A Fast and Effective Image Denoising Algorithm
Zuoyong Xiang · Research Square · 2022
Abstract Classic denoising algorithms have always suffered from the disadvantage of being slow, even for the famous BM3D.This paper presents a fastst and effective image denoising algorithm, which is based on the aggregation. It is a relatively faster method that first applies other known denoising algorithms (such as DCT for Discrete Cosine Transform, Wavelet Transform, Wiener Filter Denoising, etc.), selects different chunking specifications and different parameters to obtain several different denoised images, and then aggregate them to obtain the final denoised image.However, not arbitrarily generated noise reduction image aggregation can get good results, only the right combination can get good denoising results.One of the suitable combinations is proposed in the paper guided by the above ideaExperiments show that the algorithm has good noise reduction performance, and the algorithm has the fastest running time among all classical algorithms, and the running time is only equivalent to less than one-tenth of other algorithms.