Nonlocal Mean Denoising Method Based on Shape Adaptive Search Window

Abstract

The invention discloses a non-local mean value denoising method based on a shape-adaptive search window, comprising: step 1: inputting a noise image V, and step 2: calculating the gradient image V of the noise image V in the horizontal direction x and the gradient image V in the vertical direction y ; Step 3: Calculate the structure tensor T corresponding to the noise image V (s,σ) ; Step 4: Calculate the shape adaptive search window AS of the current pixel i i ; Step 5: Divide the image block and calculate the similarity weight value w(i, j); Step 6: Calculate the denoised pixel value of the current pixel i Step 7: Scan row by row and column by row, and sequentially scan Each pixel is processed from step 4 to step 6 until all pixels are processed, and the image after denoising is output. The method of the present invention makes the estimated value closer to the real value, and has better edge and texture detail information in the image. The protection ability, the denoising effect obtained is better than the non-local mean denoising method with fixed size and shape of the search window.

Description

technical field [0001] The invention belongs to the field of image denoising in image processing technology, and in particular relates to a non-local mean value denoising method based on a shape adaptive search window. Background technique [0002] Image is an important carrier for people to record and transmit information, and it will inevitably be polluted by noise during the process of collection and transmission. Many applications related to images, such as segmentation, registration, edge extraction, etc., usually require preprocessing with effective denoising algorithms to obtain more reliable results. Therefore, image denoising has always been an important research topic in image processing. On the basis of the classic neighborhood filtering algorithm, Buades et al. proposed the Nonlocal Means (NLM) denoising algorithm based on the image self-similarity, and applied it in the denoising processing of images and videos. It is proved that its performance is better than...

AI Technical Summary

Problems solved by technology

[0006] The invention provides a non-local mean value denoising method based on a shape-adaptive search window, which solves the problem that the existing non-local mean value de-noising algorithm has a large error between the estimated value and the real value, and the denoising effect is poor The problem makes the estimated value closer to the real value, and has better protection ability for the edge and texture details in the image, and the denoising effect obtained is better than the non-local mean denoising method with fixed size and shape search window

Images