An improved image denoising algorithm based on adaptive fractional integral
Jingxue Sun, Chunyang Wang, Xuelian Liu, Huanhuan Xia · 2017
In order to improve the effect of image denoising, in the background of the salt and pepper noise, we proposed an improved adaptive image denoising algorithm based on adaptive fractional integral in this paper. First of all, the noise points was detected by the method that detect salt & pepper noise based on the number, the noise image is divided into the flat area, the image edge and the noise points in three parts. Then use the discriminate function of “noise-edge” to do the second detection to the noise points and edge points which is detected in the first step. At the same time, adaptive fractional order was constructed according to the local statistical information and the structural characteristics. In the end, do the denoising to those suspicious noise points with the fractional integral mask. Compared with the traditional fractional integral denoising algorithm, MATLAB experiments show that the proposed adaptive algorithm effectively preserve the noise points and edge points that are misjudged and and realize adaptive the fractional integral order. The adaptive algorithm can not only remove the noises largely but also keep the edges and texture details well.