Image Sampling and

Quantization

IMAGE SAMPLING AND QUANTIZATION

 Quantization

 Audio signals are continuous in time and amplitude

 Audio signal must be digitized in both time and amplitude to be
represented in binary form.
 Discrete in time by sampling – Nyquist
 Discrete in amplitude by quantization
 Once samples have been captured, they must be made discrete in
amplitude.
 Analog to Digital Converter
 Quantization
 Converts actual sample values (usually voltage measurements)
into an integer approximation
 Process of rounding off a continuous value so that it can be
represented by a fixed number of binary digits
 Tradeoff between number of bits required and error
 Human perception limitations affect allowable error
 Specific application affects allowable error
 Two approaches to quantization
 Rounding the sample to the closest integer.
 (e.g. round 3.14 to 3)
 Create a Quantizer table that generates a staircase pattern of
values based on a step size.
 There are numerous ways to acquire images.
 To generate digital images from sensed data.
 The output of most sensors is a continuous voltage
waveform whose amplitude and spatial behavior are
related to the physical phenomenon being sensed.
 To create a digital image, we need to convert the
continuous sensed data into digital form.
This involves two processes:
 sampling
 quantization.
 Basic Concepts in Sampling and
Quantization

 The basic idea behind sampling and quantization is

illustrated in Fig. 2.16.
 Figure 2.16(a) shows a continuous image f that we want to
convert to digital form.
 An image may be continuous with respect to the x- and y-
coordinates, and also in amplitude.
 To convert it to digital form, we have to sample the
function in both coordinates and in amplitude.
 Digitizing the coordinate values is called sampling
 Digitizing the amplitude values is called quantization.
 The one-dimensional function in Fig. 2.16(b) is a plot of
amplitude (intensity level) values of the continuous
image along the line segment AB in Fig. 2.16(a).
 The random variations are due to image noise.
 To sample this function, we take equally spaced samples
along line AB, as shown in Fig. 2.16(c).
 The spatial location of each sample is indicated by a
vertical tick mark in the bottom part of the figure.
 The samples are shown as small white squares
superimposed on the function
 The set of these discrete locations gives the sampled function.
However, the values of the samples still span (vertically) a
continuous range of intensity values.
 In order to form a digital function, the intensity values also must
be converted(quantized) into discrete quantities.
 The right side of Fig. 2.16(c) shows the intensity scale divided
into eight discrete intervals, ranging from black to white.
 The vertical tick marks indicate the specific value assigned to
each of the eight intensity intervals.
 The continuous intensity levels are quantized by assigning
one of the eight values to each sample.
• The assignment is made depending on the vertical proximity of a
sample to a vertical tick mark.
• The digital samples resulting from both sampling and
quantization are shown in Fig. 2.16(d).
• Starting at the top of the image and carrying out this procedure
line by line produces a two-dimensional digital image.
• It is implied in Fig. 2.16 that, in addition to the number of
discrete levels used, the accuracy achieved in quantization is
highly dependent on the noise content of the sampled signal.
 The method of sampling is determined by the sensor
arrangement used to generate the image.
 When an image is generated by a single sensing element
combined with mechanical motion, as in Fig. 2.13.
 However, spatial sampling is accomplished by selecting
the number of individual mechanical increments at
which we activate the sensor to collect data.
• Mechanical motion can be made very exact so, in
principle, there is almost no limit as to how fine we can
sample an image using this approach.
• In practice, limits on sampling accuracy are determined
by other factors, such as the quality of the optical
components of the system.
 When a sensing strip is used for image acquisition,
the number of sensors in the strip establishes the
sampling limitations in one image direction.
 Mechanical motion in the other direction can be
controlled more accurately,
 but it makes little sense to try to achieve sampling
density in one direction that exceeds the sampling
limits established by the number of sensors in the
other.
 Quantization of the sensor outputs completes the
process of generating a digital image
 When a sensing array is used for image acquisition, there is no
motion and the number of sensors in the array establishes the
limits of sampling in both directions.
 Quantization of the sensor outputs is as before.
 Figure 2.17 illustrates this concept.
 Figure 2.17(a) shows a continuous image projected onto the
plane of an array sensor.
• Figure 2.17(b) shows the image after sampling and quantization.
• Clearly, the quality of a digital image is determined to a large
degree by the number of samples and discrete intensity levels
used in sampling and quantization.
• Image content is also an important consideration in choosing
these parameters
 Representing Digital Images

 Let f(s,t) represent a continuous image function of two continuous

variables, s and t.
 We convert this function into a digital image by sampling and
quantization,
 Suppose that we sample the continuous image into a 2-D array, , containing
M rows and N columns, where are discrete coordinates. For notational
clarity and convenience, we use integer values for these discrete
coordinates
 For example, the value of the digital image at the origin is f(0,)) , and the
next coordinate value along the first row is . Here, the notation (0, 1) is
used to signify
 the second sample along the first row. It does not mean that these are the
values
 of the physical coordinates when the image was sampled. In general, the
 value of the image at any coordinates(x,y) is denoted by f(x,y) , where x and
y
 are integers. The section of the real plane spanned by the coordinates of an
 image is called the spatial domain, with x and y being referred to as spatial
 variables or spatial coordinates.
 The representation of an M×N numerical array as

 f (0, 0) f (0,1) ... f (0, N  1) 

 f (1, 0) f (1,1) ... f (1, N  1) 
f ( x, y )  
 ... ... ... ... 
 
