause changes in the flavor, color, tenderness, and
color) as well as chemical constituents of the sample by
appearance of red meat (Xiong et al., 2014). As such, de-
spectral analysis (ElMasry et al., 2012a). Hyperspectral
tection of chemical attributes is important to ensure con-
imaging provides three-dimensional information on the
sumer health and satisfaction as well as reduce the meat
sample. Three-dimensional hyperspectral cubes, i.e., (x, y,
industry’s economic losses. Reflectance hyperspectral im-
λ), where λ is the wavelength, can be acquired using point- aging in the Vis/NIR range (900 to 1700 nm) has been used
scan, line-scan, and area-scan methods. A hyperspectral in conjunction with chemometrics models for determination
imaging system typically consists of a CCD camera (or an of different quality attributes, such as water, protein, and fat
InGaAs camera for the NIR region), an imaging spectro- content in beef of different breeds (ElMasry et al., 2013);
scope, a light source, a sample holding platform, a comput- color, pH, and tenderness of fresh beef (ElMasry et al.,
er, and related software (Leiva-Valenzuela et al., 2014; Qin 2012a); and tenderness of 14-day cooked beef (Naganathan
et al., 2013; Siedliska et al., 2014; Tao and Peng, 2014). et al., 2008b). These studies reported the use of partial least
Figure 2 is a schematic of a hyperspectral imaging system. squares regression (PLSR) to reduce the high dimensionali-
Hyperspectral imaging systems can be categorized into ty of the hyperspectral data. Naganathan et al. (2008a) re-
Vis/NIR imaging (He et al., 2014; Pierna et al., 2012; Yu et ported that a PLSR model could classify beef as tender,
al., 2014), fluorescence imaging (Cho et al., 2013; Davis et intermediate, and tough with classification accuracy of 78%
al., 2014; Noh et al., 2007), and Raman imaging (Qin et al., based on slice shearing force (SSF) analysis. The authors
2011). The light source in a hyperspectral system can be a
defined SSF ≤ 205.8 N as tender meat, SSF between 205.8
point source (Erkinbaev et al., 2014; Tao and Peng, 2014)
and 254.8 N as intermediate, and SSF > 254.8 N as tough
or a diffuse source (Dai et al., 2014; Everard et al., 2014;
meat. In a similar study, a hyperspectral imaging system
Lu and Ariana, 2013; Siedliska et al., 2014). The infor-
(496 to 1036 nm) was used to measure the tenderness of
mation obtained from hyperspectral image acquisition is
beef (Cluff et al., 2008). The authors reported fitting the
used for detecting internal as well as external qualities of
optical scattering profiles obtained from the hyperspectral
the sample.
58(5): 1371-1386 1373

images using a modified Lorentzian function. Stepwise nation (Tao and Peng, 2014). Similarly, a combination of
regression was used to identify the feature wavelengths to Lorentzian and Gompertz functions was used to extract
predict the tenderness of the samples in terms of WBSF useful information from hyperspectral imaging data to pre-
with a correlation coefficient (R) of 0.67. dict total viable count (TVC) in pork stored at 10°C for 1 to
Application of a time series hyperspectral imaging sys- 15 days after slaughter (Tao and Peng, 2015). PCA was
tem (380 to 1700 nm) was reported to determine the water reported to extract characteristic pictures from hyperspec-
distribution in beef during dehydration using principal tral images in the spectral range of 430 to 960 nm to evalu-
component analysis (PCA) (Wu et al., 2013). A successive ate TVC in pork meat 24 h postmortem (Huang et al.,
projections algorithm (SPA) was used to select the feature 2013b).
wavelengths, and a multi-linear regression (MLR) model Freshness is an essential index of meat quality and di-
was developed to predict the water content in beef slices rectly influences storage and logistics. TVB-N, pH, and
with a coefficient of determination of 0.95. color are the most important factors for meat freshness as-
sessment. Among the three parameters, TVB-N is the most
Lamb
important (Li et al., 2011c; Zhang et al., 2012). Hyperspec-
Potential application of near-infrared (NIR) hyperspec-
tral imaging has been used with different chemometrics
tral imaging has been reported for determination of the
modeling methods to predict TVB-N in pork as a measure
chemical composition (water, fat, and protein content) of
of its freshness (Chen et al., 2013; Li et al., 2015). Li et al.
lamb meat (Kamruzzaman et al., 2012a; Pu et al., 2014).
(2013) developed a multispectral imaging system to collect
Kamruzzaman et al. (2013) reported a PLSR method to
scattering images of pork samples and selected the wave-
detect water, fat, and protein content in lamb with coeffi-
lengths of 517, 550, 560, 580, 600, 760, 810, and 910 nm
cients of determination of 0.88, 0.88, and 0.63, respective-
for TVB-N modeling.
ly. Feature wavelengths were extracted for the PLSR mod-
Salting is used to produce high-quality of meat and meat
eling to reduce the computational complexity, thus reduc-
products by avoiding (or delaying) microbial growth and
ing the time for detection of each sample. Pu et al. (2014)
thus increasing the product’s stability. Meat quality can
used a combination of uninformative variable elimination
deteriorate during salting, resulting in nutritional loss and
(UVA) and successive projections algorithm (SPA) to ex-
texture problems. Moisture content has been considered a
tract feature wavelengths (eight different feature wave-
major source of growth for microbial organisms in meat,
lengths for fat, protein, and water). From the feature wave-
and it is considered a key parameter affecting meat texture
lengths, seven wavelengths were selected to develop MLR
(Ruiz-Ramirez et al., 2005; Serra et al., 2005). Microbial
models. The MLR models were reported to have correla-
growth or texture loss influences the shelf life of salted
tion coefficients of calibration of 0.95, 0.80, and 0.91 for
meat and meat products. Hyperspectral reflectance spectra
fat, protein, and water content, respectively. The UVA and
(RS), absorbance spectra (AS), and Kubelka-Munk spectra
SPA method greatly reduced the high-dimension hyper-
(KMS) in the wavelength range of 1000 to 2500 nm were
spectral data, and the MLR model had superior results
reported for prediction of moisture content in pork slices
compared to PLSR. Similarly, hyperspectral imaging (900
during salting (Liu et al., 2014b). The same researchers
to 1700 nm) was applied for non-destructive determination
reported application of hyperspectral imaging in the spec-
of the tenderness of lamb meat (Kamruzzaman et al.,
tral range of 400 to 100 nm to predict color and pH in salt-
2012b). The same group of researchers also reported the
ed pork (Liu et al., 2014a).
potential use of NIR hyperspectral imaging (900 to
To reduce the hyperspectral data, hyperspectral imaging
1700 nm) for discrimination of three different lamb mus-
can be implemented in a multispectral imaging mode. Ma
cles, i.e., semitendinosus (ST), longissimus dorsi (LD), and
et al. (2014) used a multispectral imaging system in the
psoas major (PM), of Charollais breed (Kamruzzaman et
Vis/NIR region to determine the aerobic plate count (APC)
al., 2011).
in cooked pork sausages. A novel algorithm was reported
Pork for feature selection to develop linear and non-linear re-
Similar to beef and lamb, hyperspectral imaging has gression of L*, a*, and b* color from multispectral images
been extensively used for non-destructive determination of of pork samples (Sharifzadeh et al., 2014).
the quality attributes of pork. A hyperspectral imaging sys-
Other Meats
tem in the NIR region (900 to 1700 nm) was used to deter-
Measuring the quality attributes of poultry meat is im-
mine the protein, moisture, and fat content in intact and
portant to ensure public health and consumer satisfaction
minced pork (Barbin et al., 2013). A PLSR model was de-
and standardize the meat processing procedure for econom-
veloped to select a few feature wavelengths from the high-
ic benefit. As with red meats, hyperspectral imaging has
dimension hyperspectral data. The authors reported coeffi-
been reported for detection of different quality attributes of
cients of determination of 0.92, 0.88, and 0.94 for protein,
poultry meat.
fat, and moisture content, respectively, which are better
Hyperspectral imaging in the visible range of 400 to
than those for beef (ElMasry et al., 2013) and lamb
1000 nm was reported for measuring the springiness of
(Kamruzzaman et al., 2012a).
fresh chicken meat (Xiong et al., 2015). Hyperspectral im-
A novel method including the modified Gompertz func-
aging has also been reported to determine the quality of
tion was used to extract scattering characteristics of pork
chicken meat during processing. Fusion of hyperspectral
from spatially resolved hyperspectral images (400 to
transmittance and reflectance imaging was reported for
1100 nm) for evaluation of tenderness and E. coli contami-

detection of embedded bone fragments in chicken fillets ment of protein content. That study also reported the poten-
(Yoon et al., 2008). In discussing the potential industrial tial of Vis/NIR spectroscopy with a first-derivative data
application of the imaging technology, the authors men- processing algorithm for simultaneous detection of various
tioned that variation in meat thickness influenced the imag- quality parameters. A Vis/NIR spectroscopy system with a
ing results, which can be overcome by better control of conveyor belt for on-line detection of meat quality showed
meat slicing thickness. encouraging results for application in industry (Prieto et al.,
Feng et al. (2013) applied a hyperspectral imaging sys- 2009a). Vis/NIR reflectance spectroscopy in the spectral
tem in reflectance mode for detection of Enterobacteri- range of 350 to 1800 nm was used to measure color, cook-
acease on chicken fillets. The authors reported the devel- ing loss, texture, and sensory characteristics of beef.
opment of a bacterial prediction model by PLSR using full Vis/NIR spectroscopy (350 to 2500 nm) was also used for
wavelengths in the spectral region of 430 to 1450 nm and on-line classification of beef carcasses (cows, bulls, steers,
achieved a correlation coefficient of determination (R2) and heifers) based on ultimate pH (measured 48 h postmor-
greater than 0.82. Similarly, Feng and Sun (2013a) used tem) (Reis and Rosenvold, 2014). The authors reported that
line scanning hyperspectral imaging (900 to 1700 nm) for optical spectroscopy with PLS can replace existing meth-
rapid and non-destructive determination of Pseudomonas ods of detection in on-line classification of beef carcasses.
counts in chicken breast fillets. The hyperspectral images In addition to measurement of quality attributes, studies
were correlated with the actual counts of Pseudomonas by have also been performed to classify meat based on fresh-
PLSR. The authors also reported determination of TVC in ness. Infrared spectroscopy has proven to be a potential tool
chicken breast fillets using NIR hyperspectral imaging in for detecting beef spoilage by micro-organisms (Argyri et
the spectral range of 910 to 1700 nm (Feng and Sun., al., 2010; Ellis et al., 2004; Kodogiannis et al., 2014). Fou-
2013b). In addition to detection of chicken quality attrib- rier transform infrared (FTIR) spectroscopy (4000 to
utes, research has been reported on prediction of moisture, 400 cm-1) and machine learning algorithms were used to
color, and pH in cooked turkey ham by NIR hyperspectral detect beef fillet spoilage in two different storage condi-
imaging in the spectral range of 900 to 1700 nm (Iqbal et tions (chill temperature and abuse temperature) (Argyri et
al., 2013). The research showed the ability of NIR hyper- al., 2010). The artificial neural network algorithm correctly
spectral imaging to detect different quality attributes simul- classified fresh samples with 91.7% accuracy, semi-fresh
taneously. samples with 81.2% accuracy, and spoiled samples with
94.1% accuracy. The detection accuracy for semi-fresh
QUALITY INSPECTION BY SPECTROSCOPIC METHODS samples was lower than that for the other two classes of
Spectroscopic methods such as visual and NIR, Raman samples. It was reasoned that the fresh samples and spoiled
spectroscopy, and x-ray are often reported for rapid and samples that were close to semi-fresh samples, and vice
non-destructive detection of different quality attributes of versa, reduced the accuracy of the algorithm.
meat. Spectroscopic methods generally use point scanning. Vis/NIR spectroscopy has also been used for measuring
Different from line scanning or area scanning, point scan- different indexes in cattle fed sunflower seed or flaxseed
ning can be used for more homogenous samples. Acquisi- (Prieto et al., 2014). The authors reported the application of
tion of spectral signals from a few representative points on NIR spectroscopy in the range of 400 to 2498 nm for ana-
the sample creates a greater chance for variance in the lyzing the meat’s chemical composition, quality attributes,
measurement accuracy and stability if the sample is are and fatty acid content. Infrared technology was also used to
chemically or physically uneven. However, for homoge- measure beef quality during processing (Perisic et al.,
nous samples, spectroscopic methods can be reliable, accu- 2013). Excessive intake of sodium has a negative influence
rate, rapid, and simpler compared to other methods. on human health, and attempts have been made to replace
NaCl (major source of sodium in meat) with other salts.
Beef
However, this might influence the quality attributes of
Vis/NIR spectroscopy has been applied for detection of
meat. FTIR and Vis/NIR measurements were made of the
multiple quality attributes of beef, such as chemical com-
effects of KCl and MgSo4, as substitutes for NaCl, on the
position (Prieto et al., 2006), characterization of beef mus-
quality of beef samples. MgSo4 was acceptable when it was
cle on the basis of cooked WBSF (Xia et al., 2007), and
applied at a mixture ratio of 1:3. Morsy and Sun (2013)
measurement of quality attributes of ground and intact beef
reported on the application of NIR spectroscopy with PLSR
in terms of pH, color (L*, a*, b*), aging loss, cooking loss,
to detect fresh and frozen-thawed minced beef adulterated
and WBSF (Marchi et al., 2013). While these studies re-
with pork, fat trimmings, and offal. The authors reported
ported the potential of Vis/NIR for measurement of beef
high accuracy (coefficients of determination >0.9 for all
quality attributes, the measurement accuracy was generally
three adulterations) for minced beef samples; however, the
not satisfactory because scanning a single point at a time on
detection accuracy was considerably lower for frozen-
each sample can cause poor performance of the model.
thawed samples.
Sample homogeneity is a crucial factor for development of
efficient chemometrics models using spectroscopic meth- Lamb
ods. Ripoll et al. (2008) reported good measurement results A hand-held Raman-based optical instrument was used
with NIR reflectance spectroscopy for tenderness (R2 = for measurement of shear force in lamb m. longissimus
0.98), myoglobin (R2 = 91), WHC (R2 = 0.82), and WBSF lumborum (Fowler et al., 2014a). A 671 nm laser source
(R2 = 0.7). However, the correlation was low for measure- was used in the optical device to measure the shear force by
58(5): 1371-1386 1375

