Figures from Gonzalez and Woods, Digital Image Processing, Second Edition, 2002.

1
Lecture 17

• Figures from Gonzalez and Woods, Digital Image
  Processing, Second Edition, 2002.

2
Psuedocolor Processing

• Produce RGB image (usually) from gray levels.
• Also possible to transform a number of monochrome
  images.

3
Chapter 6 Color Image Processing
4
Chapter 6 Color Image Processing
5
Typical Transformations

• Sinusiodal functions with differing start points

6
Chapter 6 Color Image Processing
7
Chapter 6 Color Image Processing
8
Chapter 1 Introduction
9
Chapter 1 Introduction
10
Chapter 6 Color Image Processing
11
Chapter 6 Color Image Processing
12
Color Transformation
13
Chapter 6 Color Image Processing
14
Color Transformations (2)

• Also possible in
• CMYK components
• HSI components

15
Chapter 6 Color Image Processing
16
Modify through intensity
17
Chapter 6 Color Image Processing
18
The Wheel of Hues

• Notice how complementary colors are opposite one
  another

19
Chapter 6 Color Image Processing
20
Effect of Complementary Transformation

• Notice difference in RGB vs. HSI model

21
Chapter 6 Color Image Processing
22
Color Slicing
23
Chapter 6 Color Image Processing
24
Chapter 6 Color Image Processing
25
Chapter 6 Color Image Processing
26
Histogram Processing

• Done to intensity component. Should not be done
  to RGB components.
• Next figure
• Histogram equalization to intensity
• Then increased saturation

27
Chapter 6 Color Image Processing
28
Smoothing and Sharpening

• Smoothing filters (like averaging or low pass)
• Or sharpening (like Laplacian or high pass)
• Are usually done on R,G,B components
• Or on intensitywith slightly different results

29
Chapter 6 Color Image Processing
30
Chapter 6 Color Image Processing
31
Chapter 6 Color Image Processing
32
Chapter 6 Color Image Processing
33
Color Segmentation

• As done in HSI space
• Color represented by hue image
• Saturation used to perform masking
• Intensity contains no color information and not
  used much for segmentation

34
Chapter 6 Color Image Processing
35
Color Segmentation

• Done with distances in RGB space
• Either Euclidean or maximum are easy to do
• In fact, looking at this example, my thought was,
  oh, thats what you want!

36
Chapter 6 Color Image Processing
37
Chapter 6 Color Image Processing
38
Color Noise

• Handled naturally in RGB mode.
• Standard color models make sense

39
Chapter 6 Color Image Processing
40
Chapter 6 Color Image Processing
41
Chapter 6 Color Image Processing
42
Chapter 6 Color Image Processing
43
Chapter 6 Color Image Processing