PART I

IMAGE PROCESSING
INSTRUCTOR: DR. NGUYEN NGOC TRUONG MINH
SCHOOL OF ELECTRICAL ENGINEERING, INTERNATIONAL UNIVERSITY (VNU-HCMC)

Ho Chi Minh City, Spring 2024

Lecture 11 – Color Image Processing 2

LECTURE XI –
COLOR IMAGE PROCESSING
INSTRUCTOR: DR. NGUYEN NGOC TRUONG MINH
SCHOOL OF ELECTRICAL ENGINEERING, INTERNATIONAL UNIVERSITY (VNU-HCMC)

Ho Chi Minh City, Spring 2024

Lecture 11 – Color Image Processing 3

LECTURE CONTENT
• What are the most important concepts and terms related to color
perception?
• What are the main color models used to represent and quantify color?
• How are color images represented in MATLAB?
• What is pseudo-color image processing and how does it differ from full-color
image processing?
• How can monochrome image processing techniques be extended to color
images?
• Chapter Summary – What we have learned?
• Problems
Lecture 11 – Color Image Processing 4

11.1 THE PSYCHOPHYSICS OF COLOR

• Color perception is a psychophysical phenomenon that combines two main

components:
1. The physical properties of light sources (usually expressed by their
spectral power distribution (SPD)) and surfaces (e.g., their absorption
and reflectance capabilities).
2. The physiological and psychological aspects of the human visual system
(HVS).
Lecture 11 – Color Image Processing 5

11.1.1 BASIC CONCEPTS

• The perception of color starts with a chromatic light source, capable of

emitting electromagnetic radiation with wavelengths between
approximately 400 and 700 nm. Part of that radiation reflects on the
surfaces of the objects in a scene and the resulting reflected light reaches
the human eye, giving rise to the sensation of color.
• An object that reflects light almost equally in all wavelengths within the
visible spectrum is perceived as white, whereas an object that absorbs
most of the incoming light, regardless of the wavelength, is seen as black.
Lecture 11 – Color Image Processing 6

11.1.1 BASIC CONCEPTS

• The perception of several shades of gray between pure white and pure
black is usually referred to as achromatic. Objects that have more selective
properties are considered chromatic, and the range of the spectrum that
they reflect is often associated with a color name.
Lecture 11 – Color Image Processing 7

11.1.1 BASIC CONCEPTS

• For example, an object that absorbs most of the energy within the 565–
590 nm wavelength range is considered yellow.
• A chromatic light source can be described by three basic quantities:
➢ Intensity (or Radiance): the total amount of energy that flows from the light
source, measured in watts (W).
➢ Luminance: a measure of the amount of information an observer perceives
from a light source, measured in lumen (lm). It corresponds to the radiant
power of a light source weighted by a spectral sensitivity function
(characteristic of the HVS).
➢ Brightness: the subjective perception of (achromatic) luminous intensity.
Lecture 11 – Color Image Processing 8

11.1.1 BASIC CONCEPTS

• The human retina (the surface at the back of the eye where images are
projected) is coated with photosensitive receptors of two different types:
cones and rods.
• Rods cannot encode color but respond to lower luminance levels and
enable vision under darker conditions.
• Cones are primarily responsible for color perception and operate only
under bright conditions. There are three types of cone cells (L cones, M
cones, and S cones, corresponding to long (≈ 610 nm), medium (≈ 560
nm), and short (≈ 430 nm) wavelengths, respectively).
Lecture 11 – Color Image Processing 9

11.1.1 BASIC CONCEPTS

Lecture 11 – Color Image Processing 10

11.1.1 BASIC CONCEPTS

• The existence of three specialized types of cones in the human eye was
hypothesized more than a century before it could be confirmed
experimentally by Thomas Young and his trichromatic theory of vision in
1802.

• The secondary colors of light, obtained by additive mixtures of the

primaries, two colors at a time, are magenta (or purple) = red + blue, cyan
(or turquoise) = blue + green, and yellow = green + red (Figure 11.2a).
Lecture 11 – Color Image Processing 11

11.1.1 BASIC CONCEPTS

• For color mixtures using pigments (or paints), the primary colors are
magenta, cyan, and yellow and the secondary colors are red, green, and blue
(Figure 11.2b). It is important to note that for pigments a color is named
after the portion of the spectrum that it absorbs, whereas for light a color
is defined based on the portion of the spectrum that it emits.
• Consequently, mixing all three primary colors of light results in white (i.e., the
entire spectrum of visible light), whereas mixing all three primary colors of
paints results in black (i.e., all colors have been absorbed, and nothing
remains to reflect the incoming light).
Lecture 11 – Color Image Processing 12