correlating the Raman spectral signal with the value meas- cal changes in fresh minced pork meat (Papadopoulou et
ured by traditional methods. Poor correlation was reported al., 2011). A handheld Raman prototype with a 671 nm
for predicting the shear force in lamb muscle. The same microsystem diode laser was reported for in situ detection
researchers reported using Raman spectroscopy for predict- of microbial spoilage in pork meat (Sowoidnich et al.,
ing the tenderness of fresh lamb meat samples, but the re- 2010). A novel technology based on mid-Fourier transform
sult was not promising (Fowler et al., 2014b). NIR reflec- infrared spectroscopy (MID-FTIR) was reported for detec-
tance spectroscopy was used in a correlation between the tion of Trichinellaspiralis infection in pig muscles (Gomez-
spectral signal and sensory parameters that are directly re- De-Anda et al., 2012). Spectroscopy in the spectral range of
lated to the eating quality of lamb meat (Andres et al., 4000 to 650 cm-1 with attenuated total reflectance and soft
2007). The authors reported R2 values of less than 0.40 for independent modeling of class analogy was used to classify
predicting the texture, juiciness, flavor, abnormal flavor, the samples with 100% accuracy for both recognition and
and overall quality when correlating the spectral signal with rejection based on sample infection.
the scores from a tasting panel. However, the authors re- NIR spectroscopy in the spectral range of 1150 to
ported significant increases in the prediction accuracy for 2250 nm was used to characterize absorption, scattering,
the best and worst samples. This shows that the NIR spec- and angular scattering distribution of pig adipose tissues for
tral signal is informative for detecting quality attributes of non-destructive analysis of intact adipose tissue for indus-
lamb meat. The same study reported R2 values of 0.84 and trial application (Zamora-Rojas et al., 2013). In the late
0.67 for predicting intramuscular fat and water content, 1990s, NIR reflectance (900 to 1800 nm) was used for
respectively, in lamb meat. measurement of WHC and drip loss in fresh pork with cor-
NIR spectroscopy has been used extensively for meas- relations greater than 80% (Forrest et al., 2000). Three
urement of chemical constituents in lamb meat, and prom- years later, Pedersen et al. (2003) reported that the infrared
ising results have been reported. NIR reflectance spectros- region from 1800 to 900 cm-1 contained the best predictive
copy was used to measure chemical constituents such as information for measuring WHC of pork meat. The authors
ash, dry matter, crude protein, and fat in mutton (Viljoen et used infrared spectroscopy with PLSR to measure WHC in
al., 2007) with correlation coefficients of 0.97, 0.96, 1.00, pork with a correlation coefficient of 0.89. The results were
and 1.00, respectively. The same method was used to de- confirmed in on-line measurement of pork WHC using FT-
termine mineral contents in lamb meat samples, with corre- IR spectroscopy.
lation coefficients ranging from 0.86 (for Zn and K) to 0.92 An effort was made to develop an on-line optical spec-
for Mg, but low accuracies were found for Cu, B, Mn, Ca, troscopy system to measure quality attributes of fresh pork
and Al. Guy et al. (2011) applied NIR reflectance spectros- (Liao et al., 2010). A conveyor belt combined with Vis/NIR
copy for detection of fatty acids in ground lamb meat sam- spectroscopy (350 to 1100 nm) was designed to acquire
ples and intact unground samples. The results showed that spectral signals from fresh pork to measure the intramuscu-
NIR reflectance spectroscopy predicted the fatty acid con- lar fat, protein, water, pH, and shear force rapidly and non-
tent in ground meat more accurately than in unground meat. destructively. Later, Liao et al. (2012) developed another
The differences in the degree of homogeneity of the ground on-line system for non-destructive measurement of pH val-
and unground meat samples could be one reason for the ues in fresh pork using Vis/NIR spectroscopy.
difference in accuracy.
Other Meats
Sun et al. (2012) reported a novel method for detecting
Extensive studies have been reported on applications of
the quality of lamb meat. The authors used NIR reflectance
optical spectroscopic methods for detection of different
spectroscopy in the spectral range of 950 to 1650 nm to
quality and safety attributes of poultry. NIR spectroscopy
classify lamb meat according to geographic origin and also
has been used extensively for detection of fatty acid content
predicted δ13C and δ15N with high accuracy. Dian et al.
in different parts of chicken meat with encouraging results;
(2008) classified lamb carcasses based on feeding method
however, the method did not perform well when measure-
for quality detection using Vis/NIR spectroscopy in the
ments were carried out in real-time (Marchi et al., 2012;
spectral range of 400 to 2500 nm.
Riovanto et al., 2012; Zhou et al., 2012). Short-wavelength
Pork NIR spectroscopy (400 to 1000 nm) was reported for detec-
NIR spectroscopy has shown promising results for de- tion of moisture content, water activity, pH, TVB-N, ATP
tection of chemical constituents such as fat, moisture, and breakdown compounds, and mesophilic bacteria, which
protein in pork (Perez-Marin et al., 2009; Zamora-Rojas et were regarded as the useful parameters for prediction of
al., 2013). A study of non-destructive measurement of pork chicken freshness (Grau et al., 2011). Like red meats, poul-
freshness with storage time was reported using FT-NIR try is also susceptible to bacterial spoilage. Studies have
spectroscopy in the spectral range of 10,000 to 4000 cm-1 been reported on determining the spoilage of chicken meat
(Chen et al., 2011). Different mathematical models, such as during aerobic storage using NIR and fluorescence spec-
linear discriminant analysis (LDA), back-propagation arti- troscopy (Sahar and Dufour, 2014; Sahar et al., 2011). A
ficial neural network (BP-ANN), and k-nearest neighbor recent study showed the potential of NIR spectroscopy (400
(kNN), were used to develop prediction models, among to 2500 nm) for quality assessment of chicken meat based
which BP-ANN was reported to have superior results over on pale, soft, and exudative (PSE), L*, pH, and WHC
the other models. FTIR spectroscopy in the spectral range (Barbin et al., 2015). Other spectroscopic methods such as
of 4000 to 650 cm-1 was reported for monitoring biochemi- Raman (excitation by 785 nm) also showed promising re-

sults for establishing spectral fingerprints to monitor the vision and image processing have been reported in the lit-
WHC of broilers based on growth rate (Phongpa-Ngain et erature for detection of different quality attributes of beef.
al., 2014). Examples include prediction of shrinkage of ellipsoid beef
joints (Zheng et al., 2007); prediction of color, marbling,
QUALITY INSPECTION BY MACHINE VISION and surface texture (Jackman et al., 2008); measurement of
AND IMAGE PROCESSING texture features to classify beef as tough or tender (Li et al.,
Machine vision has proven to be effective for quality de- 2001); prediction of skeletal maturity based on cartilage
tection of agricultural products (Dowlati et al., 2012; ossification in the thoracic vertebrae (Hatem et al., 2003);
ElMasry et al., 2012c; Khazaei et al., 2013; Kurtulmus and measurement of CIE L*, a*, b*, hue angle, and chroma
Kavdir, 2014; Wang et al., 2011). Machine vision methods (Larrain et al., 2008); automatic segmentation of the longis-
have been widely used for surface quality detection of dif- simus dorsi muscle and marbling (Jackman et al., 2009);
ferent meats and meat products. However, due to the lim- prediction of percentage surface metmyoglobin on fresh
ited information in the digital images provided by machine beef, which causes wide variation in surface color (Demos
vision, this method cannot be used in chemometrics model- et al., 1996); and classification of fresh and stained meat
ing to predict the chemical composition and internal quality samples based on marbling in the longissimus thoracis
attributes of meat. muscle (Pena et al., 2013).
Beef Pork
Machine vision has been used extensively for measuring Machine vision and image processing have been exten-
color, fat, and other physical parameters of meat. Studies sively reported for detecting different physical quality at-
have reported developments of novel hardware for machine tributes, such as marbling, back fat thickness, color, texture
vision systems and software algorithms for image pro- features, and intramuscular fat content. Digital image anal-
cessing to extract useful information for non-destructive ysis was used for quantitative description of marbling fat
and rapid detection of physical quality attributes of meat. A and its relationship with intramuscular fat content and shear
machine vision system was developed using a color digital force variation in pork (Faucitano et al., 2005). RGB-based
camera to measure the fat color of beef for quality grading digital color analysis was applied for prediction of marbling
(Chen et al., 2010). An image processing algorithm was in pork (Huang et al., 2013a). The region of interest ac-
reported for prediction of intramuscular fat content in beef quired from the sample image was used for extraction of
m. longissimus dorsi (Du et al., 2008). The image pro- marbling using a wide line detector and gray-level co-
cessing algorithm initially removed noise in the original occurrence matrix. Prediction models were developed indi-
image, followed by automatic segmentation of the image vidually using red, green, and blue channels as well as us-
into regions of interest using kernel fuzzy c-means cluster- ing the RGB three-channel with a color wheel (or an RGB
ing. Finally, intramuscular fat content was extracted from LED array). The prediction model developed based on the
the image for development of a prediction model. Similar- green channel had high correlation coefficients (Rc and Rv
ly, Chen and Qin (2008) reported a machine vision and of 0.94) for predicting pork marbling. Machine vision and
image processing method for segmentation of marbling. digital image processing have also been used for measure-
Because simple RGB digital images do not contain spec- ment of other physical attributes, such as lean and fat com-
tral information for chemometrics modeling, a different position based on video images (Jia et al., 2010).
approach to digital image analysis was reported to deter- Jackman et al. (2010) reported the use of color and
mine moisture content in cooked beef by machine vision wavelet texture features for classification of pork and tur-
(Zheng et al., 2006). The authors reported extraction of key ham based on quality grades. Du and Sun (2006) re-
color features (means and standard deviations) in terms of ported a computer vision method to detect pores and poros-
RGB and HSI (hue, saturation, and intensity). The color in ity in pork ham in an area between 6.73 × 10-3 and 2.02 ×
the sample images was correlated to the moisture content 10-1 mm2. Chmiel et al. (2011b) reported on distinguishing
obtained by chemical analysis, which showed encouraging PSE (pale, soft, exudative) and RFN (reddish pink, firm,
correlation coefficients of 0.56 (for models built by PLSR) normal) in m. longissimus dorsi using color components
and 0.45 (for models built by a neural network algorithm). (lightness parameters L*). Similar studies have been report-
The study showed a potential for machine vision and image ed on classification of pre-sliced ham (based on color) and
processing for detection of chemical contents in meat. its relationship with consumer responses (Iqbal et al.,
Research on the relationship of color to chemical, physi- 2010); classification of pre-sliced pork ham using neural
cal, and biological changes can enhance the accuracy of networks (Valous et al., 2010); changes in the color, chem-
image processing for rapid and non-destructive detection of ical composition, and texture of m. longissimus dorsi in
meat quality attributes. Application of a digital camera with pigs due to management system (Zapotoczny et al., 2014);
an auxiliary lighting system was reported for prediction of evaluation of pork carcass color (Lu et al., 2000); and
troponin-T degradation in beef longissimus using texture measurement of pork belly fat (Uttaro and Zawadski,
features from color images (Sun et al., 2014). The authors 2010). To reduce the difficulty of meat quality classifica-
reported the use of stepwise and support vector machine tion due to high variability and complex color distribution,
(SVM) methods to classify the samples based on degrada- Valous et al. (2009) calibrated the color obtained by a
tion, with satisfactory accuracy. computer vision system using a polynomial transform pro-
A wide range of methods and applications in machine cedure and the transform proposed by the sRGB standard.
58(5): 1371-1386 1377

