Abstract: In computer vision applications feature detection algorithm plays a very important role. Widely used features in many stereo vision work, are robust against the change of perspective like edge elements, corners, line segments, and curve segments. The major types of stereo image matching are Intensity-based stereo matching and feature-based stereo matching. The intensity based stereo matching requires depth calculation and this might get complex and costly. While the feature based stereo image matching is easier than depth calculation. This review paper describes the method to match stereo image with the help of Harris corner detection algorithm. Algorithm contains Gaussian smoothing filter for noise reduction so that false corners can be avoided. The architecture contains flexible threshold operator for cor-ner detection and it requires less time. The system will be implemented on System Generator to reduce the bulkiness of the system and to maximize the speed of operation.

Keywords: Harris corner detection, stereo matching, Gaussian smoothing, System Generator.