11.1.1 BASIC CONCEPTS

Lecture 11 – Color Image Processing 13

11.1.1 BASIC CONCEPTS

• The use of the expression primary colors to refer to red, green, and blue may
lead to a common misinterpretation: that all visible colors can be obtained
by mixing different amounts of each primary color, which is not true.
• A related phenomenon of color perception, the existence of color
metamers, may have contributed to this confusion. Color metamers are
combinations of primary colors (e.g., red and green) perceived by the HVS
as another color (in this case, yellow) that could have been produced by a
spectral color of fixed wavelength (of ≈ 580 nm).
Lecture 11 – Color Image Processing 14

11.1.2 THE CIE XYZ CHROMATICITY DIAGRAM

• In color matching experiments performed in the late 1920s, subjects were
asked to adjust the amount of red, green, and blue on one patch that were
needed to match a color on a second patch. The results of such
experiments are summarized in Figure 11.3.
• The existence of negative values of red and green in this figure means that the
second patch should be made brighter (i.e., equal amounts of red, green,
and blue has to be added to the color) for the subjects to report a perfect
match. Since adding amounts of primary colors on the second patch
corresponds to subtracting them in the first, negative values can occur.
Lecture 11 – Color Image Processing 15

11.1.2 THE CIE XYZ CHROMATICITY DIAGRAM

• The amounts of three

primary colors in a
three-component
additive color model
(on the first patch)
needed to match a test
color (on the second
patch) are called
tristimulus values.
Lecture 11 – Color Image Processing 16

11.1.2 THE CIE XYZ CHROMATICITY DIAGRAM

Lecture 11 – Color Image Processing 17

11.1.2 THE CIE XYZ CHROMATICITY DIAGRAM

• This model, whose color matching functions are shown in Figure 11.4, is
known as the CIE XYZ (or CIE 1931) model. The tristimulus values of X,Y,
and Z in Figure 11.4 are related to the values of R, G, and B in Figure 11.3
by the following linear transformations:

and
Lecture 11 – Color Image Processing 18

11.1.2 THE CIE XYZ CHROMATICITY DIAGRAM

• The CIE XYZ color space was designed so that the 𝑌 parameter
corresponds to a measure of the brightness of a color. The chromaticity of
a color is specified by two other parameters, 𝑥 and 𝑦, known as
chromaticity coordinates and calculated as
𝑋
𝑥=
𝑋+𝑌+𝑍
𝑌
𝑦=
𝑋+𝑌+𝑍
where 𝑥 and 𝑦 are also called normalized tristimulus values. The third
normalized tristimulus value, 𝑧, can just as easily be calculated as
𝑍
𝑧=
𝑋+𝑌+𝑍
Lecture 11 – Color Image Processing 19

11.1.2 THE CIE XYZ CHROMATICITY DIAGRAM

• The resulting CIE XYZ chromaticity diagram shows a horseshoe-shaped

outer curved boundary, representing the spectral locus of wavelengths (in
nm) along the visible light portion of the electromagnetic spectrum.
• The line of purples on a chromaticity diagram joins the two extreme
points of the spectrum, suggesting that the sensation of purple cannot be
produced by a single wavelength: it requires a mixture of shortwave and
longwave light and for this reason purple is referred to as a nonspectral color. All
colors of light are contained in the area in (𝑥, 𝑦) bounded by the line of
purples and the spectral locus, with pure white at its center.
Lecture 11 – Color Image Processing 20

11.1.2 THE CIE XYZ CHROMATICITY DIAGRAM

• The inner triangle in Figure 11.6a represents a color gamut, that is, a range
of colors that can be produced by a physical device, in this case a CRT
monitor. Different color image display and printing devices and
technologies exhibit gamuts of different shape and size, as shown in Figure
11.6.

• As a rule of thumb, the larger the gamut, the better the device’s color
reproduction capabilities.
Lecture 11 – Color Image Processing 21

11.1.2 THE CIE XYZ CHROMATICITY DIAGRAM

Lecture 11 – Color Image Processing 22

11.1.2 THE CIE XYZ CHROMATICITY DIAGRAM

Lecture 11 – Color Image Processing 23

11.1.3 PERCEPTUALLY UNIFORM COLOR SPACES

• One of the main limitations of the CIE XYZ chromaticity diagram lies in
the fact that a distance on the 𝑥𝑦 plane does not correspond to the degree of
difference between two colors. This was demonstrated in the early 1940s by
David MacAdam, who conducted experiments asking subjects to report
noticeable changes in color (relative to a starting color stimulus).
• The result of that study is illustrated in Figure 11.7, showing the resulting
MacAdam ellipses: regions on the chromaticity diagram corresponding to all
colors that are indistinguishable, to the average human eye, from the color at
the center of the ellipse.
Lecture 11 – Color Image Processing 24