The authors applied the algorithm to evaluate the color and spectroscopic and machine vision methods. Hyperspectral
presence of pores and fat-connective tissue in pre-sliced imaging has the flexibility of functioning as a spectroscopic
pork, turkey, and chicken ham with encouraging accuracy. method, a machine vision method, or a combination of
both. However, hyperspectral imaging has so far had lim-
Lamb
ited applications in industrial processing because hyper-
Although machine vision and image analysis have prov-
spectral images contain huge quantities of data, which re-
en useful for quantifying the quality of different meat and
duces the detection speed (Cho et al., 2007). Cho et al.
meat products, little research has been reported on meas-
(2007) proposed an innovative method of implementing a
urement of lamb quality attributes. Geometric and texture
hyperspectral imaging system in multispectral imaging
analyses of images obtained from a color digital camera
mode, which enabled fast, online detection without sacrific-
were used to predict cooked lamb tenderness (Chandraratne
ing accuracy. A number of studies have reported on the use
et al., 2006a). Similarly, a machine vision system consist-
of multispectral methods for detection of meat quality at-
ing of three CCD digital color cameras was reported for
tributes. Studies have also been reported on applications of
grading lamb carcasses (Chandraratne et al., 2006b). The
multispectral methods independently.
acquired images included lean area, marbling, subcutane-
Variation in light intensity in the sample is another
ous fat, intramuscular fat, bone, and background. The au-
shortcoming that influences the stability of optical methods
thors reported that image analysis using geometric and tex-
in practical applications. The light intensity is reduced at
ture features is a useful tool for lamb classification. The
the edges of samples, and in the hidden areas of carcasses,
same researchers also reported on statistical and neural
which influences optical signal acquisition. When uneven
network analyses of the extracted image features for classi-
samples are illuminated, large differences in light intensity
fication of lamb carcasses (Chandraratne et al., 2007).
occur at different surface locations of the same sample,
Other Meats which influence the measurement accuracy by the system.
Machine vision and image analysis have been used for A band ratio algorithm can reduce the variation in light
detection of fat content in chicken and turkey (Chmiel et intensity (Park et al., 2006).
al., 2011a), to estimate live weight of broilers for monitor- For rapid, non-destructive, real-time application in in-
ing growth (Mollah et al., 2010), and for inspection of dustrial processing, research must focus on simultaneous
poultry carcasses for wholesomeness (Chao et al., 2000, detection of different quality attributes by the same system.
2007). Implementing multiples systems, each dedicated to detec-
tion of a particular quality attribute, is expensive and tech-
nically challenging. However, only a few studies have re-
CHALLENGES AND FUTURE WORK ported simultaneous detection of multiple quality attributes
Table 1 summarizes research studies that have been re- by the same system (Tao et al., 2012). Studies are thus
ported since 2003 on hyperspectral imaging, Vis/NIR spec- needed on the relationships between optical measurements
troscopy, and machine vision methods for quality detection and the different physical and chemical properties of meat
of meat and meat products. Vis/NIR has proven to be a to help develop algorithms for rapid prediction of multiple
potential tool for non-destructive quality detection of meat quality attributes by the same system.
and meat products. Although a number of studies have Although many studies have reported using different op-
shown that Vis/NIR is useful for assessing physical, chemi- tical instruments and statistical analysis methods, the num-
cal, biological, and sensory quality attributes of meat and ber of samples used in many reported studies is often quite
meat products, few studies on industrial applications have small, which creates a higher probability of overfitting for
been reported. The accuracy, repeatability, and stability of the calibration model. Future studies should thus use larger
optical methods are highly dependent on the surrounding numbers of samples to develop and test calibration models
environment. For example, lighting can be a limiting factor for improved model performance.
for optical methods because the environmental conditions Practical applications of optical technology are needed
in the laboratory are highly controlled, which may not be for quality detection of processed meat and for detection of
the case in industrial settings. The meat samples in a pro- adulterants in raw and processed meat. Except for a few
cessing line often vary greatly in size and quality, which studies, most reported studies have focused on fresh meat.
can affect online measurement accuracy. The distance be- Quality detection of processed meat in packaging and dur-
tween the sample and the optical sensor must also be con- ing storage and transport is important for quality control.
sidered for ensuring precise measurements. An uneven Several studies have reported that packaging conditions
sample surface is another major source of optical measure- have a direct influence on meat quality and shelf life. High-
ment error. Development of conveyor systems equipped pressure processing and modified-atmosphere packaging
with distance sensors to maintain a constant distance be- were reported to influence microorganism activity and pH
tween the sample and the optical sensor can help reduce in pork products, which in turn influenced the flavor of the
measurement error, thus enabling optical system applica- meat (Olmo et al., 2014). Precise packaging conditions are
tions in industry. required to maintain the shelf life and quality of meat and
Hyperspectral imaging has proven to be superior to meat products. Sade et al. (2013) found different types of
spectroscopy and machine vision because it acquires in- enterobacteria in meat and poultry samples packaged in
formation that is far beyond what can be obtained by both modified-atmosphere packaging. Lee (2010) reported seri-
ous effects on the quality and safety of meat products due

Table 1. Application of optical technologies for meat quality detection.[a]
Quality Optical Wavelength Analysis Correlation
Meat Attribute Technology Range Method Coefficient Reference
Beef Protein content HSI 900-1700 nm PLSR 0.86 ElMasry et al. (2013)
Spectroscopy 1000-1800 nm PLSR 0.98 Su et al. (2014)
Moisture content HSI 900-1700 nm PLSR 0.89 ElMasry et al. (2013)
HSI 1000-2500 nm MLR 0.92 Liu et al. (2014b)
Marbling HSI 400-1100 nm - 0.92 Li et al. (2011b)
MV - - Satisfactory Chen and Qin (2008)
MV - - R = 1 and R = 0.96 Jackman et al. (2009)
Fat content HSI 900-1700 nm PLSR 0.84 ElMasry et al. (2013)
HSI 1000-2300 nm PLSR 0.90 Kobayashi et al. (2010)
Scanner 760-1040 nm - - Wold et al. (2011)
Micro-organisms HSI 400-1100 nm MLR 0.96 for TVC Peng et al. (2009)
MSI 405-970 nm PLS-DA, >80%, Panagou et al. (2014)
PLSR 0.78-0.86
Tenderness HSI 400-1100 nm MLR 0.91 Wu et al. (2012)
HSI 900-1700 nm PLSR 0.83 ElMasry et al. (2012a)
HSI 400-1000 nm PCA 96.4% Naganathan et al. (2008a)
HSI 922-1739 nm LDA 83.3% (tough), 75% (tender) Cluff et al. (2013)
Color HSI 900-1700 nm PLSR 0.88 (L*), 0.81 (b*) ElMasry et al. (2012a)
pH HSI 900-1700 nm PLSR 0.73 ElMasry et al. (2012a)
Spectroscopy 350-2500 nm PLSR 90% Reis and Rosenvold (2014)
WHC HSI 910-1700 nm PLSR 0.89 (full spectrum), ElMasry et al. (2011)
0.87 (selected wavelengths)
Pork Protein content Spectroscopy 350-1100 nm PLSR 0.75 Liao et al. (2010)
HSI 900-1700 nm PLSR 0.92 Barbin et al. (2013)
Moisture content Spectroscopy 350-1100 nm PLSR 0.79 Liao et al. (2010)
Marbling Computer image analysis - - - Faucitano et al. (2005)
Fat content Spectroscopy 350-1100 nm PLSR 0.76 Liao et al. (2010)
Micro-organisms Fluorescence 200-900 nm PLSR >0.90 (plate count) Oto et al. (2013)
FT-NIR spectroscopy 10,000-4000 cm-1 SI-PLSR 0.80 (for TVB-N) Cai et al. (2011)
Vis/NIR camera 550-1000 nm PLSR 0.89 (for TVB-N using Wang et al. (2013)
selected wavelengths)
Tenderness FT-NIR spectroscopy 10,000-4000 cm-1 SI-PLSR 0.70 Cai et al. (2011)
HSI and MV - PLSR 0.75 Barbin et al. (2013)
Color Spectroscopy 400-2500 nm M-PLRS Poor Cozzolino et al. (2003)
HSI 400-1000 nm ANN 0.86 Qiao et al. (2007)
pH Spectroscopy 350-1100 nm PLSR 0.82 Liao et al. (2010)
HSI 400-1000 nm ANN 0.55 Qiao et al. (2007)
WHC Spectroscopy 400-2500 nm PLSR <0.62 Maja et al. (2010)
Lamb Protein content HSI 900-1700 nm PLSR 0.63 Kamruzzaman et al. (2012a)
Fat content HSI 900-1700 nm PLSR 0.88 Kamruzzaman et al. (2012a)
Spectroscopy 400-2498 nm - 0.84 Andres et al. (2007)
Moisture content HSI 900-1700 nm PLSR 0.88 Kamruzzaman et al. (2012a)
Spectroscopy 400-2498 nm - 0.67 Andres et al. (2007)
pH HSI 900-1700 nm PLSR 0.65 Andres et al. (2007)
HSI 900-1700 nm PLSR 0.65 Kamruzzaman et al. (2012b)
Color HSI 900-1700 nm PLSR 0.91 Kamruzzaman et al. (2012b)
Tenderness HSI 900-1700 nm PLSR 0.84 (WBSF), Kamruzzaman et al. (2013)
0.69 (sensory tenderness)
Drip loss HSI 900-1700 nm PLSR 0.77 Kamruzzaman et al. (2012b)
Poultry Micro-organisms HSI 930-450 nm PLSR >0.89 Feng et al. (2013)
in chicken HSI 900-1700 nm PLSR 0.97 (for TVC) Feng and Sun (2013b)
Ingesta HSI 400-1000 nm Image 89% (ingesta), Yoon et al. (2011)
processing 98% (fecal material)
Moisture in turkey ham HSI 900-1700 nm PLSR 0.88 Iqbal et al. (2013)
Color in turkey ham HSI 900-1700 nm PLSR 0.74 Iqbal et al. (2013)
pH in turkey ham HSI 900-1700 nm PLSR 0.81 Iqbal et al. (2013)
Adulterants In minced beef FTIR spectroscopy 4000-650 cm-1 PLSR 0.99 Meza-Marquez et al. (2010)
In fresh and frozen- Spectroscopy 400-2500 nm PLS-DA 100% Morsy and Sun (2013)
thawed minced beef
[a]
HSI = hyperspectral imaging, MV = machine vision, MSI = multispectral imaging, PLSR = partial least squares regression, and M-PLSR = modified PLSR.
to manipulation of the packaging material. Packaged meat spoilage during storage and transport. Hence, there is an
products are susceptible to contamination before packaging urgent need for technology that can monitor the quality
and due to the packaging material, and are thus prone to attributes of packaged meat products.
58(5): 1371-1386 1379

