基于不同代数结构的图象形态滤波

Image morphological filter based on different algebra structure

用于表示图象的模型空间上的形态算子不仅与模型空间的几何结构有关,同时还依赖于模型空间的代数结构.通过在目标空间中引入加性及乘性代数结构的方法,提出了基于目标空间不同代数结构的加性及乘性膨胀和腐蚀运算、以及以此为基础所建立的形态滤波,并通过对于其性质的讨论研究了形态滤波的算法及性能.指出了加性及乘性开、闭滤波在噪声滤除上由于目标空间代数结构的不同产生差异的原因.在此基础上,还对具有不同代数结构的灰值形态滤波进行了讨论.

The morphological operators are interrelated with geometry structure and algebra structure of model space used to represent an image.Two morphological operations with addition and multiple algebra structure for binary image-dilationa and erosion operation are proposed firstly.Their basic properties and algorithms,which established in equivalence representation of dilation and erosion,are investigated.Through the morphological opening opeator's different functions in image processing, show that the two different algebra structure will effect the morphological operator in some degree.