11.1.3 PERCEPTUALLY UNIFORM COLOR SPACES

Lecture 11 – Color Image Processing 25

11.1.3 PERCEPTUALLY UNIFORM COLOR SPACES

In M AT L A B
• The IPT has an extensive support for conversion among CIE color spaces.
Converting from one color space to another is usually accomplished by
using function makecform (to create a color transformation structure that
defines the desired color space conversion) followed by applycform, which
takes the color transformation structure as a parameter.
Lecture 11 – Color Image Processing 26

11.1.3 PERCEPTUALLY UNIFORM COLOR SPACES

TA B L E 11.1 IP T F unctions for CIE X Y Z a nd CIE L A B Color S pa ces

Function Description
xyz2double Converts an M ×3 or M × N × 3 array of XYZ color values to double
xyz2uint16 Converts an M ×3 or M × N × 3 array of XYZ color values to uint16
lab2double Converts an M ×3 or M × N × 3 array of L*a*b* color values to double
lab2uint16 Converts an M ×3 or M × N × 3 array of L*a*b* color values to uint16
lab2uint8 Converts an M ×3 or M × N × 3 array of L*a*b* color values to uint8
whitepoint Returns a 3×1 vector of XYZ values scaled so that Y = 1
Lecture 11 – Color Image Processing 27

11.1.4 ICC PROFILES

• An ICC (International Color Consortium) profile is a standardized

description of a color input or output device, or a color space, according
to standards established by the ICC. Profiles are used to define a mapping
between the device source or target color space and a profile connection
space (PCS), which is either CIELAB (L*a*b*) or CIEXYZ.
In M AT L A B
• The IPT has several functions to support ICC profile operations. They are
listed in Table 11.2.
Lecture 11 – Color Image Processing 28

11.1.4 ICC PROFILES

TA B L E 11.2 IP T F unctions for ICC P rofile M a nipula tion

Function Description
iccread Reads an ICC profile into the MATLAB workspace
iccfind Finds ICC color profiles on a system, or a particular ICC color profile
whose description contains a certain text string
iccroot Returns the name of the directory that is the default system repository
for ICC profiles
iccwrite Writes an ICC color profile to disk file
Lecture 11 – Color Image Processing 29

11.2 COLOR MODELS

• A color model (also called color space or color system) is a specification

of a coordinate system and a subspace within that system where each
color is represented by a single point.
• There have been many different color models proposed over the last 400
years. Contemporary color models have also evolved to specify colors for
different purposes (e.g., photography, physical measurements of light, color
mixtures, etc.). In this section, we discuss the most popular color models
used in image processing.
Lecture 11 – Color Image Processing 30

11.2.1 THE RGB COLOR MODEL

• The RGB color model is based on a Cartesian coordinate system whose
axes represent the three primary colors of light (R, G, and B), usually
normalized to the range [0, 1] (Figure 16.8). The eight vertices of the
resulting cube correspond to the three primary colors of light, the three
secondary colors, pure white, and pure black. Table 16.3 shows the R, G, and B
values for each of these eight vertices.
• RGB color coordinates are often represented in hexadecimal notation, with
individual components varying from 00 (decimal 0) to FF (decimal 255).
For example, a pure (100% saturated) red would be denoted FF0000,
whereas a slightly desaturated yellow could be written as CCCC33.
Lecture 11 – Color Image Processing 31

11.2.1 THE RGB COLOR MODEL

• The number of discrete values of R, G, and B

is a function of the pixel depth, defined as the
number of bits used to represent each pixel: a
typical value is 24 bits = 3 image planes x 8
bits per plane.
• The resulting cube—with more than 16
million possible color combinations.
Lecture 11 – Color Image Processing 32

11.2.1 THE RGB COLOR MODEL

TA B L E 11.3 R, G, a nd B Va lues for E ig ht Representa tive

Colors Corresponding to the Vertices of the RGB Cube
Color Name R G B
Black 0 0 0
Blue 0 0 1
Green 0 1 0
Cyan 0 1 1
Red 1 0 0
Magenta 1 0 1
Yellow 1 1 0
White 1 1 1
Lecture 11 – Color Image Processing 33

11.2.1 THE RGB COLOR MODEL

Lecture 11 – Color Image Processing 34

11.2.2 THE CMY AND CMYK COLOR MODELS

• The CMY model is based on the three primary colors of pigments (cyan,
magenta, and yellow). It is used for color printers, where each primary color
usually corresponds to an ink (or toner) cartridge. Since the addition of equal
amounts of each primary to produce black usually produces unacceptable,
muddy looking black, in practice, a fourth color, blacK, is added, and the
resulting model is called CMYK.
• The conversion from RGB to CMY is straightforward:

• The inverse operation (conversion from CMY to RGB), although it is equally

easy from a mathematical standpoint, is of little practical use.
Lecture 11 – Color Image Processing 35

