A COMPARATIVE ANALYSIS OF STEREO VISION ALGORITHMS

Abstract

There has been rapid development and research in the field of computer vision aiming towards the design of complex autonomous system that helps human in simplifying the job which consumes a lot of time and effort. The theme of this field has been to duplicate the abilities of human vision by electronically perceiving and understanding an image. This image understanding can be seen as the disentangling of symbolic information from image data using models constructed with the aid of geometry, physics, statistics, and learning theory. This thesis aims at study and analysis of different stereo matching algorithms by comparing their performances based on their accuracy and computational speed. This thesis report includes study of several stereo matching techniques such as Sum of Absolute Differences (SAD), Sum of Squared Differences (SSD), Normalized Cross Correlation (NCC) and Adaptive Window. SAD, SSD and NCC are simple local area-based algorithms which are faster in computation but have relatively lower accuracy. Adaptive window algorithm considers the variance of the error also, which further helps in improving accuracy without much affecting computation speed.