Meat adulteration not only causes serious damage to the Argyri, A. A., Panagou, E. Z., Tarantilis, P. A., Polysiou, M., &
reputation of the meat industry but also influences consum- Nychas, G.-J. E. (2010). Rapid qualitative and quantitative
er health and perceptions. Certain meats and meat products detection of beef fillets spoilage based on Fourier transform
that are accepted by consumers of a certain community infrared spectroscopy data and artificial neural networks.
Sensors Actuators B, 145(1), 146-154.
may not be acceptable to another community. Current http://dx.doi.org/10.1016/j.snb.2009.11.052.
methods for detecting meat adulterants rely on HPLC, gas Argyri, A. A., Doulgeraki, A. I., Blana, V. A., Pangou, E. Z., &
chromatography, and liquid chromatography for measuring Nychas, G.-J. E. (2011). Potential of a simple HPLC-based
fatty acid composition, histidine dipeptides, and proteins approach for the identification of the spoilage status of minced
(Ashoor et al., 1988; Carnegie et al., 1983; Verbeke and beef stored at various temperatures and packaging systems. Intl.
Brabander, 1995). These measurement methods are com- J. Food Microbiol., 150(1), 25-33.
plex and cannot be used for routine inspection. It is thus http://dx.doi.org/10.1016/j.ijfoodmicro.2011.07.010.
important to develop non-destructive technology to monitor Ashoor, S. H., Monte, W. G., & Stiles, P. G. (1988). Liquid
meat adulterants, in addition to meat quality detection. Re- chromatographic identification of meats. J. Assoc. Official Anal.
Chem., 71(2), 397-403.
search has been reported on detection of pork, fat trimming, Balasubramanian, S., Panigrahi, S., Logue, C. M., Gu, H., &
and offal adulterants in fresh and frozen-thawed minced Marchello, M. (2009). Neural networks-integrated metal oxide-
beef using NIR spectroscopy (Morsy and Sun, 2013); on based artificial olfactory system for meat spoilage identification.
turkey adulterants in minced beef using UV-VIS, NIR, and J. Food Eng., 91(1), 91-98.
MIR spectroscopy (Alamprese et al., 2013); and on horse http://dx.doi.org/10.1016/j.jfoodeng.2008.08.008.
meat in mixtures with other meats by protein content using Barbin, D. F., ElMasry, G., Sun, D.-W., & Allen, P. (2013). Non-
Raman spectroscopy (Zajac et al., 2014). However, none of destructive determination of chemical composition in intact and
these methods has been reported for practical use, and chal- minced pork using near-infrared hyperspectral imaging. Food
lenges remain in the development of reliable methods for Chem., 138(2-3), 1162-1171.
http://dx.doi.org/10.1016/j.foodchem.2012.11.120.
industrial application. Databases of the chemical properties Barbin, D. F., Kaminishikawahara, C. M., Soares, A. L., Mizubuti,
of adulterants that are suitable for use by optical systems I. Y., Grespan, M., Shimokomaki, M., & Hirooka, E. Y. (2015).
can be useful for industrial applications, as meat adulterants Prediction of chicken quality attributes by near-infrared
are often identified by their chemical properties. spectroscopy. Food Chem., 168, 554-560.
Finally, challenges remain in reducing the size of optical http://dx.doi.org/10.1016/j.foodchem.2014.07.101.
systems, developing hardware suitable for use in industrial Boselli, E., Pacetti, D., Curzi, F., & Frega, N. G. (2008).
environments, and designing control software with an easy- Determination of phospholipid molecular species in pork meat
to-use graphical user interface (GUI). Most hyperspectral by high-performance liquid chromatography tandem mass
imaging, multispectral imaging, and Raman spectroscopic spectrometry and evaporative light scattering detection. Meat
Sci., 78(3), 305-313.
systems are bulky and are limited to laboratory use. Com- http://dx.doi.org/10.1016/j.meatsci.2007.06.017.
pact hardware that is suitable for industrial use will be at- Cai, J., Chen, Q., Wan, X., & Zhao, J. (2011). Determination of
tractive to the meat industry. Studies have reported on the total volatile basic nitrogen (TVB-N) content and Warner-
detection of meat quality attributes using hand-held optical Bratzler shear force (WBSF) in pork using Fourier transform
devices and conveyor belt optical systems (Guo et al., near-infrared (FT-NIR) spectroscopy. Food Chem., 126(1),
2012a, 2012b; Wan et al., 2014; Wang et al., 2012; Yang et 1354-1360. http://dx.doi.org/10.1016/j.foodchem.2010.11.098.
al., 2012; Zhao et al., 2012). More studies on such Carnegie, P. R., Illic, M. Z., Etheridge, M. O., & Collins, M. G.
handheld devices are needed under different experimental (1983). Improved high-performance liquid chromatography
conditions for different types of samples (i.e., different method for analysis of histidine dipeptides anserine, carnosine,
and balenine present in fresh meat. J. Chromatog., 26(1), 153-
sample sizes, breeds, and slaughterhouses) to evaluate their 157. http://dx.doi.org/10.1016/S0021-9673(01)87933-4.
potential for industrial application. Chandraratne, M. R., Samarasinghe, S., Kulasiri, D., &
Bickerstaffe, R. (2006a). Prediction of lamb tenderness using
ACKNOWLEDGEMENTS image surface texture features. J. Food Eng., 77(3), 492-499.
The authors wish to thank the Special Fund for Agro- http://dx.doi.org/10.1016/j.jfoodeng.2005.06.063.
Scientific Research in the Public Interest (Project No. Chandraratne, M. R., Kulasiri, D., Frampton, C., Samarasinghe, S.,
201003008) of China for providing funding support for this & Bickerstaffe, R. (2006b). Prediction of lamb carcass grades
research. using features extracted from lamb chop images. J. Food Eng.,
74(1), 116-124.
http://dx.doi.org/10.1016/j.jfoodeng.2005.02.012.
Chandraratne, M. R., Kulasiri, D., & Samarasinghe, S. (2007).
REFERENCES Classification of lamb carcass using machine vision:
Alamprese, C., Casale, M., Sinelli, N., Lanteri, S., & Casiraghi, E. Comparison of statistical and neural network analyses. J. Food
(2013). Detection of minced beef adulteration with turkey meat Eng., 82(1), 26-34.
by UV-vis, NIR, and MIR spectroscopy. LWT Food Sci. Tech., http://dx.doi.org/10.1016/j.jfoodeng.2007.01.003.
53(1), 225-232. http://dx.doi.org/10.1016/j.lwt.2013.01.027. Chao, K., Park, B., Chen, Y. R., Hruschka, W. R., & Wheaton, F.
Andres, S., Murray, I., Navajas, E. A., Fisher, A. V., Lambe, N. R., W. (2000). Design of a dual-camera system for poultry carcasses
& Bunger, L. (2007). Prediction of sensory characteristics of inspection. Appl. Eng. Agric., 16(5), 581-587.
lamb meat samples by near-infrared reflectance spectroscopy. http://dx.doi.org/10.13031/2013.5291.
Meat Sci., 76(3), 509-516. Chao, K., Yang, C.-C., Chen, Y.-R., Kim, M. S., & Chan, D. E.
http://dx.doi.org/10.1016/j.meatsci.2007.01.011. (2007). Fast line-scan imaging system for broiler carcass

inspection. Sensory Instrum. Food Qual., 1, 62-71. http://dx.doi.org/10.1016/j.jfoodeng.2007.07.026.
http://dx.doi.org/10.1007/s11694-007-9008-y. Davis, R. W., Jones, H. D. T., Collins, A. M., Ricken, J. B.,
Chen, K., & Qin, C. (2008). Segmentation of beef marbling based Sinclair, M. B., Timlin, J. A., & Singh, S. (2014). Label-free
vision threshold. Comput. Elec. Agric., 62(2), 223-230. measurement of algal triacylglyceride production using
http://dx.doi.org/10.1016/j.compag.2008.01.002. fluorescence hyperspectral imaging. Algal Res., 5, 181-189.
Chen, K., Sun, X., Qin, C., & Tang, X. (2010). Color grading of http://dx.doi.org/10.1016/j.algal.2013.11.010.
beef fat by using computer vision and support vector machine. Demos, B. P., Gerrard, D. E., Gao, X., Tan, J., & Mandigo, R. W.
Comput. Elec. Agric., 70(1), 27-32. (1996). Utilization of image processing to quantitate surface
http://dx.doi.org/10.1016/j.compag.2009.08.006. metmyglobin on fresh beef. Meat Sci., 43(3-4), 265-274.
Chen, Q., Cai, J., Wan, X., & Zhao, J. (2011). Application of http://dx.doi.org/10.1016/0309-1740(95)00061-5.
linear/non-linear classification algorithms in discrimination of Dian, P. H. M., Andueza, D., Jestin, M., Prado, I. N., & Prache, S.
pork storage time using Fourier transform near-infrared (FT- (2008). Comparision of visible and near-infrared reflectance
NIR) spectroscopy. LWT Food Sci. Tech., 44(10), 2053-2058. spectroscopy to discriminate between pasture-fed and
Chen, Q., Zhang, Y., Zhao, J., & Hui, Z. (2013). Non-destructive concentrate-fed lamb carcasses. Meat Sci., 80(4), 1157-1164.
measurement of total viable basic nitrogen (TVB-N) content in http://dx.doi.org/10.1016/j.meatsci.2008.05.009.
salted pork in jelly using a hyperspectral imaging technique Dowlati, M., Mohtasebi, S. S., & Guardia, M. D. L. (2012).
combined with efficient hypercube processing algorithms. Anal. Application of machine-vision technique to fish-quality
Methods, 5(22), 6382-6388. assessment. TrAC Trends Anal. Chem., 40, 168-179.
http://dx.doi.org/10.1039/c3ay40436f. http://dx.doi.org/10.1016/j.trac.2012.07.011.
Chmiel, M., Slowinski, M., & Dasiewicz, K. (2011a). Application Du, C.-J., & Sun, D.-W. (2006). Automatic measurement of pores
of computer vision system for estimation of fat content in and porosity in pork ham and their correlation with processing
poultry meat. Food Control, 22(8), 1424-1427. time, water content, and texture. Meat Sci., 72(2), 294-302.
http://dx.doi.org/10.1016/j.foodcont.2011.03.002. http://dx.doi.org/10.1016/j.meatsci.2005.07.016.
Chmiel, M., Slowinski, M., & Dasiewicz, K. (2011b). Lightness of Du, C.-J., & Sun, D.-W. (2008). Multi-classification of pizza using
the color measured by computer image analysis as a factor for computer vision and support vector machine. J. Food Eng.,
assessing the quality of pork meat. Meat Sci., 88(3), 566-570. 86(2), 234-242.
http://dx.doi.org/10.1016/j.meatsci.2011.02.014. http://dx.doi.org/10.1016/j.jfoodeng.2007.10.001.
Cho, B.-K., Chen, Y.-R., & Kim, M. S. (2007). Multispectral Du, C.-J., Sun, D.-W., Jackman, P., & Allen, P. (2008).
detection of organic residue on poultry processing plant Development of hybrid image processing algorithm for
equipment based on hyperspectral reflectance imaging automatic evaluation of intramuscular fat content in beef m.
technique. Comput. Elec. Agric., 57(2), 177-189. longissimus dorsi. Meat Sci., 80(4), 1231-1237.
http://dx.doi.org/10.1016/j.compag.2007.03.008. http://dx.doi.org/10.1016/j.meatsci.2008.05.036.
Cho, B.-K., Kim, M. S., Baek, I.-S., Kim, D.-Y., Lee, W.-H., Kim, Ellis, D. I., Broadhurst, D., & Goodacre, R. (2004). Rapid and
J., Bae. H., & Kim, Y.-S. (2013). Detection of cuticle defects on qualitative detection of the microbial spoilage of beef by Fourier
cherry tomatoes using hyperspectral fluorescence imagery. transform infrared spectroscopy and machine learning. Anal.
Postharvest Biol. Tech., 76, 40-49. Chim. Acta, 514(2), 193-201.
http://dx.doi.org/10.1016/j.postharvbio.2012.09.002. http://dx.doi.org/10.1016/j.aca.2004.03.060.
Clariana, M., Gratacos-Cubarsi, M., Hortos, M., Garcia-Regueiro, J. ElMasry, G., Sun, D.-W., & Allen, P. (2011). Non-destructive
A., & Castellari, M. (2010). Analysis of seven purines and determination of water-holding capacity in fresh beef by using
pyrimidines in pork meat products by ultra-high performance NIR hyperspectral imaging. Food Res. Intl., 44(9), 2624-2633.
liquid chromatography tandem mass spectrometry. J. http://dx.doi.org/10.1016/j.foodres.2011.05.001.
Chromatog. A, 1217(6), 4294-4299. ElMasry, G., Sun, D.-W., & Allen, P. (2012a). Near-infrared
http://dx.doi.org/10.1016/j.chroma.2010.04.033. hyperspectral imaging for predicting color, pH, and tenderness
Cluff, K., Naganathan, G., Subbiah, J., Lu, R., Calkins, C. R., & of fresh beef. J. Food Eng., 110(1), 127-140.
Samal, A. (2008). Optical scattering in beef steak to predict http://dx.doi.org/10.1016/j.jfoodeng.2011.11.028.
tenderness using hyperspectral imaging in the vis-NIR region. ElMasry, G., Barbin, D. F., Sun, D.-W., & Allen, P. (2012b). Meat
Sensory Instrum. Food Qual., 2(3), 189-196. quality evaluation by hyperspectral imaging technique: An
http://dx.doi.org/10.1007/s11694-008-9052-2. overview. Crit. Rev. Food Sci. Nutr., 52(8), 689-711.
Cluff, K., Naganathan, G. K., Subbiah, J., Samal, A., & Calkins, C. http://dx.doi.org/10.1080/10408398.2010.507908.
R. (2013). Optical scattering with hyperspectral imaging to ElMasry, G., Cubero, S., Molto, E., & Blasco, J. (2012c). In-line
classify longissimus dorsi muscle based on beef tenderness sorting of irregular potatoes by using automated computer-based
using multivariate modeling. Meat Sci., 95(1), 42-50. machine vision system. J. Food Eng., 112(1-2), 60-68.
http://dx.doi.org/10.1016/j.meatsci.2013.04.014. http://dx.doi.org/10.1016/j.jfoodeng.2012.03.027.
Cozzolino, D., Barlocco, N., Vadell, A., Ballesteros, F., & Gallieta, ElMasry, G., Sun, D.-W., & Allen, P. (2013). Chemical-free
G. (2003). The use of visible and near-infrared reflectance assessment and mapping of major constituents in beef using
spectroscopy to predict colour on both intact and homogenised hyperspectral imaging. J. Food Eng., 117(2), 235-245.
pork muscle. LWT Food Sci.Tech., 36(2), 195-202. http://dx.doi.org/10.1016/j.jfoodeng.2013.02.016.
Dai, Q., Cheng, J.-H., Sun, D.-W., & Zeng, X.-A. (2014). Potential Erkinbaev, C., Herremans, E., Trong, N. N. D., Jakubczyk, E.,
of hyperspectral imaging for non-invasive determination of Verbpven, P., Nocolai, B., & Saeys, W. (2014). Contactless and
mechanical properties of prawn (Metapenaeus ensis). J. Food non-destructive differentiation of microstructures of sugar foams
Eng., 136, 64-72. by hyperspectral scatter imaging. Innov. Food Sci. Emerg. Tech.,
http://dx.doi.org/10.1016/j.jfoodeng.2014.03.013. 24, 131-137. http://dx.doi.org/10.1016/j.ifset.2013.08.007.
Damez, J.-L., Clerjon, S., Abouelkaram, S., & Lepetit, J. (2008). Everard, C. D., Kim, M. S., & Lee, H. (2014). A comparison of
Beef meat electrical impedance spectroscopy and anisotropy hyperspectral reflectance and fluorescence imaging techniques
sensing for non-invasive early assessment of meat spoilage. J. for detection of contaminants on spinach leaves. J. Food Eng.,
Food Eng., 85(1), 116-122. 143, 139-145. http://dx.doi.org/10.1016/j.jfoodeng.2014.06.042.
58(5): 1371-1386 1381

