Image denoising algorithm based on the median morphological filter

With respect to the instance of image denoising by morphological filter that causes useful information missing, a new median morphological filter algorithm is presented. The algorithm does the median value operation through erosion and dilation operations, improves and optimizes structuring element units. The experiment proves that the new method overcomes the inherent inadequate of traditional morphological limit operation, and effectively removes the impulse noise of image; especially in low signal to noise ratio environment, the denoising performance has obvious advantages than the traditional morphological filter and median filter algorithms. Therefore, it has a broad application prospect in image processing.