Professional Elective –Il

Computer Vision
Course Code: CET 4011B Credits: 3 TH+1 Lab=4
 Syllabus
Unit 1: Image Processing Fundamentals
Introduction : Types of Computer Images, Satellite Images, Medical Images, Image File Formats,
Components of Image Processing System, Fundamentals Steps in Image Processing, Dimensions of
image, Image Operations.
Image Formation and Low-Level Processing: Human Vision System, Computer Vision System,
Stereo Vision, Geometric Cameras and projection models, noise models, Human color perception.

Unit 2: Image Enhancement and Segmentation Techniques

Image Enhancement: Point processing, Gray Level Slicing, Thresholding
Transformations, Histogram Processing, Filtering with morphological operators,
Intensity transformations, contrast stretching, histogram equalization, Correlation
and convolution, smoothing filters,
sharpening filters.

2/1/2024 Computer Vision(PE-I) Unit 1 2024-245 TY S6 2

 Syllabus
Unit 2: Image Enhancement and Segmentation Techniques:
Image Segmentation: Classification of image segmentation techniques,
thresholding-based image segmentation, edge-based segmentation, edge
detection, edge linking, Hough transform, watershed transform, clustering
techniques, region approach.

Unit 3: Feature Detection and Extraction

Edge Detection: Mathematical concepts, Operators based on first order derivative
(Roberts, Prewitt and Sobel), Laplacian of Gaussian (LoG).
Corners Detection: Harris and Orientation Histogram, Scale-Invariant Feature
Transform (SIFT), Speeded Up Robust Features (SURF), Histogram of Oriented
gradients (HOG).

2/1/2024 Computer Vision(PE-I) Unit 1 2024-245 TY S6 3

 Syllabus
Unit 3: Feature Detection and Extraction
Feature extraction: Spatial Features, Amplitude, transform based features, Histogram
based statistical features, based on statistical moments (e.g., mean, variance, kurtosis,
etc), Shape/geometry-based features and moment-based features (Radii, perimeter,
area, compactness, max boundary rectangle, orientation etc.), Texture features (GLCM
and texture features, Gabor features), Color features.

2/1/2024 Computer Vision(PE-I) Unit 1 2024-245 TY S6 4

 Syllabus
Unit 4: Fundamentals of Video Processing
Introduction Analog Video, Digital Video, 3D Video, Video Quality, Video standards
Motion Estimation and Tracking, Motion Models.
Motion Estimation: Differential Methods: Lukas— Kanade Method, Horn—Schunk
Motion Estimation, Motion Tracking Kanade—Lucas—Tomasi Tracking, Mean-Shift
Tracking, Particle-Filter Tracking, Active-Contour Tracking.
Video Segmentation:
Change Detection: Shot-Boundary Detection, Background Subtraction, Motion
Segmentation: Dominant-Motion Segmentation, Multiple-Motion Segmentation,
Region-Based Motion Segmentation: Fusion of Color and Motion.

2/1/2024 Computer Vision(PE-I) Unit 1 2024-245 TY S6 5

 Syllabus
Unit 5: Applications of Computer Vision:
Object Recognition: simple object recognition methods, Shape correspondence and
shape matching, contour based representation, Region based representation, Patterns
and pattern classification.
Introduction to Satellite Image Processing: Concepts and Foundations of Remote
Sensing, Multispectral, Thermal, and Hyper Spectral Sensing, Earth Resource Satellites
Operating in the Optical Spectrum.
Introduction to Medical Image Processing: Basic steps of medical image Processing,
Medical Image Enhancement, Segmentation, Medical Image Analysis (Images of X-
ray/ CT Scan/ MRI).

