Implementation Of Harris Corner Matching Based On FPGA

Abstract

A method for applying FPGA to realize the Harris corner detection and matching, aiming to overcome the problem that it is slow to fetch feature information from the image which show the planar motion of robot. Considering the theory of Harris corner detection and the structure of FPGA, the algorithm is improved in order to simplify calculation and simulated. The simulation result show that the algorithm which guarantee of corner extraction could decrease the number of redundancy corners and raise the real-time performance of system. Otherwise it can also lower the difficulty of design and reduce the hardware overhead, in favor of analyzing the movement of robot.