11.2.3 THE HSV COLOR MODEL

• Color models such as the RGB and CMYK described previously are very
convenient to specify color coordinates for display or printing,
respectively. They are not, however, useful to capture a typical human
description of color. After all, none of us goes to a store looking for a
FFFFCC shirt to go with the FFCC33 jacket we got for our birthday.
• Rather, the human perception of color is best described in terms of hue,
saturation, and lightness. Hue describes the color type, or tone, of the color
(and very often is expressed by the “color name”), saturation provides a
measure of its purity (or how much it has been diluted in white), and
lightness refers to the intensity of light reflected from objects.
Lecture 11 – Color Image Processing 36

11.2.3 THE HSV COLOR MODEL

• For representing colors in a way that is closer to the human description, a

family of color models have been proposed. The common aspect among
these models is their ability to dissociate the dimension of intensity (also
called brightness or value) from the chromaticity—expressed as a
combination of hue and saturation—of a color.
Lecture 11 – Color Image Processing 37

11.2.3 THE HSV COLOR MODEL

• The HSV (sometimes called HSB) color model can be obtained by looking
at the RGB color cube along its main diagonal (or gray axis), which results in
a hexagon-shaped color palette. As we move along the main axis in the
pyramid in Figure 11.10, the hexagon gets smaller, corresponding to
decreasing values of V , from 1 (white) to 0 (black).
• For any hexagon, the three primary and the three secondary colors of
light are represented in its vertices. Hue, therefore, is specified as an angle
relative to the origin (the red axis by convention). Finally, saturation is
specified by the distance to the axis: the longer the distance, the more
saturated the color.
Lecture 11 – Color Image Processing 38

11.2.3 THE HSV COLOR MODEL

Lecture 11 – Color Image Processing 39

11.2.3 THE HSV COLOR MODEL

• Figure 11.11 shows an alternative representation of the HSV color model
in which the hexcone is replaced by a cylinder. Figure 16.12 shows yet
another equivalent three-dimensional representation for the HSV color
model, as a cone with circular-shaped base.
• In summary, the main advantages of the HSV color model (and its closely
related alternatives) are its ability to match the human way of describing
colors and to allow for independent control over hue, saturation, and
intensity (value). The ability to isolate the intensity component from the
other two—which are often collectively called chromaticity
components—is a requirement in many color image processing
algorithms.
Lecture 11 – Color Image Processing 40

11.2.3 THE HSV COLOR MODEL

Lecture 11 – Color Image Processing 41

11.2.3 THE HSV COLOR MODEL

Lecture 11 – Color Image Processing 42

11.2.4 THE YIQ (NTSC) COLOR MODEL

• The NTSC color model is used in the American standard for analog
television, which will be described in more detail in Chapter 20. One of the
main advantages of this model is the ability to separate grayscale contents
from color data, a major design requirement at a time when emerging
color TV sets and transmission equipment had to be backward compatible
with their B&W predecessors. In the NTSC color model, the three
components are luminance (Y ) and two color-difference signals hue (I)
and saturation (Q).
• Conversion from RGB to YIQ can be performed using the transformation
Lecture 11 – Color Image Processing 43

11.2.5 THE YCBCR COLOR MODEL

• The YCbCr color model is the most popular color representation for digital
video. In this format, one component represents luminance (Y ), while the
other two are color-difference signals: Cb (the difference between the
blue component and a reference value) and Cr (the difference between
the red component and a reference value).
• Conversion from RGB to YCbCr is possible using the transformation
Lecture 11 – Color Image Processing 44

11.3 REPRESENTATION OF COLOR IMAGES IN

MATLAB
• Color images are usually
represented as RGB (24 bits per
pixel) or indexed with a palette
(color map), usually of size 256.
These representation modes are
independent of file format
(although GIF images are usually
indexed, and JPEG images typically
are not). In this section, we provide
a more detailed analysis of color
image representation in MATLAB.
Lecture 11 – Color Image Processing 45

11.3.1 RGB IMAGES

• An RGB color image in MATLAB corresponds to a 3D array of dimensions

MxNx3, where M and N are the image’s height and width (respectively) and 3 is
the number of color planes (channels). Each color pixel is represented as a triple
containing the values of its R, G, and B components (Figure 11.13). Each
individual array of size MxN is called a component image and corresponds to one of
the color channels: red, green, or blue. The data class of the component images
determines their range of values.
• For RGB images of class double, the range of values is [0.0, 1.0], whereas for
classes uint8 or uint16, the ranges are [0, 255] and [0, 65535], respectively. RGB
images typically have a bit depth of 24 bits per pixel (8 bits per pixel per
component image), resulting in a total number of (28)3 = 16, 777, 216 colors.
Lecture 11 – Color Image Processing 46

