Computer Vision

Introduction

Dr. Surya Prakash

 Human Vision
– How does human visualize/understand the  real 
world

 Computer Vision
– Understanding  the real world with the help of 
computer

2
 Introduction
 Human Vision
–It can do amazing things like:
• Recognize people and objects
• Navigate through obstacles
• Understand mood in the scene
• Imagine stories

3
 Introduction
 Computer Vision:
–Main task involves intelligent 
interpretation of pictures or scenes
• Understanding  the input image/ scene
• Extracting the information from the 
scene 
• Description of the scene information

4
 Components of a
Computer Vision System

Camera

Lighting

Computer

Scene

Scene Interpretation
5
 Computer Vision

Computer Images/Scenes
Images/Scenes
Vision representation

Visualization

6
 Introduction
 Human Vision Vs. Computer Vision

What a human sees

What a computer sees 7
 Computer Vision Tasks

 Computer Vision deals with
– Intelligent interpretation of pictures or scenes

 Majority of the tasks in the fields of Computer 
Vision deals with
– Understanding  the input image/ scene
– Extracting the information from the scene 
– Description of the scene information

8
 Computer Vision Tasks/Applications
 Image restoration and filtering
 Image representation 
 Image description
 Feature extraction
 Image segmentation
 Image matching

9
 Computer Vision Tasks/Applications
 Image synthesis
 Image reconstruction
 Range data processing
 Stereo image processing
 Image morphology

10
Computer Vision – Selection of Techniques
 Techniques used to solve a particular problem:
– depends on the application domain
– nature of data being considered as input

 Choice of techniques/features
– are problem specific
– Are chosen after some amount of trial & error
– experience helps in choosing techniques/features

11
 Computer Vision
 Example 1:
– Analysis of remote sensed images (satellite 
imagery) may involve techniques of texture 
analysis
– However, these techniques may not be helpful in 
analyzing  images of indoor/outdoor scenes
 Example 2:
– Character Recognition Vs. Speech Recognition

12
Computer Vision Vs. Image Processing

 In Computer Vision
– Input: Images, scenes 
– Output: 
• Intelligent interpretation of pictures or scenes
• Output may be represented in the form of scalar, vector 
or matrix

13
 Image Processing Tasks
 Digital Image processing (DIP)
– Input: Image (an array of pixels)
– Output: Image (another transformed array of 
pixels )
– This transformation in some way represents an 
improvement to the original array

Image Processing 
Transformation

Input output 14
 Image Processing Tasks
 DIP also deals with improving the 
pictorial information
– improves the contrast of the image
– removes noise
– removes blurring caused by movement of 
the camera during image acquisition
– it may correct geometrical distortions
caused by the lens

15
 Courses Related to Computer Vision
 It uses the concepts and techniques from 
– Digital Image Processing
– Pattern Recognition
– Artificial Intelligence
– Artificial Neural Networks
– Computer Graphics
– Optimization Techniques
– Fuzzy Theory
– VLSI and Computer Architecture

16
 Topics Covered in this Course
 Image Processing
– Histogram Processing
– Fourier Transform
– Convolution and Correlation
– Image Enhancement
– Edge Detection
– Morphological Operations
 Hough Transform 
 Statistical Techniques
– Least Square Fitting
– Principal Component Analysis
17
 Topics Covered in this Course
 Feature detection and representation
– SIFT and SURF
 Segmentation
– Texture Analysis and Segmentation
 Optical Flow
 Object Segmentation
– Active Contour
 Object Recognition
 Pattern Classification
18
Applications Areas of Computer Vision
 Character Recognition
 Document Processing
 Biometric Applications
– verification of persons using biometric characteristics
– Biometric characteristics:  face, iris & fingerprint
 Forensic Analysis
 Pose and Gesture Identification
 Automatic Inspection of Industrial Products
 Industrial Process Monitoring
19
Applications Areas of Computer Vision
 Biomedical Engineering (Diagnosis and 
surgery)

 Military surveillance and target identification

 Navigation and mobility
– for robots and unmanned vehicles ‐ land, air and 
underwater

 Remote sensing (using satellite imagery)
20
Applications Areas of Computer Vision
 Geographic Information Systems (GIS)

 Safety and security (night vision)

 Traffic monitoring

 Sports (training and incident analysis)

 Entertainment and virtual reality

21
 Pre‐requisite
• Mathematics:
• Basic Linear Algebra
• Probability
• Calculus
• Programming
• Basic Data Structures
• Programming Knowledge
• No Prior knowledge of Computer Vision
Required
22
 Evaluation/Grading
Component Credit Marks
Mid Sem 25
End Sem 35
Attendance 5
Quiz ‐
Assignment 5
Project Mid‐sem 10
Project End‐sem 20
Total 100 23
 Text Books
 Digital Image Processing, R. C. Gonzalez 
and R. E., Woods; Addison Wesley; 1992.
 Computer Vision: A Modern Approach, D. 
A. Forsyth and J. Ponce,
 Fundamentals of Digital Image 
Processing, Anil K Jain, Prentice Hall of 
India, 1995.

24
 References
 Journals:
– IEEE Transactions on Pattern Analysis and Machine 
Intelligence 
– IEEE Transactions on Image Processing 
– Pattern Recognition (PR), Elsevier
– Image and Vision Computing (IVC), Elsevier
– Pattern Recognition Letters (PRL), Elsevier
– Computer Vision, Graphics & Image Processing, 
Elsevier
– International Journal of Computer Vision (IJCV), 
Springer

25
 References
 Online Resources
– CV Online: 
http://homepages.inf.ed.ac.uk/rbf/CVonline/
– Computer Vision Homepage: 
http://www.cs.cmu.edu/~cil/vision.html

26