Real Time Sobel Square Edge Detector for Night Vision Analysis

Abstract

Vision analysis with low or no illumination is gaining more and more attention recently, especially in the fields of security surveillance and medical diagnosis. In this paper, a real time sobel square edge detector is developed as a vision enhancer in order to render clear shapes of object in targeting scenes, allowing further analysis such as object or human detection, object or human tracking, human behavior recognition, and identification on abnormal scenes or activities. The method is optimized for real time applications and compared with existing edge detectors. Program codes are illustrated in the content and the results show that the proposed algorithm is promising to generate clear vision data with low noise.