Colour Image Processing

UNIT - 5

Trusha Patel, CE Dept, CSPIT, CHARUSAT, Changa 1

Use of color

Powerful descriptor that

often simplifies object identification

Trusha Patel, CE Dept, CSPIT, CHARUSAT, Changa 2

Use of color

Human can discern

thousand of color shades and intensities
compared to
about an two dozen shades of gray

Trusha Patel, CE Dept, CSPIT, CHARUSAT, Changa 3

 Color image processing

Full color Pseudocolor

processing processing

Input image acquired with full-color sensors Assigning color to

such as color TV camera & color scanner monochrome intensity / Range of intensities

Trusha Patel, CE Dept, CSPIT, CHARUSAT, Changa 4

 COLOR FUNDAMENTALS

Trusha Patel, CE Dept, CSPIT, CHARUSAT, Changa 5

Quantities to describe quality of chromatic light source
• Radiance
 Total amount of energy that flows from light source
 Measured in watts (W)

Trusha Patel, CE Dept, CSPIT, CHARUSAT, Changa 6

Quantities to describe quality of chromatic light source
• Radiance
• Luminance
 Measure amount of energy an observer perceives
 Measured in lumens (lm)

Trusha Patel, CE Dept, CSPIT, CHARUSAT, Changa 7

Quantities to describe quality of chromatic light source
• Radiance
• Luminance
• Brightness
 Subjective descriptor
 Practically not possible to measure

Trusha Patel, CE Dept, CSPIT, CHARUSAT, Changa 8

• Primary color of light
(secondary color of pigment)
• Red (R)
• Green (G)
• Blue (B)

• Primary color of pigment

(secondary color of light)
• Magenta (R + B)
• Cyan (G + B)
• Yellow (R + G)

Trusha Patel, CE Dept, CSPIT, CHARUSAT, Changa 9

Characteristics used to distinguish one color from another

• Brightness
• embodies the chromatic notion of intensity

Trusha Patel, CE Dept, CSPIT, CHARUSAT, Changa 10

Characteristics used to distinguish one color from another

• Brightness
• Hue
• Represents dominant color as perceived by an observer
• E.g. when said object is red, yellow, orange means specifying its hue

Trusha Patel, CE Dept, CSPIT, CHARUSAT, Changa 11

Characteristics used to distinguish one color from another

• Brightness
• Hue
• Saturation
• Refers to the relative purity or amount of white light mixed with a hue

Trusha Patel, CE Dept, CSPIT, CHARUSAT, Changa 12

Characteristics used to distinguish one color from another

• Hue & Saturation together called Chromaticity

• Amount of red, green, blue needed to form any particular color are
called Tristimulus values, denoted by X, Y, Z

Trusha Patel, CE Dept, CSPIT, CHARUSAT, Changa 13

• Colors can also be specified
by CIE chromaticity
diagram

x (red)
y (green)
z (blue)

Trusha Patel, CE Dept, CSPIT, CHARUSAT, Changa 14

 COLOR MODELS

Trusha Patel, CE Dept, CSPIT, CHARUSAT, Changa 15

(1) RGB color model
• (R)Red (G)Green (B)Blue

• Each color appears in its primary spectral components of red, green,

blue

• Model based on Cartesian coordinate system

Trusha Patel, CE Dept, CSPIT, CHARUSAT, Changa 16

(1) RGB color model

• Cube composed of
(28)3 = 1,67,77,216 colors

Trusha Patel, CE Dept, CSPIT, CHARUSAT, Changa 17

(1) RGB color model
• Color image can be acquired by using three filters sensitive to red,
green and blue respectively

• When color scene with monochrome camera with one of these filters,
the result is monochrome image whose intensity is proportional to the
response of that filter

• Three monochrome images produces by three filters, that are RGB

component images of color scene

Trusha Patel, CE Dept, CSPIT, CHARUSAT, Changa 18

(1) RGB color model

Trusha Patel, CE Dept, CSPIT, CHARUSAT, Changa 19

(1) RGB color model
• Many system are limited to 256 colors (24-bit RGB image)

