This paper designed a new edge detection[1,3] operator based on Krisch edge detector and direction encoding and combining the local average, while proposed heuristic search algorithm in the noise image based on segmentation edge of the self-enhanced. First, use small-scale Gaussian filtering in the noise image, and then use this edge detector for edge detection, and trajectory of each sub-search self-reinforcing, and finally according to the degree of self-reinforcing accumulation of noise to obtain the image edge. The results show that: This algorithm can effectively extract the real object from the noise, and to maximize the retention of details. And the performance better than Krisch and has some speed advantage
Image Edge Detector Based on Analog Correlator and Neighbor Pixels
