The edge extraction of Yellow River Model based on the principle of SUSAN

This paper analyzes the principle and availability of SUSAN algorithm of edge detection. Based on this, a method based on improved SUSAN algorithm to extract the boundary features of the Yellow River Model images is put forward. The method utilizes the self-adaptive threshold of gray level difference, and then combines the method of increasing the boundary pixels contrast to set a geometric threshold of its ups and collars. Finally, automatic identification and extraction of the Yellow River boundary is achieved. Experiments show that this method of edge extraction is good for de-noising. Its boundary is clear, true, meticulous, and location is accurate. It is convenient to measure parameters of the feature extraction, target recognition and matching for the next step.