Improved median filtering denoising algorithm and analysis
Jianxiong Wang, Jiwen Liu, Yaohui Li, Yulan Wang · 2012
This paper presents a statistical sort, the smallest neighbourhood linear variance denoising algorithm. And it is based on the median denoising algorithm theoretical. We not only analyze the performance of variance threshold to algorithm from theoretical and experimental, but also analyze this algorithm parameter adjustment strategy to achieve the optimized selection of the threshold. The present algorithm does not need to have any relevant information of the image noise of the known. Only uses the 3×3 filter window, it can effectively remove a variety of different densities noise and also be able to effectively protect the detail of the image borders information and the thin solid line.