Image denoising method and device, equipment and medium

Abstract

The invention discloses an image denoising method and device, equipment and medium, and the method comprises the steps: classifying each pixel block in a search window based on a preset clustering algorithm, wherein the pixel blocks comprise a reference block and a candidate block, the reference block is a rectangular pixel block which takes a current to-be-denoised pixel point as a center and has a side length of N, and the candidate blocks are pixel blocks selected by sliding a neighborhood window in the search window; respectively determining a first distance between the reference block and each candidate block, and respectively determining a second distance between the reference block and a clustering center of a category to which each candidate block belongs; and based on the first distance and the second distance corresponding to each candidate block, respectively determining an NLM filtering weight between the current pixel point to be denoised and the central point of each candidate block; and carrying out denoising processing on the current pixel point to be denoised based on the NLM filtering weight. In this way, the noise reduction risk caused by noise can be reduced, and the image denoising effect is improved.

Description

technical field [0001] The present application relates to the technical field of image processing, and in particular to an image denoising method, device, equipment, and medium. Background technique [0002] Image denoising is a commonly used image processing technology. In the prior art, the NLM algorithm (Non Local Means, non-local mean) can be used to denoise the image. The NLM algorithm uses the similarity of the adjacent points and blocks of the natural image to estimate the original image. NLM assumes that in the case of image noise, the similarity of blocks can reflect the similarity of points in the case of no noise. [0003] In the existing NLM algorithm, when the noise level is relatively low, the NLM algorithm often has a good profit, but the greater the noise, the variance of the distance between blocks increases, and the similarity between the similarity of the block and the similarity of the point The correlation gradually decreases, that is, the noise reduces...

AI Technical Summary

Problems solved by technology

[0003] In the existing NLM algorithm, when the noise level is relatively low, the NLM algorithm often has a good profit, but the greater the noise, the variance of the distance between blocks increases, and the similarity between the similarity of the block and the similarity of the point The correlation gradually decreases, that is, the noise reduces the confidence of the inter-block distance, making the inter-block distance tend to be random, and the point with low similarity with the current point obtains a higher NLM filter weight in the weighting process, so that in the noise When the level is high, the denoising effect of the NLM algorithm will be greatly reduced

Images