EXAMPLE 11.1

• The following MATLAB sequence can be used to open, verify the size (in
this case, 384 x 512 x 3) and data class (in this case, uint8), and display an
RGB color image (Figure 11.14).
I = imread(’peppers.png’);
size(I)
class(I)
subplot(2,2,1), imshow(I), title(’Color image (RGB)’);
subplot(2,2,2), imshow(I(:,:,1)), title(’Red component’);
subplot(2,2,3), imshow(I(:,:,2)), title(’Green component’);
subplot(2,2,4), imshow(I(:,:,3)), title(’Blue component’);
Lecture 11 – Color Image Processing 47

EXAMPLE 11.1
Lecture 11 – Color Image Processing 48

11.3.2 INDEXED IMAGES

• An indexed image is a matrix of integers (X), where each integer refers to a
particular row of RGB values in a secondary matrix (map) known as a color
map. The image can be represented by an array of class uint8,
uint16,ordouble. The color map array is an M x 3 matrix of class double,
where each element’s value is within the range [0.0, 1.0]. Each row in the
color map represents R (red),G (green), and B (blue) values, in that order.
• The indexing mechanism works as follows (Figure 11.15): if X is of class
uint8 or uint16, all components with value 0 point to the first row in map,
all components with value 1 point to the second row, and so on. If X is of
class double, all components with value less than or equal to 1.0 point to
the first row and so on.
Lecture 11 – Color Image Processing 49

11.3.2 INDEXED IMAGES

• MATLAB has many built-in color

maps (which can be accessed
using the function colormap),
briefly described in Table 11.4. In
addition, you can easily create
your own color map by defining
an array of class double and size
M x 3, where each element is a
floating-point value in the range
[0.0, 1.0].
Lecture 11 – Color Image Processing 50

11.3.2 INDEXED IMAGES

TA B L E 11.4 Color M aps in M AT L A B
Name Description Name Description
hsv Hue–saturation–value color map colorcube Enhanced color-cube color map
hot Black–red–yellow–white color map vga Windows color map for 16 colors
gray Linear gray-scale color map jet Variant of HSV
bone Gray scale with tinge of blue color map prism Prism color map
copper Linear copper-tone color map cool Shades of cyan and magenta color map
pink Pastel shades of pink color map autumn Shades of red and yellow color map
white All white color map spring Shades of magenta and yellow color map
flag Alternating red, white, blue, and black
winter Shades of blue and green color map
color map
lines Color map with the line colors summer Shades of green and yellow color map
Lecture 11 – Color Image Processing 51

EXAMPLE 11.2

• The following MATLAB sequence can be used to load a built-in indexed

image, verify its size (in this case, 200 x 300) and data class (in this case,
double), verify the color map’s size (in this case, 81 x 3) and data class (in
this case, double), and display the image (Figure 11.16).
load clown
size(X)
class(X)
size(map)
class(map)
imshow(X,map), title(’Color (Indexed)’)
Lecture 11 – Color Image Processing 52

EXAMPLE 11.2
Lecture 11 – Color Image Processing 53

11.3.2 INDEXED IMAGES

• MATLAB has many useful functions for manipulating indexed color images:
➢ If we need to approximate an indexed image by one with fewer colors, we can use MATLAB’s
imapprox function.
➢ Conversion between RGB and indexed color images is straightforward, thanks to functions
rgb2ind and ind2rgb.
➢ Conversion from either color format to their grayscale equivalent is equally easy, using
functions rgb2gray and ind2gray.
➢ We can also create an index image from an RGB image by dithering the original image using
function dither.
➢ Function grayslice creates an indexed image from an intensity (grayscale) image by
thresholding and can be used in pseudocolor image processing.
➢ Function gray2ind converts an intensity (grayscale) image into its indexed image equivalent. It
is different from grayslice. In this case, the resulting image is monochrome, just as the original
one; only the internal data representation has changed.
Lecture 11 – Color Image Processing 54

11.4 PSEUDOCOLOR IMAGE PROCESSING

• The purpose of pseudocolor image processing techniques is to enhance a
monochrome image for human viewing purposes. Their rationale is that
subtle variations of gray levels may very often mask or hide regions of
interest within an image. This can be particularly damaging if the masked
region is relevant to the application domain (e.g., the presence of a tumor
in a medical image).
• Since the human eye is capable of discerning thousands of color hues and
intensities, compared to only less than 100 shades of gray, replacing gray
levels with colors leads to better visualization and enhanced capability for
detecting relevant details within the image.
Lecture 11 – Color Image Processing 55

11.4 PSEUDOCOLOR IMAGE PROCESSING

• The typical solution consists of using a color lookup table (LUT)

designed to map the entire range of (typically 256) gray levels to a
(usually much smaller) number of colors. For better results, contrasting
colors should appear in consecutive rows in the LUT.
• The term pseudocolor is used to emphasize the fact that the assigned
colors usually have no correspondence whatsoever with the true colors
that might have been present in the original image.
Lecture 11 – Color Image Processing 56

