A Research Of Corner Detection Algorithms

Corner is important feature in the analysis and understanding of images or computer graphics. Corners are image points that show a strong intensity change. The detection of corner points in images is essential for many tasks such as machine vision, image processing. The reason that this approach is so popular is that corner points have a great deal of information, providing a sufficient constraint to vision processing and a reliable match between images, and having a great improvemence on speed. In military field, technique of corner detection is used for target recognition, shape analysis and so on, and technique of corner matching can be used for target tracking and target location.This paper attempts to analyze two algorithms of corner detection based on gray-level, and put forward improved method in fixed threshold of SUSAN operator, shortage in detection capability to X-shape corner and shortage in noise sensitivity of Mic operator. Experimental results show that the algorithm has better effect than before.