Faucitano, L., Huff, P., Teuscher, F., Gariepy, C., & Wegner, J. Hatem, I., Tan, J., & Gerrard, D. E. (2003). Determination of animal
(2005). Application of computer image analysis to measure pork skeletal maturity by image processing. Meat Sci., 65(3), 999-
marbling characteristics. Meat Sci., 69(3), 537-543. 1004. http://dx.doi.org/10.1016/S0309-1740(02)00318-2.
http://dx.doi.org/10.1016/j.meatsci.2004.09.010. Haugen, J.-E., Lundby, F., Wold, J. P., & Veberg, A. (2006).
Feng, W., Yao, X., Zhu, Y., Tian, Y. C., & Cao, W. X. (2008). Detection of rancidity in freeze-stored turkey meat using a
Monitoring leaf nitrogen status with hyperspectral reflectance in commercial gas-sensor array system. Sensors Actuators B,
wheat. European J. Agron., 28(3), 394-404. 116(1-2), 78-84. http://dx.doi.org/10.1016/j.snb.2005.12.060.
http://dx.doi.org/10.1016/j.eja.2007.11.005. He, H.-J., Wu, D., & Sun, D.-W. (2014). Rapid and non-destructive
Feng, Y.-Z., & Sun, D.-W. (2013a). Near-infrared hyperspectral determination of drip loss and pH distribution in farmed Atlantic
imaging in tandem with partial least squares regression and salmon (Salmo salar) fillets using visible and near-infrared (vis-
genetics algorithm for non-destructive determination and NIR) hyperspectral imaging. Food Chem., 156(1), 394-401.
visualization of Pseudomonas loads in chicken fillets. Talanta, http://dx.doi.org/10.1016/j.foodchem.2014.01.118.
109, 74-83. http://dx.doi.org/10.1016/j.talanta.2013.01.057. Huang, H., Liu, L., Ngadi, M. O., & Gariepy, C. (2013a). Prediction
Feng, Y.-Z., & Sun, D.-W. (2013b). Determination of total visible of pork marbling scores using pattern analysis techniques. Food
count (TVC) in chicken breast fillets by near-infrared Control, 31(1), 224-229.
hyperspectral imaging and spectroscopic transforms. Talanta, http://dx.doi.org/10.1016/j.foodcont.2012.09.034.
105, 244-249. http://dx.doi.org/10.1016/j.talanta.2012.11.042. Huang, L., Zhao, J., Chen, Q., & Zhang, Y. (2013b). Rapid
Feng, Y.-Z., ElMasry, G., Sun, D.-W., Scannell, A. G. M., Walsh, detection of total viable count (TVC) in pork meat by
D., & Morcy, N. (2013). Near-infrared hyperspectral imaging hyperspectral imaging. Food Res. Intl., 54(1), 821-828.
and partial least squares regression for rapid and reagentless http://dx.doi.org/10.1016/j.foodres.2013.08.011.
determination of Enterobacteriaceae on chicken fillet. Food Iqbal, A., Valous, N. A., Mendoza, F., Sun, D.-W., & Allen, P.
Chem., 138(2-3), 1829-1836. (2010). Classification of pre-sliced pork and turkey ham
http://dx.doi.org/10.1016/j.foodchem.2012.11.040. qualities based on image colour and textural features and their
Forrest, J. C., Morgan, M. T., Borggaard, C., Rasmussen, A. J., relationships with consumer responses. Meat Sci., 84(3), 455-
Jespersen, B. L., & Andersen, J. R. (2000). Development of 465. http://dx.doi.org/10.1016/j.meatsci.2009.09.016.
technology for the early post-mortem prediction of water- Iqbal, A., Sun, D.-W., & Allen, P. (2013). Prediction of moisture,
holding capacity and drip loss in fresh pork. Meat Sci., 55(1), color, and pH in cooked, pre-sliced turkey hams by NIR
115-122. http://dx.doi.org/10.1016/S0309-1740(99)00133-3. hyperspectral imaging system. J. Food Eng., 117(1), 42-51.
Fowler, S. M., Schmidt, H., Ven, R. V. D., Wynn, P., & Hopkins, http://dx.doi.org/10.1016/j.jfoodeng.2013.02.001.
D. L. (2014a). Raman spectroscopy compared against traditional Jackman, P., Sun, D.-W., Du, C.-J., Allen, P., & Downey, G.
predictors of shear force in lamb m. longissimus lumborum. (2008). Prediction of beef eating quality from color, marbling,
Meat Sci., 98(4), 652-656. and wavelet texture features. Meat Sci., 80(4), 1273-1281.
http://dx.doi.org/10.1016/j.meatsci.2014.06.042. http://dx.doi.org/10.1016/j.meatsci.2008.06.001.
Fowler, S. M., Schmidt, H., Ven, R. V. D., Wynn, P., & Hopkins, Jackman, P., Sun, D.-W., & Allen, P. (2009). Automatic selection
D. L. (2014b). Predicting tenderness of fresh ovine of beef longissimus dorsi muscle and marbling by an adaptable
semimembranosus using Raman spectroscopy. Meat Sci., 97(4), algorithm. Meat Sci., 83(2), 187-194.
597-601. http://dx.doi.org/10.1016/j.meatsci.2014.02.018. http://dx.doi.org/10.1016/j.meatsci.2009.03.010.
Gomez, A. H., He, Y., & Pereira, A. G. (2006). Non-destructive Jackman, P., Sun, D.-W., Allen, P., Valous, N. A., Mendoza, F., &
measurement of acidity, soluble solids, and firmness of Satsuma Ward, P. (2010). Identification of important image features for
mandarin using vis/NIR spectroscopy technique. J. Food Eng., pork and turkey ham classification using color and wavelet
77(2), 313-319. texture features and genetic selection. Meat Sci., 84(4), 711-717.
http://dx.doi.org/10.1016/j.jfoodeng.2005.06.036. http://dx.doi.org/10.1016/j.meatsci.2009.10.030.
Gomez-De-Anda, F., Dorantes-Alvarez, L., Gallardo-Velazquez, T., Jamshidi, B., Minaei, S., Mohajerani, E., & Ghassemian, H. (2012).
Osorio-Revilla, G., Calderon-Dominguez, G., Labat, P. M., & Reflectance vis/NIR spectroscopy for nondestructive taste
de-la-Rosa-Arana, J.-L. (2012). Determination of Trichinella characterization of valencia oranges. Comput. Elec. Agric., 85,
spiralis in pig muscles using mid-Fourier transform infrared 64-69. http://dx.doi.org/10.1016/j.compag.2012.03.008.
spectroscopy (mid-FTIR) with attenuated total reflectance Ji, W., Zhao, D., Cheng, G., Xu, B., Zhang, Y., & Wang, J. (2012).
(ATR) and soft independent modeling of class analogy Automatic recognition vision system guided for apple harvesting
(SIMCA). Meat Sci., 91(3), 240-246. robot. Comput. Elec. Eng., 38(5), 1186-1195.
http://dx.doi.org/10.1016/j.meatsci.2012.01.019. http://dx.doi.org/10.1016/j.compeleceng.2011.11.005.
Grau, R., Sanchez, A. J., Giron, J., Iborra, E., Fuentes, A., & Barat, Jia, J., Schinckel, A. P., Forrest, J. C., Chen, W., & Wagner, J. R.
J. M. (2011). Nondestructive assessment of freshness in (2010). Prediction of lean and fat composition in swine carcasses
packaged sliced chicken breast using SW-NIR spectroscopy. from ham area measurement with image analysis. Meat Sci.,
Food Res. Intl., 44(1), 331-337. 85(2), 240-244. http://dx.doi.org/10.1016/j.meatsci.2010.01.005.
http://dx.doi.org/10.1016/j.foodres.2010.10.011. Kamruzzaman, M., ElMasry, G., Sun, D.-W., & Allen, P. (2011).
Guo, H., Jiang, F. C., Peng, Y. K., Wan, L., & Yang, W. L. (2012a). Application of NIR hyperspectral imaging for discrimination of
Development of a rapid beef quality detection device. J. Food lamb muscles. J. Food Eng., 104(3), 332-340.
Saf. Qual., 3(6), 617-620. http://dx.doi.org/10.1016/j.jfoodeng.2010.12.024.
Guo, H., Peng, Y. K., Jiang, F. C., & Yang, W. L. (2012b). Kamruzzaman, M., ElMasry, G., Sun, D.-W., & Allen, P. (2012a).
Development of conveyable beef-marbling detection system. Non-destructive prediction and visualization of chemical
Trans. Chinese Soc. Agric. Mach., 43, 207-210. composition in lamb meat using NIR hyperspectral imaging and
Guy, F., Prache, S., Thomas, A., Bauchart, D., & Andueza, D. multivariate regression. Innov. Food Sci. Emerg. Tech., 16, 218-
(2011). Prediction of lamb meat fatty acid composition using 226. http://dx.doi.org/10.1016/j.ifset.2012.06.003.
near-infrared reflectance spectroscopy (NIRS). Food Chem., Kamruzzaman, M., ElMasry, G., Sun, D.-W., & Allen, P. (2012b).
127(3), 1280-1286. Prediction of some quality attributes of lamb meat using near-
http://dx.doi.org/10.1016/j.foodchem.2011.01.084. infrared hyperspectral imaging and multivariate analysis. Anal.