2/1/2024 Computer Vision(PE-I) Unit 1 2024-245 TY S6 6

 Learning Resources:
Text Books:
1. R.C. Gonzalez and R.E. Woods, Digital Image Processing, Addison- Wesley, 1992
2. Richard Szeliski, Computer Vision: Algorithms and Applications, Springer-Verlag
London Limited 2011.
3. A. Forsyth, J. Ponce, Computer Vision: A Modern Approach, Pearson Education,
2003.
4. R. Davies, Computer & Machine Vision, Fourth Edition, Academic Press, 2012

2/1/2024 Computer Vision(PE-I) Unit 1 2024-245 TY S6 7

 Unit 2:
Image Enhancement and Segmentation Techniques:

2/1/2024 Computer Vision(PE-I) Unit 1 2024-245 TY S6 8

 Introduction: Image Enhancement

2/1/2024 Computer Vision(PE-I) Unit 1 2024-245 TY S6 9

Introduction: Image Enhancement

10
Point Processing

11
Point Processing

2/1/2024 Computer Vision(PE-I) Unit 1 2024-245 TY S6 12

Point Processing
Point processing is a technique in image processing that modifies the
intensity of pixels to enhance an image.
● Point processing operates on individual pixels, changing their value based on a
predefined function

● The new pixel value is calculated without considering the values of surrounding
pixels

● Point processing can be used to change the brightness, contrast, or emphasis

on certain gray levels

13
 Examples of Point Processing
● Contrast stretching: A basic gray level transformation that improves contrast
● Thresholding: A point processing technique
● Logarithmic transformations: A point processing technique that modifies pixel
intensity
● Power-law transformations: A point processing technique that modifies pixel
intensity
● Piecewise linear transformations: A point processing technique that modifies
pixel intensity
● Intensity level slicing: A basic gray level transformation
● Bit plane slicing: A basic gray level transformation

14
Point Processing

2/1/2024 Computer Vision(PE-I) Unit 1 2024-245 TY S6 15

Image Negative
• A negative of an image is an image where its lightest areas appear as
darkest and the darkest areas appear as lightest.
• The appearance change from lightest to darkest and darkest to lightest
is basically done in gray scale image and refers to the change of pixel
intensity values from highest to lowest and lowest to highest.

2/1/2024 Computer Vision(PE-I) Unit 1 2024-245 TY S6 16

Image Negative

17
18
Image Negatives

import cv2
import numpy as np

# Read original image

image = cv2.imread("C:\\Users\\Dell\\OneDrive\\Desktop\CV_Image\\Img3.jpg")

# Max intensity based on quantization

L = image.max()

# Subtract each intensity from max to obtain negative

negative = L - image

cv2.imshow('original', image)
cv2.imshow('negative', negative)
cv2.waitKey(0)
cv2.destroyAllWindows() 19
Contrast Stretching expands the range of intensity levels in an Image
Extreme Contrast Stretching yields Thresholding. Thresholded image
has maximum contrast with only 0 and 255 intensities
Brightness enhancement is shifting intensities to higher values

20
Basic Gray Level Transformations
Linear
Logarithmic
Power Law

21
Logarithmic Transform

22
Logarithmic transformation

Logarithmic transformation is divided into two

types:

1. Log transformation
2. Inverse log transformation

The formula for Logarithmic transformation

s = c log(r + 1)

Here, s and r are the pixel values for input and

output image. And c is constant. In the
formula, we can see that 1 is added to each
pixel value this is because if pixel intensity is
zero in the image then log(0) is infinity so, to
have minimum value one is added.
23
Logarithmic Transformation

24
Logarithmic Transformation

25
26
Power Law Transformation

27
28
29
30
31
32
33
Contrast Stretching
It expands the range of intensity values in an image.
Contrast is difference between intensity levels of darker and brighter
pixels.
It is done in 3 ways
1. Multiplying each input pixel intensity value by a constant scalar.
2. Using Histogram equivalent.
3. applying a transform which make dark pixel darker by assigning
slope of < 1 and bright portion brighter by assigning slope > 1.