 f ( M  1, 0) f ( M  1,1) ... f ( M  1, N  1) 
 The representation of an M×N numerical array as

 a0,0 a0,1 ... a0, N 1 

 a a1,1 ... a1, N 1 
A  1,0 
 ... ... ... ... 
 
 aM 1,0 aM 1,1 ... aM 1, N 1 
 The representation of an M×N numerical array in
MATLAB

 f (1,1) f (1, 2) ... f (1, N ) 

 f (2,1) f (2, 2) ... f (2, N ) 
f ( x, y )  
 ... ... ... ... 
 
 f ( M ,1) f ( M , 2) ... f (M , N )
 Discrete intensity interval [0, L-1], L=2k

 The number b of bits required to store a M × N digitized

image

b=M×N×k
 As Fig. 2.18 shows, there are three basic ways to represent f(x,y).
 Figure 2.18(a) is a plot of the function, with two axes determining
spatial location
 and the third axis being the values of f (intensities) as a function of
the two spatial
 variables x and y. Although we can infer the structure of the image
in this
 example by looking at the plot, complex images generally are too
detailed and
 difficult to interpret from such plots.This representation is useful
when working
 with gray-scale sets whose elements are expressed as triplets of the
form (x,y,z)
 , where x and y are spatial coordinates and z is the value of f at
coordinates (x,y)
 The representation in Fig. 2.18(b) is much more common. It shows
 as it would appear on a monitor or photograph. Here, the intensity of each
 point is proportional to the value of f at that point. In this figure, there are only
 three equally spaced intensity values. If the intensity is normalized to the interval
 [0, 1], then each point in the image has the value 0, 0.5, or 1. A monitor
 or printer simply converts these three values to black, gray, or white, respectively,
 as Fig. 2.18(b) shows. The third representation is simply to display the
 numerical values of f(x,y) as an array (matrix). In this example, f is of size 600 x 600
 elements, or 360,000 numbers. Clearly, printing the complete array
 would be cumbersome and convey little information.When developing algorithms,
 however, this representation is quite useful when only parts of the
 image are printed and analyzed as numerical values. Figure 2.18(c) conveys
 this concept graphically
 Numerical arrays are used for processing and
algorithm development.
 In equation form, we write the representation of an
M x N numerical array as
 origin of a digital image is at the
 top left, with the positive x-axis extending downward and the positive y-
axis
 extending to the right.This is a conventional representation based on the
fact
 that many image displays (e.g., TV monitors) sweep an image starting at
the
 top left and moving to the right one row at a time. More important is the
fact
 that the first element of a matrix is by convention at the top left of the
array, so
 choosing the origin of f(x,y) at that point makes sense mathematically.
Keep
 in mind that this representation is the standard right-handed Cartesian
coordinate
 system with which you are familiar.† We simply show the axes pointing
 downward and to the right, instead of to the right and up
 we define the dynamic range of an imaging
 system to be the ratio of the maximum measurable
intensity to the minimum
 detectable intensity level in the system. As a rule, the upper limit is
determined
 by saturation and the lower limit by noise (see Fig. 2.19). Basically,
dynamic
 range establishes the lowest and highest intensity levels that a system can
represent
 and, consequently, that an image can have. Closely associated with this
concept
 is image contrast, which we define as the difference in intensity between
 the highest and lowest intensity levels in an image.When an appreciable
number
 of pixels in an image have a high dynamic range, we can expect the image
 to have high contrast. Conversely, an image with low dynamic range
typically
 has a dull, washed-out gray look
 The number of intensity levels corresponding to each
 value of k is shown in parentheses.When an image
can have 2 power k intensity levels,
 it is common practice to refer to the image as a “k-bit
image.”
 Spatial and Intensity Resolution

 spatial resolution can be stated in a number of ways,

 with line pairs per unit distance, and dots (pixels)
per unit distance being
 among the most common measures. Suppose that we
construct a chart with
 alternating black and white vertical lines, each of
width W units (W can be
 less than 1).The width of a line pair is thus 2W, and
there are ½ W line pairs
 per unit distance
 A widely used definition of image resolution is
 the largest number of discernible line pairs per unit
distance (e.g., 100 line
 pairs per mm). Dots per unit distance is a measure of
image resolution used
 commonly in the printing and publishing industry.
In the U.S., this measure
 usually is expressed as dots per inch (dpi).
 the number of bits used to quantize intensity as the
intensity resolution
 Image Interpolation

 Interpolation is a basic tool used extensively in tasks

such as zooming, shrinking,
 rotating, and geometric corrections.
 (shrinking and
 zooming), which are basically image resampling
methods.
 interpolation is the process of using known data to
estimate
 values at unknown locations.
 nearest neighbor interpolation because it
 assigns to each new location the intensity of its
nearest neighbor in the original
 image