Intro to Computer Vision

CMSC 25040/35040
Introduction to Computer Vision

General information

Instructor:	Pedro F. Felzenszwalb
Email:	pff (at) cs.uchicago.edu
Office:	Ryerson 162B

TA: Eric Purdy
Email: epurdy (at) cs.uchicago.edu

Lecture: MWF 11:30-12:20 Ryerson 251

Overview

This course is an introduction to some of the fundamental concepts in computer vision. We will cover selected topics in low-level image analysis, perceptual grouping, 3D reconstruction, motion estimation and object recognition. Our approach will be to focus on mathematical models and efficient algorithms.

There is no required textbook for the course. This webpage will have handouts covering some of the material from class.

Lectures

01. Intro
02. Filtering and Convolution
03. Edge detection
04. Hough transform
05. Corner detection (harris)
06. Simple stereo geometry
07. Window-based stereo matching
08. Stereo matching with DP
09. Stereo with graph cuts
10. Optical-flow
11. Estimating constant/parametric flow
12. Perceptual organization
13. Image segmentation
14. Graph-based image segmentation
15. Saliency network
16. Saliency network continued
17. Bayesian justification for the saliency network
18. Object detection with rigid templates
19. Part based models (pictorial structures)
20. Generalized distance transforms
21. More on part based models
22. Linear combination of views
23. Statistical interpretation of template matching
24. Statistical interpretation of part-based models