11.4.1 INTENSITY SLICING

• The technique of intensity (or density) slicing is the simplest and best-
known pseudocoloring technique.
• If we look at a monochrome image as if it were a 3D plot of gray levels
versus spatial coordinates, where the most prominent peaks correspond
to the brightest pixels, the technique corresponds to placing several planes
parallel to the coordinate plane of the image (also known as the xy plane).
Lecture 11 – Color Image Processing 57

11.4.1 INTENSITY SLICING

• Each plane “slices” the 3D function in the area of intersection, resulting in

several gray-level intervals. Each side of the plane is then assigned a
different color.
• Figure 11.17 shows an example of intensity slicing using only one slicing
plane at 𝑓(𝑥, 𝑦) = 𝑙𝑖 and Figure 11.18 shows an alternative
representation, where the chosen colors are indicated as 𝑐1 , 𝑐2 , 𝑐3 , and
𝑐4 . The idea can easily be extended to M planes and M + 1 intervals.
Lecture 11 – Color Image Processing 58

11.4.1 INTENSITY SLICING

Lecture 11 – Color Image Processing 59

11.4.1 INTENSITY SLICING

Lecture 11 – Color Image Processing 60

EXAMPLE 11.3

In M AT L A B
• Intensity slicing can be accomplished using the grayslice function.
• Figure 11.19 shows an example of pseudocoloring with intensity slicing
using 16 levels, the same input image (a), and three different color maps
summer (b), hsv (c), and jet (d).
Lecture 11 – Color Image Processing 61

EXAMPLE 11.3
Lecture 11 – Color Image Processing 62

11.6 TUTORIAL: PSEUDOCOLOR IMAGE PROCESSING

G o al

• The goal of this tutorial is to learn how to display grayscale images using
pseudocolors in MATLAB.

O bjectives

• Learn how to use the grayslice function to perform intensity slicing.

• Learn how to specify color maps with a custom number of colors.

Lecture 11 – Color Image Processing 63

11.6 TUTORIAL: PSEUDOCOLOR IMAGE PROCESSING

P rocedure
We will start by exploring the grayslice function on a gradient image.
1. Create and display a gradient image.
I = repmat(uint8([0:255]),256,1);
figure, subplot(1,2,1), subimage(I), title(’Original Image’);
2. Slice the image and display the results.
I2 = grayslice(I,16);
subplot(1,2,2), subimage(I2,colormap(winter(16))), ...
title(’Pseudo-colored with "winter" colormap’)
Q uestio n 1. Why did w e use the subimage functio n to display the
images (instead o f the familiar imshow )?
Lecture 11 – Color Image Processing 64

11.6 TUTORIAL: PSEUDOCOLOR IMAGE PROCESSING

Q uestio n 2. What do es the value 16 represent in the functio n call fo r

gray slice?
Q uestio n 3. In the statement subimage(I2,co lo rmap(w inter(16))), w hat
do es the value 16 represent?
In the above procedure, we sliced the image into equal partitions—this is
the default for the grayslice function. We will now learn how to slice the
range of grayscale values into unequal partitions.
3. Slice the image into unequal partitions and display the result.
levels = [0.25*255, 0.75*255, 0.9*255];
I3 = grayslice(I,levels);
figure, imshow(I3,spring(4))
Lecture 11 – Color Image Processing 65

11.6 TUTORIAL: PSEUDOCOLOR IMAGE PROCESSING

Q uestio n 4. T he o riginal image co nsists o f values in the range [0, 255].

If o ur o riginal image values ranged [0.0, 1.0], how w o uld the above
co de change?
Now that we have seen how pseudocoloring works, let us apply it to an
image where this visual information might be useful.
4. Clear all variables and close any open figures.
5. Load and display the mri.jpg.
I = imread(’mri.jpg’);
figure, subplot(1,2,1), subimage(I), title(’Original Image’);
Lecture 11 – Color Image Processing 66

11.6 TUTORIAL: PSEUDOCOLOR IMAGE PROCESSING

6. Pseudocolor the image.

I2 = grayslice(I,16);
subplot(1,2,2), subimage(I2, colormap(jet(16))), ...
title(’Pseudo-colored with "jet" colormap’);
Q uestio n 5. In the previo us steps, w e have specified how many co lo rs w e
want in o ur co lo r map. If w e do no t specify this number, how do es
M AT L A B determine how many co lo rs to return in the co lo r map?
Lecture 11 – Color Image Processing 67

11.7 TUTORIAL: FULL-COLOR IMAGE PROCESSING

G o al
• The goal of this tutorial is to learn how to convert between color spaces
and perform filtering on color images in MATLAB.
O bjectives
• Learn how to convert from RGB to HSV color space using the rgb2hsv function.
• Learn how to convert from HSV to RGB color space using the hsv2rgb function.
• Explore smoothing and sharpening in the RGB and HSV color spaces.
Lecture 11 – Color Image Processing 68

