Robust edge detection in noisy images

摘要: We describe a new edge detector based on the robust rank-order (RRO) test which is useful alternative to Wilcoxon test, using rxr window for detecting edges of all possible orientations in noisy images. Our method testing whether spatially partitioned into two subregions having significant differences local gray-level value between adjacent pixel neighborhoods given pixel, an edge-height model extract some sufficient height from images corrupted with noises. Some experiments statistical detectors and T-test, well-known Canny our RRO are carried out synthetic real by both Gaussian impulse noise. The results show that performance proposed appears be most variations noise, performing well noise distributions tested.