An improved image denoising method based on the pulse coupled neural network

In this paper, we proposed an image denoising filtering method based on PCNN (pulse coupled neural network) for images polluted by the pulse noise. The method firstly determined the position of pixels polluted by noise according to the fire capture features of PCNN. Then, a similar median filtering step was applied in the processing of noised pixels. Finally, the iterations and the sizes of filter window were chosen according to the noise intensity, which realized the adaptive-filtering of images. The main result stated that, compared to mean filtering, median filtering and adapted-median filtering, the method we proposed was not only effective in denoising and keeping details of images, but also showing good performances in different SNR conditions.