Image Denoising Research Based on Extreme Filtering
Lihua Sun, Enliang Zhao · Proceedings of the 2018 3rd International Conference on Automation, Mechanical Control and Computational Engineering (AMCCE 2018) · 2018
Median filtering algorithm can make the edge details of the image fuzzy when it is used to remove noise of an image.In this paper a method based on extreme filtering is put forward for image denoising.Firstly extreme filtering algorithm is used to determine whether a pixel is a signal point or a suspicious noise point in a small 3 3 × window, then the same extreme filtering algorithm is used in an expanded 5 5 × window.The suspicious noise points are further defined as signal points and noise points with some pixels in the small window.Median filtering is used to process the noise points while the original pixel values are retained for the signal points.This method is used in numerical experiments.The results show that the proposed method for image denoising can effectively avoid fuzzy image while texture details of image is due to the loss and can obtain relatively clearer image.The denoised image can retain the image edges of the image and can improve the peak signal-to-noise ratio.