11.7 TUTORIAL: FULL-COLOR IMAGE PROCESSING

P rocedure
We will start by exploring the rgb2hsv function.
1. Load the onions.png image and display its RGB components.
I = imread(’onion.png’);
figure, subplot(2,4,1), imshow(I), title(’Original Image’);
subplot(2,4,2), imshow(I(:,:,1)), title(’R component’);
subplot(2,4,3), imshow(I(:,:,2)), title(’G component’);
subplot(2,4,4), imshow(I(:,:,3)), title(’B component’);
2. Convert the image to HSV and display its components.
Ihsv = rgb2hsv(I);
subplot(2,4,6), imshow(Ihsv(:,:,1)), title(’Hue’);
subplot(2,4,7), imshow(Ihsv(:,:,2)), title(’Saturation’);
subplot(2,4,8), imshow(Ihsv(:,:,3)), title(’Value’);
Lecture 11 – Color Image Processing 69

11.7 TUTORIAL: FULL-COLOR IMAGE PROCESSING

Q uestio n 1. Why do w e no t display the HSV equivalent o f the image?

When viewing the components of an RGB image, the grayscale visualization
of each component is intuitive because the intensity within that component
corresponds to how much of the component is being used to generate the
final color.Visualization of the components of an HSV image is not as
intuitive.You may have noticed that when displaying the hue, saturation, and
value components, hue and saturation do not give you much insight as to
what the actual color is. The value component, on the other hand, appears
to be a grayscale version of the image.
Lecture 11 – Color Image Processing 70

11.7 TUTORIAL: FULL-COLOR IMAGE PROCESSING

3. Convert the original image to grayscale and compare it with the value
component of the HSV image.
Igray = rgb2gray(I);
figure, subplot(1,2,1), imshow(Igray), title(’Grayscale’);
subplot(1,2,2), imshow(Ihsv(:,:,3)), title(’Value component’);
Q uestio n 2. How do es the gray scale versio n o f the o riginal image and
the value co mpo nent o f the HSV image co mpare?
Procedures for filtering a color image will vary depending on the color
space being used. Let us first learn how to apply a smoothing filter on an
RGB image.
Lecture 11 – Color Image Processing 71

11.7 TUTORIAL: FULL-COLOR IMAGE PROCESSING

4. Apply a smoothing filter to each component and then reconstruct the
image.
fn = fspecial(’average’);
I2r = imfilter(I(:,:,1), fn);
I2g = imfilter(I(:,:,2), fn);
I2b = imfilter(I(:,:,3), fn);
I2(:,:,1) = I2r;
I2(:,:,2) = I2g;
I2(:,:,3) = I2b;
figure, subplot(1,2,1), imshow(I), title(’Original Image’);
subplot(1,2,2), imshow(I2), title(’Averaged Image’);
Lecture 11 – Color Image Processing 72

11.7 TUTORIAL: FULL-COLOR IMAGE PROCESSING

Q uestio n 3. Do the results co nfirm that the R G B equivalent o f averaging a
gray scale image is to average each co mpo nent o f the R G B image
individually ?
Now let us see what happens when we perform the same operations on an HSV
image. Note that in these steps, we will use the hsv2rgb function to convert the
HSV image back to RGB so that it can be displayed.
5. Filter all components of the HSV image.
Ihsv2h = imfilter(Ihsv(:,:,1), fn);
Ihsv2s = imfilter(Ihsv(:,:,2), fn);
Ihsv2v = imfilter(Ihsv(:,:,3), fn);
Ihsv2(:,:,1) = Ihsv2h;
Ihsv2(:,:,2) = Ihsv2s;
Ihsv2(:,:,3) = Ihsv2v;
Lecture 11 – Color Image Processing 73

11.7 TUTORIAL: FULL-COLOR IMAGE PROCESSING

6. Display the results.

figure, subplot(2,3,1), imshow(Ihsv(:,:,1)), ...
title(’Original Hue’);
subplot(2,3,2), imshow(Ihsv(:,:,2)), ...
title(’Original Saturation’);
subplot(2,3,3), imshow(Ihsv(:,:,3)), ...
title(’Original Value’);
subplot(2,3,4), imshow(Ihsv2(:,:,1)), ...
title(’Filtered Hue’);
Lecture 11 – Color Image Processing 74

11.7 TUTORIAL: FULL-COLOR IMAGE PROCESSING

subplot(2,3,5), imshow(Ihsv2(:,:,2)), ...

