Mathematical Morphological Edge Detectors For Noisy Images Corrupted By Impulses
脉冲噪声污染图象中的数学形态边缘检测器

This paper extensivey studies the application of mathematical morphology to the field of edge detecting. The definitions of morphological differentiating operators are systematically presented. In the mean time, several anti-noises morphological differentiating operators are developed to deal with the images corrupted by impulses. It is shown by computer simulations that, by appropriately using the four basic operators of mathematics morphology, the influence of impulsive noise on edge detecting will be eliminated by morphological differentiating operators. The morphological edge detectors will outperform other edge detectors.