Stereo Vision and Depth Perception in Computer Vision - Lecture 16

Explore stereo vision techniques in this 33-minute lecture from the University of Central Florida's Computer Vision series. Delve into shape from stereo, stereo pairs, and depth maps. Learn about rectification, correspondence using search, and various correlation-based stereo methods including block matching, sum of squares differences, minimum absolute difference, and maximum matching pixel count. Examine cross correlation, normalized correlation, mutual correlation, and simulated annealing. Analyze results with window correlation and discover practical applications of stereo vision. Access the accompanying presentation for visual aids and consult the provided reading material for further study.

Intro
Shape from Stereo
Stereo Pairs and Depth Maps (from Szeliski's book)
Rectification
Correspondence using Search
Correlation Based Stereo Methods
Block Matching
Sum of Squares Differences (SSD)
Minimum Absolute Difference
Maximum Matching Pixel Count
Cross Correlation
Normalized Correlation
Mutual Correlation
Simulated Annealing
Results with window correlation
Applications of Stereo (from Szeliski's book)
Reading Material