Chim. Acta, 714, 57-67. Li, Y., Shan, J., Peng, Y., & Gao, X. (2011b). Nondestructive
http://dx.doi.org/10.1016/j.aca.2011.11.037. assessment of beef-marbling grade using hyperspectral imaging
Kamruzzaman, M., ElMasry, G., Sun, D.-W., & Allen, P. (2013). technology. In Proc. 2011 Intl. Conf. New Technology of Agric.
Non-destructive assessment of instrumental and sensory Eng. (ICAE) (pp. 779-783). Piscataway, N.J.: IEEE.
tenderness of lamb meat using NIR hyperspectral imaging. Food Li, Y., Zhang, L., Peng, Y., Tang, X., Chao, K., & Dhakal, S.
Chem., 141(1), 389-396. (2011c). Hyperspectral technique techniques for determination
http://dx.doi.org/10.1016/j.foodchem.2013.02.094. of pork freshness attributes. In Proc. SPIE 8027: Sensing for
Khazaei, N. B., Tavakoli, T., Ghassemian, H., Khoshtaghaza, M. Agricultural and Food Quality and Safety III. Bellingham,
H., & Banakar, A. (2013). Applied machine vision and artificial Wash.: SPIE. http://dx.doi.org/10.1117/12.883652
neural network for modeling and controlling of the grape drying Liao, Y.-T., Fan, Y.-X., & Cheng, F. (2010). On-line prediction of
process. Comput. Elec. Agric., 98, 205-213. fresh pork quality using visible/near-infrared reflectance
http://dx.doi.org/10.1016/j.compag.2013.08.010. spectroscopy. Meat Sci., 86(4), 901-907.
Kimiya, T., Sivertsen, A. H., & Heia, K. (2013). Vis/NIR http://dx.doi.org/10.1016/j.meatsci.2010.07.011.
spectroscopy for non-destructive freshness assessment of Liao, Y., Fan, Y., & Cheng, F. (2012). On-line prediction of pH
Atlantic salmon (Salmo salar L.) fillets. J. Food Eng., 116(3), values in pork using visible/near-infrared spectroscopy with
758-764. http://dx.doi.org/10.1016/j.jfoodeng.2013.01.008. wavelet de-noising and variable selection methods. J. Food
Kobayashi, K.-I., Matsui, Y., Maebuchi, Y., Toyota, T., & Eng., 109(4), 668-675.
Nakauchi, S. (2010). Near-infrared spectroscopy and http://dx.doi.org/10.1016/j.jfoodeng.2011.11.029.
hyperspectral imaging for prediction and visualisation of fat and Lin, H., Zhao, J., Sun, L., Chen, Q., & Zhou, F. (2011). Freshness
fatty acid content in intact raw beef cuts. J. Near Infrared measurement of eggs using infrared (NIR) spectroscopy and
Spectros., 18(5), 301-315. http://dx.doi.org/10.1255/jnirs.896. multivariate data analysis. Innov. Food Sci. Emerg. Tech., 12(2),
Kodogiannis, V. S., Pachidis, T., & Kontogianni, E. (2014). An 182-186. http://dx.doi.org/10.1016/j.ifset.2011.01.008.
intelligent based decision support system for the detection of Liu, D., Ma, J., Sun, D.-W., Pu, H., Gao, W., Qu, J., & Zeng, X.-A.
meat spoilage. Eng. Appl. Artif. Intel., 34, 23-36. (2014a). Prediction of color and pH of salted porcine meats
http://dx.doi.org/10.1016/j.engappai.2014.05.001. using visible and near-infrared hyperspectral imaging. Food
Kurtulmus, F., & Kavdir, I. (2014). Detecting corn tassels using Bioproc. Tech., 7(11), 3100-3108.
computer vision and support vector machines. Expert Syst. http://dx.doi.org/10.1007/s11947-014-1327-5.
Appl., 41(16), 7390-7397. Liu, D., Sun, D.-W., Qu, J., Zeng, X.-A., Pu, H., & Ma, J. (2014b).
http://dx.doi.org/10.1016/j.eswa.2014.06.013. Feasibility of using hyperspectral imaging to predict moisture
Larrain, R. E., Schaefer, D. M., & Reed, J. D. (2008). Use of digital content of porcine meat during salting process. Food Chem.,
images to estimate CIE color coordinates of beef. Food Res. 152, 197-204.
Intl., 41(4), 380-385. http://dx.doi.org/10.1016/j.foodchem.2013.11.107.
http://dx.doi.org/10.1016/j.foodres.2008.01.002. Lu, J., Tan, J., Shatadal, P., & Gerrard, D. E. (2000). Evaluation of
Lawrie, R. A. (1985). Chapter 5: The conversion of muscle to meat. pork color by using computer vision. Meat Sci., 56(1), 57-60.
In Meat Science (pp. 74-91). 4th Ed. Oxford, U.K.: Pergamon http://dx.doi.org/10.1016/S0309-1740(00)00020-6.
Press. Lu, R., & Ariana, D. (2013). Detection of fruit fly infestation in
Lee, K. T. (2010). Quality and safety aspects of meat products as pickling cucumbers using a hyperspectral
affected by various physical manipulations of packaging reflectance/transmittance imaging system. Postharvest
materials. Meat Sci., 86(1), 138-150. Biol.Tech., 81, 44-50.
http://dx.doi.org/10.1016/j.meatsci.2010.04.035. http://dx.doi.org/10.1016/j.postharvbio.2013.02.003.
Leemans, V., Magein, H., & Destain, M. F. (1998). Defects Lu, R., & Peng, Y. K. (2006). Hyperspectral scattering for assessing
segmentation on ‘Golden Delicious’ apples by using color peach fruit firmness. Biosyst. Eng., 93(2), 161-171.
machine vision. Comput. Elec. Agric., 20(2), 117-130. http://dx.doi.org/10.1016/j.biosystemseng.2005.11.004.
http://dx.doi.org/10.1016/S0168-1699(98)00012-X. Ma, F., Yao, J., Xie, T., Liu, C., Chen, W., Chen, C., & Zeng, L.
Leiva-Valenzuela, G. A., Lu, R., & Aguilera, J. M. (2014). (2014). Multispectral imaging for rapid and non-destructive
Assessment of internal quality of blueberries using hyperspectral determination of aerobic plate count (APC) in cooked pork
transmittance and reflectance images with whole spectra or sausages. Food Res. Intl., 62, 902-908.
selected wavelengths. Innov. Food Sci. Emerg. Tech., 24, 2-13. http://dx.doi.org/10.1016/j.foodres.2014.05.010.
http://dx.doi.org/10.1016/j.ifset.2014.02.006. Maja, P., Martin, S., Dejan, S., & Marjeta, C.-P. (2010). Application
Li, C., Peng, Y., Tang, X., & Sasao, A. (2013). A portable system of near-infrared spectroscopy to predict chemical composition of
for prediction of pork freshness parameters using multispectral meat and meat products. Tehnologija Mesa, 51(2), 133-142.
imaging technology. ASABE Paper No. 131587037. St. Joseph, Marchi, M. D., Riovanto, R., Penasa, M., & Cassandro, M. (2012).
Mich.: ASABE. At-line prediction of fatty acid profile in chicken breast using
Li, H., Chen, Q., Zhao, J., & Wu, M. (2015). Nondestructive near-infrared reflectance spectroscopy. Meat Sci., 90(3), 653-
detection of total viable basic nitrogen (TVB-N) content in pork 657. http://dx.doi.org/10.1016/j.meatsci.2011.10.009.
meat by integrating hyperspectral imaging and colorimetric Marchi, M. D., Penasa, M., Cecchinato, A., & Bittante, G. (2013).
sensor combined with a non-linear data fusion. LWT Food Sci. The relevance of different near-infrared technologies and
Tech., 63(1), 268-274. samples treatment for prediction meat quality traits in
http://dx.doi.org/10.1016/j.lwt.2015.03.052. commercial beef cuts. Meat Sci., 93(2), 329-335.
Li, J., Tan, J., & Shatadal, P. (2001). Classification of tough and http://dx.doi.org/10.1016/j.meatsci.2012.09.013.
tender beef by image texture analysis. Meat Sci., 57(4), 341-346. Mehta, S. S., & Burks, T. F. (2014). Vision-based control of robotic
http://dx.doi.org/10.1016/S0309-1740(00)00105-4. manipulator for citrus harvesting. Comput. Elec. Agric., 102,
Li, P., Lee, S.-H., & Hsu, H.-Y. (2011a). Review on fruit harvesting 146-158. http://dx.doi.org/10.1016/j.compag.2014.01.003.
method for potential use of automatic fruit harvesting systems. Meza-Marquez, O. G., Gallardo-Velazquez, T., & Osorio-Revilla,
Procedia Eng., 23, 351-366. G. (2010). Application of mid-infrared spectroscopy with
http://dx.doi.org/10.1016/j.proeng.2011.11.2514. multivariate analysis and soft independent modeling of class
58(5): 1371-1386 1383