34
35
Piecewise transformation
Piecewise transformation is a
spatial domain method used for
enhancing the group of pixels
falling in the defined range. The
pixels are grouped into different
groups based on specified range
and each group has its own linear
transformation and slope. The
commonly used piecewise
transformations are,
thresholding, contrast stretching,
gray-level slicing and bit-plane
slicing

36
37
38
39
Contrast Stretching

40
Contrast Stretching

41
Contrast Stretching

42
 Contrast Stretching

# Find line equations by calculating slopes

def Contrast_stretch(p, r1, s1, r2, s2):
pixelVal_vec = np.vectorize(Contrast_stretch)
if (0 <= p and p <= r1):
equation = (s1 / r1)*p # Contrast stretching
contrast = pixelVal_vec(image, r1, s1, r2, s2)
elif (r1 < p and p <= r2):
equation = ((s2 - s1)/(r2 - r1))*(p - r1)+s1 cv2.imwrite('contrast.jpg', contrast)
else: cv2.imshow('original', image)
cv2.imshow('contrast.jpg', contrast)
equation = ((255 - s2)/(255 - r2))*(p - r2)+s2 cv2.waitKey(0)
return equation cv2.destroyAllWindows()

# Read original image

image = cv2.imread("C:\\Users\\Dell\\OneDrive\\Desktop\ https://medium.com/@koushikc2000/basic-operations-on-ima
CV_Image\\img1.jpg") ges-using-opencv-python-cb0d60d11911

https://www.iosrjournals.org/iosr-jece/papers/vol1-issue2/L012
# Initialize range 2023.pdf
r1 = 55
https://ninjakx.github.io/Image_Enhancement/
s1 = 40
r2 = 140
43
s2 = 200
Intensity Level Slicing

44
Grey Level Slicing(Intensity Level)

This technique is used to highlight a specific range of gray levels in a given

image.
Similar to thresholding
Other levels can be suppressed or maintained
Useful for highlighting features in an image
It can be implemented in several ways, but the two basic themes are:
One approach is to display a high value for all gray levels in the range of
interest and a low value for all other gray levels.
The second approach, based on the transformation brightens the desired
range of gray levels but preservesHanan Hardan gray levels unchanged.

45
Grey Level Slicing(Intensity Level)

46
Grey Level Slicing(Intensity Level)

47
Bit Plane Slicing

48
Example

49
50
Thresholding

51
Single Threshold

52
Types of Thresholding

53
Two Objects

54
Procedure for obtaining Global Thresholding value T

55
Histogram Transformation

56
Histogram Equalization

57
Histogram Equalization

58
Histogram Equalization

59
Histogram Equalization

60
Histogram Equalization

61
62
63
Questions based on Slide No. 1 to Slide No. 38

1. How point processing works for image enhancement? Explain image negative
method in brief.
2. What do you mean by image enhancement? Define point processing.
3. Illustrate the image negative transformation with suitable example.
4. Explain the mechanism of spatial domain filtering with suitable functions.
5. List down basics of intensity transformation(Grey Level) in image enhancement.
6. Why Log transform is used in image enhancement?
7. With necessary graphs, explain Log transformation and power law
transformation for spatial domain image enhancement.
8. Illustrate contrast stretching and draw the necessary graph and calculate the slop
for given range.
9. Explain the concept of histogram for various images with relevant diagrams.
10.Explain the histogram equalization operation in image enhancement with
necessary expressions.
64
Filtering with Morphology Operators

65
Filtering with Morphology Operators

66
Filtering with Morphology Operators

67
Erosion and Dilation

68
Filtering with Morphology Operators:Dilation

69
70
Dilation

71
Filtering with Morphology Operators:Erosion

72
Filtering with Morphology Operators:Erosion

73
Erosion

74
75
76
Open and Close Operation

77
Open and Close

78
79
Spatial Filtering Concepts

80
Spatial Filtering Concepts

81
Smoothing Filters

82
83
84
85
Smoothing Filters

86
Smoothing Filters

87
Smoothing Filters

88
Spatial Filtering Concepts

89
90
Smoothing Filters

91
Smoothing Filters

92
Smoothing Filters

93
Smoothing Filters

94
Smoothing Filters

95
96
Smoothing Filters

97
Smoothing Filters

98
Smoothing Filters

99
Smoothing Filters

100
Smoothing Filters

101
102
Handling pixels close to boundaries

103
104
105
106
107
108
109
Median Filter

110
Sharpening Filter

111
Sharpening Filter

112
Sharpening Filter

113
Sharpening Filter

114
115
116
Sharpening Filter

117
Sharpening Filter

118
Sharpening Filter

119
Sharpening Filter

120
121
Laplacian Filter

122
Laplacian Filter

123
Sharpening Filter

124
Sharpening Filter

125
Questions

1. What do you understand by dilation and erosion in morphological

image processing? Explain with example. Also give one suitable
application for each.
2. What is a structuring element? How it is used for dilation operation?

