Adaptive morphological filter based on image 2-D entropy

To maintain the detail information of images in the process of noise removal, a new adaptive morphological filter method was proposed based on the morphological filter in use. The 2-D image entropy has been used in this filter, and it has been constructed with the gray scale of every pixel and the gray scale distribution of neighbor region pixels. With the effective morphological arithmetic open and close, we set a reasonable flow in this filter. When finished the filter operation, the 2-D image entropy would always be calculated to reflect the effect of the filter operation, then the structure element for morphological arithmetic would be enlarged continuously. Based on the terminative qualification by 2-D image entropy, the filter operation ran adaptively and effectively with salt and pepper noise image. The definition of image 2-D entropy and adaptive morphological filter were discussed particularly.