analogies (SIMCA) for the detection of adulterants in minced (2003). Early prediction of water-holding capacity in meat by
beef. Meat Sci., 86(2), 511-519. multivariate vibrational spectroscopy. Meat Sci., 65(1), 581-592.
http://dx.doi.org/10.1016/j.meatsci.2010.05.044. http://dx.doi.org/10.1016/S0309-1740(02)00251-6.
Mollah, M. B. R., Hasan, M. A., Salam, M. A., & Ali, M. A. Pena, F., Molina, A., Aviles, C., Juarez, M., & Horcada, A. (2013).
(2010). Digital image analysis to estimate the live weight of Marbling in the longissimus thoracis muscle from lean cattle
broiler. Comput. Elec. Agric., 72(1), 48-52. breeds: Computer image analysis of fresh versus stained meat
http://dx.doi.org/10.1016/j.compag.2010.02.002. samples. Meat Sci., 95(3), 512-519.
Morsy, N., & Sun, D.-W. (2013). Robust linear and non-linear http://dx.doi.org/10.1016/j.meatsci.2013.05.036.
models of NIR spectroscopy for detection and quantification of Peng, Y. K., & Lu, R. (2007). Prediction of apple fruit firmness and
adulterants in fresh and frozen-thawed minced beef. Meat Sci., soluble solids content using characteristics of multispectral
93(2), 292-302. http://dx.doi.org/10.1016/j.meatsci.2012.09.005. scattering images. J. Food Eng., 82(2), 142-152.
Naganathan, G. K., Grimes, L. M., Subbiah, J., Calkins, C. R., http://dx.doi.org/10.1016/j.jfoodeng.2006.12.027.
Samal, A., & Meyer, G. E. (2008a). Partial least squares analysis Peng, Y., Zhang, J., Wu, J., & Hang, H. (2009). Hyperspectral
of near-infrared hyperspectral images for beef tenderness scattering profiles for prediction of the microbial spoilage of
prediction. Sensing Instrum. Food Qual. Saf., 2(3), 178-188. beef. In Proc. SPIE 7315: Sensing for Agricultural and Food
http://dx.doi.org/10.1007/s11694-008-9051-3. Quality and Safety. Bellingham, Wash.: SPIE.
Naganathan, G. K., Grimes, L. M., Subbiah, J., Calkins, C. R., http://dx.doi.org/10.1117/12.819424.
Samal, A., & Meyer, G. E. (2008b). Visible/near-infrared Penman, D. W. (2001). Determination of stem and calyx location on
hyperspectral imaging for beef tenderness prediction. Comput. apples using automatic visual inspection. Comput. Elec. Agric.,
Elec. Agric., 64(2), 225-233. 33(1), 7-18. http://dx.doi.org/10.1016/S0168-1699(01)00172-7.
http://dx.doi.org/10.1016/j.compag.2008.05.020. Perez-Marin, D., Sanz, E. D., Guerrero-Ginel, J. E., & Garrido-
Noh, H. K., & Lu, R. (2007). Hyperspectral laser-induced Varo, A. (2009). A feasibility study on the use of near-infrared
fluorescence imaging for assessing apple fruit quality. spectroscopy for prediction of the fatty acid profile in live
Postharvest Biol. Tech., 43, 193-201. Iberian pigs and carcasses. Meat Sci., 83(4), 627-633.
http://dx.doi.org/10.1016/j.postharvbio.2006.09.006. http://dx.doi.org/10.1016/j.meatsci.2009.07.012.
Ohk, S.-H., & Bhunia, A. K. (2013). Multiplex fiber optic biosensor Perisic, N., Afseth, N. K., Ofstad, R., Narum, B., & Kohler, A.
for detection of Listeria monocytogenes, Escherichia coli (2013). Chacterizing salt substitution in beef meat processing by
O157:H7, and Salmonella enterica from ready-to-eat meat vibrational spectroscopy and sensory analysis. Meat Sci., 95(3),
samples. Food Microbiol., 33(2), 166-171. 576-585. http://dx.doi.org/10.1016/j.meatsci.2013.05.043.
http://dx.doi.org/10.1016/j.fm.2012.09.013. Phongpa-Ngain, P., Aggrey, S. E., Mulligan, J. H., & Wicker, L.
Olmo, A. D., Calzada, J., & Nunez, M. (2014). Effect of high- (2014). Raman spectroscopy to assess water-holding capacity in
pressure processing and modified atmosphere packaging on muscle from fast and slow growing broilers. LWT Food Sci.
safety and quality of sliced and quality of sliced ready-to-eat Tech., 57(2), 696-700.
“lacón”, a cured-cooked pork meat product. Innov. Food Sci. http://dx.doi.org/10.1016/j.lwt.2014.01.028.
Emerg. Tech., 23, 25-32. Pierna, J. A. F., Vermeulen, P., Amand, O., Tossens, A., Dardenne,
http://dx.doi.org/10.1016/j.ifset.2014.03.003. P., & Baeten, V. (2012). NIR hyperspectral imaging
Oto, N., Oshita, S., Makino, Y., Kawagoe, Y., Sugiyama, J., & spectroscopy and chemometrics for the detection of undesirable
Yoshimura, M. (2013). Non-destructive evaluation of ATP substances in food and feed. Chemomet. Intel. Lab. Syst., 117,
content and plate count on pork meat surface by fluorescence 233-239. http://dx.doi.org/10.1016/j.chemolab.2012.02.004.
spectroscopy. Meat Sci., 93(3), 579-585. Prieto, N., Andres, S., Giraldez, F. J., Mantecon, A. R., & Lavin, P.
http://dx.doi.org/10.1016/j.meatsci.2012.11.010. (2006). Potential use of near-infrared reflectance spectroscopy
Pan, L., Zhu, Q., Lu, R., & McGrath, J. M. (2015). Determination (NIRS) for the estimination of chemical composition of oxen
of sucrose content in sugar beet by portable visible near-infrared meat samples. Meat Sci., 74(3), 487-496.
spectroscopy. Food Chem., 167, 264-271. http://dx.doi.org/10.1016/j.meatsci.2006.04.030.
http://dx.doi.org/10.1016/j.foodchem.2014.06.117. Prieto, N., Ross, D., Navajas, E. A., Nute, G. R., Richardson, R. I.,
Panagou, E. Z., Papadopoulou, O., Carstensen, J. M., & Nychas, G.- Hyslop, J. J., Simm, G., & Roehe, R. (2009a). On-line
J. E. (2014). Potential of multispectral imaging technology for application of visible and near-infrared reflectance spectroscopy
rapid and non-destructive determination of the microbiological to predict chemical-physical and sensory characteristics of beef
quality of beef fillets during aerobic storage. Intl. J. Food quality. Meat Sci., 83(1), 96-103.
Microbiol., 174, 1-11. http://dx.doi.org/10.1016/j.meatsci.2009.04.005.
http://dx.doi.org/10.1016/j.ijfoodmicro.2013.12.026. Preito, N., Roehe, R., Lavin, P., Batten, G., & Andres, S. (2009b).
Panigrahi, S., Balasubramanian, S., Gu, H., Logue, C. M., & Application of near-infrared reflectance spectroscopy to predict
Marchello, M. (2006). Design and development of a metal oxide meat and meat products quality: A review. Meat Sci., 83(2),
based electronic nose for spoilage classification of beef. Sensors 175-186. http://dx.doi.org/10.1016/j.meatsci.2009.04.016.
Actuators B, 119(1), 2-14. Prieto, N., Lopez-Campos, O., Aalhus, J. L., Dugan, M. E. R.,
http://dx.doi.org/10.1016/j.snb.2005.03.120. Juarez, M., & Uttaro, B. (2014). Use of near-infrared
Papadopoulou, O., Panagou, E. Z., Tassou, C. C., & Nychas, G.-J. spectroscopy for estimating meat chemical composition, quality
(2011). Contribution of Fourier transform infrared (FTIR) traits, and fatty acid content from cattle fed sunflower or
spectroscopy data on the quantitative determination of minced flaxseed. Meat Sci., 98(2), 279-288.
pork meat spoilage. Food Res. Intl., 44(10), 3264-3271. http://dx.doi.org/10.1016/j.meatsci.2014.06.005.
http://dx.doi.org/10.1016/j.foodres.2011.09.012. Pu, H., Sun, D.-W., Ma, J., Liu, D., & Kamruzzaman, M. (2014).
Park, B., Lawrence, K. C., Windham, W. R., & Smith, D. P. (2006). Hierarchical variable selection for predicting chemical
Performance of hyperspectral imaging system for poultry constituents in lamb meats using hyperspectral imaging. J. Food
surface fecal contaminant detection. J. Food Eng., 75(3), 340- Eng., 143, 44-52.
348. http://dx.doi.org/10.1016/j.jfoodeng.2005.03.060. http://dx.doi.org/10.1016/j.jfoodeng.2014.06.025.
Pedersen, D. K., Morel, S., Andersen, H. J., & Engelsen, S. B. Qiao, J., Wang, N., Ngadi, M. O., Gunenc, A., Monroy, M.,

Gariepy, C., & Prasher, S. O. (2007). Prediction of drip-loss, pH, Sun, B. (2014). Development of near-infrared reflectance
and color for pork using a hyperspectral imaging technique. spectroscopy to predict chemical composition with a wide range
Meat Sci., 76(1), 1-8. of variability in beef. Meat Sci., 98(2), 110-114.
http://dx.doi.org/10.1016/j.meatsci.2006.06.031. http://dx.doi.org/10.1016/j.meatsci.2013.12.019.
Qin, J., Chao, K., & Kim, M. S. (2011). Investigation of Raman Sun, S., Gua, B., Wei, Y., & Fan, M. (2012). Classification of
chemical imaging for detection of lycopene changes in tomatoes geographical origins and prediction of δ13C and δ15N values of
during postharvest ripening. J. Food Eng., 107(3-4), 277-288. lamb meat by near-infrared reflectance spectroscopy. Food
http://dx.doi.org/10.1016/j.jfoodeng.2011.07.021. Chem., 135(2), 508-514.
Qin, J., Chao, K., Kim, M. S., Lu, R., & Burks, T. F. (2013). http://dx.doi.org/10.1016/j.foodchem.2012.05.004.
Hyperspectral and multispectral imaging for evaluating food Sun, X., Chen, K. J., Berg, E. P., Newman, D. J., Schwartz, C. A.,
safety and quality. J. Food Eng., 118(2), 157-171. Keller, W. L., & Carlin, K. R. M. (2014). Prediction of troponin-
http://dx.doi.org/10.1016/j.jfoodeng.2013.04.001. T degradation using color image texture features in 10 d aged
Reis, M. M., & Rosenvold, K. (2014). Early on-line classification of beef longissimus steaks. Meat Sci., 96(2A), 837-842.
beef carcasses based on ultimate pH by near-infrared http://dx.doi.org/10.1016/j.meatsci.2013.09.012.
spectroscopy. Meat Sci., 96(2A), 862-869. Tao, F., & Peng, Y. K. (2014). A method for nondestructive
ttp://dx.doi.org/10.1016/j.meatsci.2013.10.016. prediction of pork meat quality and safety attributes by
Riovanto, R., Marchi, M. D., Cassandro, M., & Penasa, M. (2012). hyperspectral imaging technique. J. Food Eng., 126, 98-106.
Use of near-infrared transmittance spectroscopy to predicts fatty http://dx.doi.org/10.1016/j.jfoodeng.2013.11.006.
acid composition of chicken meat. Food Chem., 134(4), 2459- Tao, F., & Peng, Y. K. (2015). A nondestructive method for
2464. http://dx.doi.org/10.1016/j.foodchem.2012.04.038. prediction of total viable count in pork meat by hyperspectral
Ripoll, G., Alberti, P., Panea, B., Olleta, J. L., & Sanudo, C. (2008). scattering imaging. Food Bioproc. Tech., 8(1), 17-30.
Near-infrared reflectance spectroscopy for predicting chemical, http://dx.doi.org/10.1007/s11947-014-1374-y.
instrumental, and sensory quality of beef. Meat Sci., 80(3), 697- Tao, F., Peng, Y., Li, Y., Chao, K., & Dhakal, S. (2012).
702. http://dx.doi.org/10.1016/j.meatsci.2008.03.009. Simultaneous determination of tendernes and Escherichia coli
Ruiz-Ramirez, J., Arnau, J., Serra, X., & Gou, P. (2005). contamination of pork using hyperspectral scattering technique.
Relationship between water content, NaCl content, pH, and Meat Sci., 90(3), 851-857.
texture parameters in dry-cured muscles. Meat Sci., 70(4), 579- http://dx.doi.org/10.1016/j.meatsci.2011.11.028.
587. http://dx.doi.org/10.1016/j.meatsci.2005.02.007. Tong, J. H., Li, J. B., & Jiang, H. Y. (2013). Machine vision
Sade, E., Murros, A., & Bjorkroth, J. (2013). Predominant techniques for the evaluation of seedling quality based on leaf
enterobacteria on modified-atmosphere packaged meat and area. Biosyst. Eng., 115(3), 369-379.
poultry. Food Microbiol., 34(2), 252-258. http://dx.doi.org/10.1016/j.biosystemseng.2013.02.006.
http://dx.doi.org/10.1016/j.fm.2012.10.007. Uttaro, B., & Zawadski, S. (2010). Prediction of pork belly fatness
Sahar, A., & Dufour, E. (2014). Use of Fourier transform-infrared from the intact primal cut. Food Control, 21(10), 1394-1401.
spectroscopy to predict spoilage bacteria on aerobically stored http://dx.doi.org/10.1016/j.foodcont.2010.03.012.
chicken breast fillets. LWT Food Sci. Tech., 56(2), 315-320. Valous, N. A., Mendoza, F., Sun, D.-W., & Allen, P. (2009). Color
http://dx.doi.org/10.1016/j.lwt.2013.12.009. calibration of a laboratory computer vision system for quality
Sahar, A., Boubellouta, T., & Dufour, E. (2011). Synchronous evaluation of pre-sliced hams. Meat Sci., 81(1), 132-141.
front-face fluorence spectroscopy as a promising tool for the http://dx.doi.org/10.1016/j.meatsci.2008.07.009.
rapid determination of spoilage bacteria in chicken breast fillets. Valous, N. A., Mendoza, F., Sun, D.-W., & Allen, P. (2010).
Food Res. Intl., 44(1), 471-480. Supervised neural network classification of pre-sliced cooked
http://dx.doi.org/10.1016/j.foodres.2010.09.006. pork ham images using quanternionic singular values. Meat Sci.,
Serra, X., Ruiz-Ramirez, J., Arnau, J., & Gou, P. (2005). Texture 84(3), 422-430. http://dx.doi.org/10.1016/j.meatsci.2009.09.011.
parameters of dry-cured ham m. biceps femoris samples dried at Vendrell-Pascuas, S., Castellote-Bargallo, A. I., & Lopez-Sabater,
different levels as a function of water activity and water content. M. C. (2000). Determination of inulin in meat products by high-
Meat Sci., 69(2), 249-254. performance liquid chromatography with refractive index
http://dx.doi.org/10.1016/j.meatsci.2004.07.004. detection. J. Chromatog. A, 881(1-2), 591-597.
Shalaby, A. R., Salama, N. A., Abou-Raya, S. H., Emam, W. H., & http://dx.doi.org/10.1016/S0021-9673(00)00351-4.
Mehaya, F. M. (2011). Validation of HPLC method for Venel, C., Mullen, A. M., Downey, G., & Troy, D. J. (2001).
determination of tetracycline residues in chicken meat and liver. Prediction of tenderness and other quality attributes of beef by
Food Chem., 124(4), 1660-1666. near-infrared reflectance spectroscopy between 750 and 1100
http://dx.doi.org/10.1016/j.foodchem.2010.07.048. nm: Further studies. J. Near Infrared Spectros., 9(3), 185-198.
Sharifzadeh, S., Clemmensen, L. H., Borggaard, C., Stoier, S., & http://dx.doi.org/10.1255/jnirs.305.
Ersboll, B. K. (2014). Supervised features selection for linear Verbeke, R., & Brabander, H. D. (1995). Differentiation of meat
and non-linear regression of L*a*b* color from multispectral species in processed meat products through identification of
image of meat. Eng. Appl. Artif. Intel., 27, 211-227. animal fat species. In R. L. S. Patterson (Ed.), Biochemical
http://dx.doi.org/10.1016/j.engappai.2013.09.004. Identification of Meat Species (pp. 145-154). New York, N.Y.:
Siedliska, A., Baranowski, P., & Mazurek, W. (2014). Elsevier.
Classification models of bruise and cultivar detection on the Viljoen, M., Hoffman, L. C., & Brand, T. S. (2007). Prediction of
basis of hyperspectral imaging data. Comput. Elec. Agric., 106, the chemical composition of mutton with near-infrared
66-74. http://dx.doi.org/10.1016/j.compag.2014.05.012. reflectance spectroscopy. Small Ruminant Res., 69(1-3), 88-94.
Sowoidnich, K., Schmidt, H., Maiwald, M., Sumpf, B., & http://dx.doi.org/10.1016/j.smallrumres.2005.12.019.
Kronfeldt, H.-D. (2010). Application of diode-laser Raman Wan, L., Peng, Y., & Wang, C. (2014). Design of portable device
spectroscopy for in situ investigation of meat spoilage. Food for rapid and nondestructive detection of fresh meat quality.
Bioproc. Tech., 3(6), 878-882. http://dx.doi.org/10.1007/s11947- Trans. CSAE, 30(7), 243-249.
010-0360-2. Wang, C., Li, X., Wang, W., Feng, Y., Zhou, Z., & Zhan, H.
Su, H., Sha, K., Zhang, L., Zhang, Q., Xu, Y., Zhang, R., Li, H., & (2011). Recognition of worm-eaten chestnuts based on machine
58(5): 1371-1386 1385

