CDS 6334 - Visual Image Processing

Lecture 1: Introduction to Images

1. What is Computer Vision?

Computer Vision is a branch of Artificial Intelligence that enables computers to detect, process, analyze and understand visual information from images and videos.
🧠 Remember:

Human → Understands visual scenes naturally
Computer Vision → Tries to teach machines to do the same

2. Goal of Computer Vision

Convert image/video data into meaningful information for decision making.
Exam Keyword:
Image/Video → Information

3. Human Vision vs Computer Vision

Human Vision Computer Vision
Naturally interprets scenes Requires algorithms and data
Handles ambiguity well Can be fooled easily
Fast understanding Needs computation
🧠 Human vision remains more robust than machines in many situations.

4. Fields Related to Images and Videos

Field Main Purpose
Computer Graphics Create images
Image Processing Manipulate images
Computer Vision Understand images
🧠 Shortcut:

Graphics → Create
Processing → Improve
Vision → Understand

5. Three Main Areas of Computer Vision

  1. Measurement
  2. Perception and Interpretation
  3. Search and Organization
Measurement: Recover information about the real 3D world from images.
Perception & Interpretation: Recognize objects, people, activities and scenes.
Search & Organization: Find and organize visual data efficiently.

6. Measurement

Reconstructing 3D models from multiple images.
Computer vision estimates properties of the real world from visual data.

7. Perception and Interpretation

Detecting faces, recognizing objects and understanding scenes.
🧠 Think:
"What is in the image?"

8. Search and Organization

Google Image Search and image retrieval systems.
Visual data can be indexed, searched and categorized automatically.

9. Applications of Computer Vision

Exam Tip:
Be able to explain at least 3 real-world applications.

10. Biometrics and Recognition

Technology Purpose
Face Recognition Identity verification
Fingerprint Recognition User authentication
Iris Recognition High-security identification

11. Optical Character Recognition (OCR)

OCR converts images containing text into machine-readable text.
License plate recognition, digit recognition, scanned document conversion.

12. Why Computer Vision is Difficult

Computer Vision is an Ill-Posed Problem.
The real world is 3D, but images are only 2D projections.
🧠 Remember:

Many different real-world situations can produce the same image.

13. Digital Images

A digital image is composed of pixels arranged in rows and columns.
🧠 Pixel = Smallest unit of a digital image.

14. Image Processing Tasks

These are core low-level image processing operations.

15. Colour Image Processing

RGB is the most common colour representation.

16. Higher-Level Vision Tasks

  1. Representation
  2. Description
  3. Recognition
  4. Interpretation
  5. Semantic Understanding
Modern deep learning systems perform object detection and scene understanding.

17. Emerging Applications

18. Final Exam Summary

Most Important Points

  • Computer Vision: Image/Video → Information
  • Image Processing: Image → Image
  • Graphics: Create images
  • Three Areas: Measurement, Perception, Search
  • Challenge: 3D world projected to 2D images
  • OCR: Image text → Digital text
  • Applications: Face recognition, medical imaging, autonomous vehicles
  • Digital Image: Collection of pixels