FPGA implementation based on a fast median edge detection system
Aiming at the problem that the classic Sobel edge detection system has inaccurate positioning and eliminates the blurred or impossible extraction of image edge feature extraction caused by the noise difference between Gaussian and pepper and salt,an improved fast median filtering algorithm combined with the eight-direction Sobel operator is proposed.Firstly,the fast median filter-ing algorithm is used to eliminate the image noise,and then the eight-direction Sobel operator is used to improve the accuracy of edge localization and improve the accuracy of edge information extraction in the face of noise.An improved fast median edge detection sys-tem accurately locates image edges in dense mixed noise.The improved algorithm doubles the amount of computation and uses FPGA hardware to implement it,and the experimental results verify the feasibility of the algorithm.
Mixed noiseFast median filteringSobel operator in eight directionsEdge detectionFPGA
万方数据
维普