vision. Math. Comput. Model., 54(3-4), 888-894. reflectance imaging. Sensing Instrum. Food Qual. Saf., 2(3),
http://dx.doi.org/10.1016/j.mcm.2010.11.012. 197-207. http://dx.doi.org/10.1007/s11694-008-9044-2.
Wang, M. W., Peng, Y. K., Wang, C. P., & Tang, X. Y. (2012). Yoon, S. C., Park, B., Lawrence, K. C., Windham, W. R., &
Design of control system of putrid pork detection using visible Heitschmidt, G. W. (2011). Line-scan hyperspectral imaging
and near-infrared spectral technology. J. Food Saf. Qual., 3(6), system for real-time inspection of poultry carcasses with fecal
585-588. material and ingesta. Comput. Elec. Agric., 79(2), 159-168.
Wang, X., Zhao, M., Ju, R., Song, Q., Hua, D., Wang, C., & Chen, http://dx.doi.org/10.1016/j.compag.2011.09.008.
T. (2013). Visualizing quantitatively the freshness of intact fresh Yu, Z., Zhao, Y., Li, X., Shao, Y., Zhu, F., & He, Y. (2014).
pork using acousto-optical tunable filter-based visible/near- Identification of crack features in fresh jujube using vis/NIR
infrared spectral imagery. Comput. Elec. Agric., 99, 41-53. hyperspectral imaging combined with image processing.
http://dx.doi.org/10.1016/j.compag.2013.08.025. Comput. Elec. Agric., 103, 1-10.
Wold, J. P., O’Farrell, M., Hoy, M., & Tschudi, J. (2011). On-line http://dx.doi.org/10.1016/j.compag.2014.01.016.
determination and control of fat content in batches of beef Zajac, A., Hanuza, J., & Dyminska, L. (2014). Raman spectroscopy
trimmings by NIR imaging spectroscopy. Meat Sci., 89(3), 317- in determination of horse meat content in mixture with other
324. http://dx.doi.org/10.1016/j.meatsci.2011.05.001. meats. Food Chem., 156, 333-338.
Wu, D., Wang, S., Wang, N., Nie, P., He, Y., Sun, D.-W., & Yao, J. http://dx.doi.org/10.1016/j.foodchem.2014.02.002.
(2013). Application of time series hyperspectral imaging (TS- Zamora-Rojas, E., Aernouts, B., Garrido-Varo, A., Perez-Marin, D.,
HSI) for determining water distribution within beef and spectral Guerrero-Ginel, J. E., & Saeys, W. (2013). Double integrating
kinetic analysis during dehydration. Food Bioproc. Tech., 6(11), sphere measurement for estimating optical properties of pig
2943-2958. http://dx.doi.org/10.1007/s11947-012-0928-0. subcutaneous adipose tissue. Innov. Food Sci. Emerg. Tech., 19,
Wu, J., Peng, Y., Li, Y., Wang, W., Chen, J., & Dhakal, S. (2012). 218-226. http://dx.doi.org/10.1016/j.ifset.2013.04.015.
Prediction of beef quality attributes using VIS/NIR Zapotoczny, P., Kozera, W., Karpiesiuk, K., & Pawlowski, R.
hyperspectral scattering imaging technique. J. Food Eng., (2014). The use of computer-assisted image analysis in the
109(2), 267-273. evaluation of the effect of management system on changes in
http://dx.doi.org/10.1016/j.jfoodeng.2011.10.004. color, chemical composition, and texture of m. longissimus dorsi
Xia, J. J., Lee, J. W., & Yao, G. (2007). Characterizing beef in pigs. Meat Sci., 97(4), 518-528.
muscles with optical scattering and absorption coefficient in http://dx.doi.org/10.1016/j.meatsci.2014.03.010.
VIS-NIR region. Meat Sci., 75(1), 78-83. Zhang, B., Huang, W., Gong, L., Li, J., Zhao, C., Liu, C., & Huang,
http://dx.doi.org/10.1016/j.meatsci.2006.07.002. D. (2015). Computer vision detection of defective apples using
Xiong, Z., Sun, D.-W., Zeng, S.-A., & Xie, A. (2014). Recent automatic lightness correction and weighted RVM classifier. J.
developments of hyperspectral imaging systems and their Food Eng., 146, 143-151.
applications in detecting quality attributes of red meats: A http://dx.doi.org/10.1016/j.jfoodeng.2014.08.024.
review. J. Food Eng., 132, 1-13. Zhang, L., Li, Y., Peng, Y., Wang, W., Tao, F., & San, J. (2012).
http://dx.doi.org/10.1016/j.jfoodeng.2014.02.004. Determination of pork freshness attributes by hyperspectral
Xiong, Z., Sun, D.-W., Dai, Q., Han, Z., Zeng, X.-A., & Wang, L. imaging technique. Trans. CSAE, 28(7), 254-259.
(2015). Application of visible hyperspectral imaging for Zhang, L., Xu, H., & Gu, M. (2014). Use of signal to noise ratio and
prediction of springiness of fresh chicken meat. Food Anal. area change rate of spectral to evaluate the visible/NIR spectral
Methods, 8(2), 380-391. http://dx.doi.org/10.1007/s12161-014- system for fruit internal quality detection. J. Food Eng., 139, 19-
9853-3. 23. http://dx.doi.org/10.1016/j.jfoodeng.2014.04.009.
Xu, Y., Cui, L., Tian, C., Li, S., Cao, J., Liu, Z., & Zhang, G. Zhao, S. W., Peng, Y. K., Wang, W., Zhang, H. Y., Song, Y. L., &
(2012). A multiplex polymerase chain reaction coupled with Zhao, J. (2012). On-line system for prediction of lean meat
high-performance liquid chromatography assay for simultaneous percentage and backfat thickness of pig carcass. J. Food Saf.
detection of six foodborne pathogens. Food Control, 25(2), 778- Qual., 3(6), 589-594.
783. http://dx.doi.org/10.1016/j.foodcont.2011.12.014. Zheng, C., Sun, D.-W., & Zheng, L. (2006). Correlating colour to
Xue, L., J. Cai, J., Li, J., & Liu, M. (2012). Application of particle moisture content of large cooked beef joints by computer
swarm optimization (PSO) algorithm to determine dichlorvos visions. J. Food Eng., 77(4), 858-863.
residue on the surface of navel orange with vis-NIR http://dx.doi.org/10.1016/j.jfoodeng.2005.08.013.
spectroscopy. Procedia Eng., 29, 4124-4128. Zheng, C., Sun, D.-W., & Zheng, L. (2007). Predicting shrinkage of
http://dx.doi.org/10.1016/j.proeng.2012.01.631. ellipsoid beef joints as effected by water immersion cooking
Yang, W. L., Peng, Y. K., Jiang, F. C., Tang, X. Y., Guo, H., & using image analysis and neural network. J. Food Eng., 79(4),
Zhou, T. (2012). Software design of marbling grades system for 1243-1249. http://dx.doi.org/10.1016/j.jfoodeng.2006.04.010.
beef. J. Food Saf. Qual., 3(6), 608-612. Zhou, L. J., Wu, H., Li, J. T., Wang, Z. Y., & Zhang, L. Y. (2012).
Yao, L., Zhou, W., Wang, T., Liu, M., & Yu, C. (2014). Determination of fatty acid in broiler breast meat by near-
Quantification of egg yolk contamination in egg white using infrared reflectance spectroscopy. Meat Sci., 90(3), 658-664.
UV/vis spectroscopy: Prediction model development and http://dx.doi.org/10.1016/j.meatsci.2011.10.010.
analysis. Food Control, 43, 88-97. Zor, K., Castellarnau, M., Pascual, D., Pich, S., Plasencia, C.,
http://dx.doi.org/10.1016/j.foodcont.2014.02.037. Bardsley, R., & Nistor, M. (2011). Design, development, and
Yoon, S. C., Lawrence, K. C., Smith, D. P., Park, B., & Windham, application of a bioelectrochemical detection system for meat
W. R. (2008). Embedded bone fragment detection in chicken tenderness prediction. Biosensors Bioelec., 26(11), 4283-4288.
fillets using transmittance image enhancement and hyperspectral http://dx.doi.org/10.1016/j.bios.2011.04.011.
V i e w p u b l i c a t i o n s t a t s

Meatqualityinspection-Review

Uploaded by

Document Informationclick to expand document information

Copyright:

Available Formats

Meatqualityinspection-Review

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Meatqualityinspection-Review

Uploaded by

Copyright:

Available Formats

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

Optical Methods and Techniques for Meat Quality Inspection

The user has requested enhancement of the downloaded file.

Submitted for review in October 2014 as manuscript number PRS

Transactions of the ASABE

1372 TRANSACTIONS OF THE ASABE

58(5): 1371-1386 1373

1374 TRANSACTIONS OF THE ASABE

58(5): 1371-1386 1375

1376 TRANSACTIONS OF THE ASABE

58(5): 1371-1386 1377

1378 TRANSACTIONS OF THE ASABE

58(5): 1371-1386 1379

1380 TRANSACTIONS OF THE ASABE

58(5): 1371-1386 1381

1382 TRANSACTIONS OF THE ASABE

58(5): 1371-1386 1383

1384 TRANSACTIONS OF THE ASABE

58(5): 1371-1386 1385

1386 TRANSACTIONS OF THE ASABE

You might also like