一种新的基于粗糙集的数学形态学算法及其在图像边缘检测中的应用

A New Mathematical Morphological Algorithm Based on Rough Sets and It’s Application to Detecting Image Edge

从粗糙集的角度出发来研究数学形态学,指出经典数学形态学模型是不符合实际的,引入粗糙集中的粗糙度的概念改进了形态学模型,通过动态的选择结构元素来对图像进行处理。实验显示,基于粗糙集的数学形态学方法较好地解决了经典形态学处理中图像目标边界线条较粗的缺点。相比经典梯度算子,文中提出的方法又较好地保证了图像边界的连续性,同时还避开了结构元素的选择问题。

The paper studies the mathematical morphology from the point of view of rough sets, and concludes that the classical mathematical morphology desn＇t fall short of the fact. The paper improves the mathematical morphology model by introducing rough set. The new model processes the image by dynamic selecting the structure elements. Experiment shows that this algorithm resolve the disadvantage that the line of edge of target in the image is bold. Comparing with the classical grad operator, the algorithm ensures the continuity of the edge of image, meanwhile avoids the selection problems of structure elements.