Study on One-step Singular Correlation Identifying and the Most Relevant Weighted Filtering for Images

摘要

Based on the correlation properties of the adjacent pixels in images, the characteristics of image de-noising and the insufficiencies in current methods for remedying values are analyzed. A kind of image filtering, that is, one-step singular correlation identifying and the most relevant weighted filtering (OSSC-MRWF) method is put forward. The simulation experiments and comparisons with some current common methods under the salt-pepper noises are done. The results show that this method not only could accurately detecting the salt-pepper noises and effectively remedying the noise points in images, but also well protect the original information such as the edge details. The image filtering accuracy and performance criterions are further improved considerably.