Novel Binary Adaptive Morphological Operators
Yanbo Li, Junping Wang · 2018
In this paper, novel binary adaptive neighborhood morphological operators are proposed to solve the problem that the fixed structure elements cannot adapt to the whole image when processing images in classical mathematical morphology. The key of adaptive morphology is to select suitable structure elements according to the different attributes of an image. On the basis of image local features and pixel position information, the structure elements selected in this paper can be adaptively adjusted. Besides, the new structure elements have symmetry, namely they are symmetrical about the center point. New binary adaptive neighborhood morphological operators are also defined according to the new structure elements. Comparing with the experimental results of the classical morphological operations, the new operators can retain the details of the image better and be more similar to the original image, which shows the new operators is more effective and practical.