title(’Filtered Saturation’);
subplot(2,3,6), imshow(Ihsv2(:,:,3)), ...
title(’Filtered Value’);
figure, subplot(1,2,1), imshow(I), title(’Original Image’);
subplot(1,2,2), imshow(hsv2rgb(Ihsv2)), ...
title(’HSV with all components filtered’);
Lecture 11 – Color Image Processing 75

11.7 TUTORIAL: FULL-COLOR IMAGE PROCESSING

Q uestio n 4. B ased o n the results, do es it make sense to say that the

HSV equivalent o f averaging a gray scale image is to average each
co mpo nent o f the HSV image individually ?
7. Filter only the value component and display the results.
Ihsv3(:,:,[1 2]) = Ihsv(:,:,[1 2]);
Ihsv3(:,:,3) = Ihsv2v;
figure, subplot(1,2,1), imshow(I), title(’Original Image’);
subplot(1,2,2), imshow(hsv2rgb(Ihsv3)), ...
title(’HSV with only value component filtered’);
Lecture 11 – Color Image Processing 76

11.7 TUTORIAL: FULL-COLOR IMAGE PROCESSING

Q uestio n 5. How do es this result co mpare w ith the previo us o ne?
We can sharpen an HSV image following a similar sequence of steps.
8. Sharpen the HSV image and display the result.
fn2 = fspecial(’laplacian’,0);
Ihsv4v = imfilter(Ihsv(:,:,3), fn2);
Ihsv4(:,:,[1 2]) = Ihsv(:,:,[1 2]);
Ihsv4(:,:,3) = imsubtract(Ihsv(:,:,3),Ihsv4v);
figure, subplot(1,2,1), imshow(I), title(’Original Image’);
subplot(1,2,2), imshow(hsv2rgb(Ihsv4)), ...
title(’HSV sharpened’);
Q uestio n 6. How w o uld w e perfo rm the same sharpening technique o n
an R G B image?
Lecture 11 – Color Image Processing 77

WHAT HAVE WE LEARNED?

• This chapter introduced the most important concepts and terms related
to color perception, representation, and processing. Understand the meaning
of the main terms in colorimetry is a required step toward the
understanding of color image processing.
• There are many color models used to represent and quantify color
information. Some of the most popular (and their use) models are as
follows:
─ RGB, CMY(K): display and printing devices
─ YIQ,YCbCr: television and video
─ XYZ: color standardization
─ CIELAB, CIELUV: perceptual uniformity
─ HSV, HSI, HSL: intuitive description of color properties
Lecture 11 – Color Image Processing 78

WHAT HAVE WE LEARNED?

• Color images can be represented in MATLAB either as an MxNx3 array
(one per color channel) or as an MxN array of indices (pointers) to a
secondary (usually 256x3) color palette. The former representation is
called an RGB image, whereas the latter is called an indexed image.

• Pseudocolor image processing techniques assign color to pixels based on an

interpretation of the data rather than the original scene color (which may
not even be known). Full-color image processing methods, on the other
hand, process the pixel values of images whose colors usually correspond
to the color of the original scene.
Lecture 11 – Color Image Processing 79

WHAT HAVE WE LEARNED?

• Several monochrome image processing techniques, from edge detection to

histogram equalization, can be extended to color images.
• The success of applying such techniques to color images depends on the
choice of color model used to represent the images.
Lecture 11 – Color Image Processing 80

PROBLEMS

P ro blem 1. Use the MATLAB function patch to display the RGB cube in
Figure 11.9.
P ro blem 2. Write MATLAB code to add two RGB color images. Test it with
two test images (of the same size) of your choice. Are the results what you
had expected?
P ro blem 3. What is wrong with the following MATLAB code to add an
indexed color image to a constant for brightening purposes? Fix the code to
achieve the desired goal.
[X, map] = imread(’canoe.tif ’);
X = X + 20;
Lecture 11 – Color Image Processing 81

PROBLEMS

P ro blem 4. In our discussion of pseudocoloring, we stated that the intensity

slicing method described in Section 11.4.1 is a particular case of the more
general method (using transformation functions) described in Section 11.4.2.
Assuming that a 256-level monochrome image has been “sliced” into four
colors, red, green, blue, and yellow, and that each range of gray levels has the
same width (64 gray levels), plot the (staircase-shaped) transformation
functions for each color channel (R, G, and B).
Lecture 11 – Color Image Processing 82

PROBLEMS

P ro blem 5. Use the edge function in MATLAB and write a script to compute
and display the edges of a color image for the following cases:
(a) RGB image, combining the edges from each color channel by adding them up.
(b) RGB image, combining the edges from each color channel with a logical OR
operation.
(c) YIQ image, combining the edges from each color channel by adding them up.
(d) YIQ image, combining the edges from each color channel with a logical OR
operation.
Lecture 11 – Color Image Processing 83

END OF LECTURE XI

LECTURE 12 – IMAGE COMPRESSION AND

CODING