126
Types of Image Segmentation
Image segmentation can be categorized into several types, including:
○ Thresholding Segmentation: This technique involves setting a threshold value to
separate pixels based on their intensity or color.
○ Edge-based Segmentation: It identifies object boundaries by detecting edges or
gradients within the image.
○ Region-based Segmentation: This approach groups pixels based on their visual
features, color, or texture similarity.
○ Semantic segmentation: Semantic segmentation is the process of assigning a class
label to each pixel in the image so that pixels with the same label belong to the same
object or category.
○ Clustering-based Segmentation: It utilizes clustering algorithms to group pixels with
similar attributes.
○ Instance segmentation is the process of assigning a class label and an instance label
to each pixel in the image, such that pixels with the same class label and instance label
belong to the same object or instance.
127
What Is Image Segmentation Used For?

mage segmentation finds applications in various domains, including:

I

○ Medical Imaging: Segmentation aids in identifying tumors, organs, or anatomical

structures, assisting in diagnosis, treatment planning, and surgical guidance.
○ Object Recognition and Tracking: Segmentation facilitates object detection, tracking,
and recognition tasks in computer vision applications.
○ Image Editing and Forensics: Segmenting images allows for selective editing,
background removal, and image manipulation.
○ Autonomous Driving: Accurate segmentation enables object detection, lane detection,
and scene understanding, contributing to safe and efficient autonomous vehicles.
○ Fingerprint recognition: Image segmentation can help extract fingerprints from images
and enable biometric authentication, criminal identification, forensic analysis, etc.
These are just some of the examples of image segmentation applications. Image
segmentation can be useful and beneficial in many more domains and scenarios.

128
Image Segmentation

129
Image Segmentation

130
Image Segmentation

131
Image Segmentation using Threshold Method

132
Image Segmentation using Threshold Method

133
134
135
136
137
138
Procedure for obtaining Global Thresholding value T

139
 Refer Link: https://www.youtube.com/watch?v=eyTFqyM03L
k&t=447s

140
141
142
143
144
145
146
147
148
149
150
151
Image Segmentation

152
Introduction to Image Segmentation

153
Introduction to Image Segmentation

154
Region Approach

155
Region Approach

156
Region Approach

157
158
Introduction to Image Segmentation

159
160
161
https://www.youtube.com/watch?v=0kUGpgIrZIw

162
163
164
https://www.youtube.com/watch?v=0kUGpgIrZIw

165
Edge Based Segmentation

166
Edge Based Segmentation

167
168
Edge Based Segmentation

169
Edge Based Segmentation

170
Edge Based Segmentation

171
Edge Based Segmentation

172
Edge Based Segmentation

173
Hough Transform

174
Hough Transform

175
Hough Transform

176
Hough Transform

177
Hough Transform

178
Hough Transform

179
Hough Transform

180
Hough Transform

181
Hough Transform

182
Links: Hough Transform
https://universe.bits-pilani.ac.in/uploads/JNKDUBAI/hough_transform.
pdf
https://www.youtube.com/watch?v=XRBc_xkZREg

183
Clustering Technique

184
Clustering Technique

185
186
187
188
189
190
191
Clustering Technique

192
Clustering Technique

193
Important Links
https://www.youtube.com/watch?v=j3_Ck5oP5oI

https://www.youtube.com/watch?v=fiDDn_F9U74

https://www.youtube.com/watch?v=FQy3bTe2pVc
https://www.youtube.com/watch?v=j3_Ck5oP5oI

https://www.youtube.com/watch?v=fiDDn_F9U74

https://www.youtube.com/watch?v=FQy3bTe2pVc

2/1/2024 Computer Vision(PE-I) Unit 1 2024-245 TY S6 194