Algorithm of Adaptive Median Filter for Salt and Pepper Noise Based on CUDA
Xiaoyu Liu · Aeronautical Computing Technique · 2013
According to excellence and shortcoming of conventional median filter algorithm,combined with the salt and pepper noise characteristics,an effective adaptive median filter algorithm is proposed.The method works in this way:Firstly,According to the salt and pepper noise characteristics,the pixels of the image are divided into noise and signal,then the size of filtering window is adaptively adjusted according to the number of noise points,Finally,an adaptive median image smoothing algorithm is used to filter noise pixels.But with the noise density and image size increases,execution on the CPU time increases significantly.This article analyses and uses parallel processing of the graphics processor(GPU),and achieves the algorithm in the CUDA platform.The experimental results show that the proposed algorithm can not only effectively remove salt and pepper noise,preserve the edges and details,but also can reduce the time consumption significantly.As the image size increases,it can get a maximum speed up ratio of 6 000 times.