• This subset of colors is called set of safe RGB colors / set of all-
system-safe colors

Trusha Patel, CE Dept, CSPIT, CHARUSAT, Changa 20

(1) RGB color model

Trusha Patel, CE Dept, CSPIT, CHARUSAT, Changa 21

(2) CMY and CMYK color models
• (C)Cyan (M)Magenta (Y)Yellow (B)Blue

• Most devices use CMY data input or convert RGB to CMY internally

• Pure Cyan does not contain Red

Pure Magenta does not contain Green
Pure Yellow does not contain Blue

Trusha Patel, CE Dept, CSPIT, CHARUSAT, Changa 22

(2) CMY and CMYK color models
• RGB color can be obtain by subtracting CMY value from 1

• As given in figure, equal amount of

cyan, magenta and yellow should
generate black

• In practice it generates muddy-looking black,

to produce true black CMYK is introduced

Trusha Patel, CE Dept, CSPIT, CHARUSAT, Changa 23

(3) HSI color model
• (H)Hue (S)Saturation (I)Intensity

• RGB is ideal for image color generation

while
HSI for describing color in terms that are practical for human
interpretation

• Decouple the intensity component from color-carrying information

Trusha Patel, CE Dept, CSPIT, CHARUSAT, Changa 24

(3) HSI color model

Trusha Patel, CE Dept, CSPIT, CHARUSAT, Changa 25

(3) HSI color model

Trusha Patel, CE Dept, CSPIT, CHARUSAT, Changa 26

(3) HSI color model

RGB to HSI conversion

𝐻= { 𝜃 𝑖𝑓 𝐵≤ 𝐺
3 60 − 𝜃 𝐼𝑓 𝐵>𝐺

{ }
1
3 2
[ ( 𝑅 − 𝐺 ) +( 𝑅 − 𝐵 )]
𝑆 =1 − ¿ 𝜃= 𝑐𝑜𝑠
−1

( 𝑅 +𝐺 + 𝐵) 2
[ ( 𝑅 − 𝐺 ) +( 𝑅 − 𝐵 )(𝐺 − 𝐵 )]
1
2

1
𝐼= ( 𝑅+ 𝐺 + 𝐵 )
3

Trusha Patel, CE Dept, CSPIT, CHARUSAT, Changa 27

(3) HSI color model

HSI to RGB conversion

(1) RG Sector 0 ° ≤ 𝐻 ≤ 120°

𝐵=𝐼 ( 1− 𝑆)

[
𝑅= 𝐼 1+
𝑆 cos 𝐻
°
cos ⁡(60 − 𝐻 ) ]
𝐺=3 𝐼 −( 𝑅 + 𝐵)

Trusha Patel, CE Dept, CSPIT, CHARUSAT, Changa 28

(3) HSI color model

HSI to RGB conversion

(2) GB Sector 120° ≤ 𝐻 ≤ 240°

𝐻 =𝐻 −120
°
𝑅=𝐼 (1− 𝑆)

[
𝐺= 𝐼 1+
𝑆 cos 𝐻
°
cos ⁡(60 − 𝐻 ) ]
𝐵=3 𝐼 −( 𝑅+ 𝐺)

Trusha Patel, CE Dept, CSPIT, CHARUSAT, Changa 29

(3) HSI color model

HSI to RGB conversion

(3) BR Sector 240° ≤ 𝐻 ≤ 360°

𝐻 =𝐻 −240
°
𝐺=𝐼 (1 −𝑆)

[
𝐵= 𝐼 1+
𝑆 cos 𝐻
°
cos ⁡( 60 − 𝐻 ) ]
𝑅=3 𝐼 −( 𝐺+ 𝐵)

Trusha Patel, CE Dept, CSPIT, CHARUSAT, Changa 30