Canny Edge Detection Algorithm Based on Improved Sequential Statistical Filter

This paper describes an improved Canny edge detection algorithm to detect edge of infrared image better through overcoming the complexity and intensity of infrared image noise acquired by missile-borne radar seeker. Based on the traditional Canny algorithm architecture, the improved Sequential Statistical Filtering (ISSF) algorithm is used to replace Gauss filtering algorithm, and the P-entropy method (PEM) and OTSU threshold (PEM and OTSU, POM) are used to self-adaptively select the two thresholds for edge extraction. Experiments show that the algorithm can eliminate noise interference better and has higher edge detection effect for infrared images.