An image denoising method based on an adaptive weighted total variational model

Abstract

The invention discloses an image denoising method based on an adaptive weight total variational model. An adaptive weighted total variational model is constructed. A non-local mean filter is used to process the noisy image to obtain an estimated de-noisy image. After preprocessing, four edge operators are used to detect the noisy image so as to obtain and edge detection image, and the adaptive normal parameters are defined according to the four edge operators. The pixels of the edge detection image are reset according to the adaptive normal form parameters to obtain the weight matrix of the noisy image. The denoised image is obtained by substituting the estimated denoised image, the adaptive normal form parameters and the weight matrix into the adaptive weighted total variational model. The invention can self-adaptively separate the edge region and the smooth region of the image according to the image characteristics, and the image texture and the edge are kept while noise is removed,thereby avoiding the generation of the step effect.

Description

technical field [0001] The invention belongs to the technical field of image processing, and relates to an image denoising method based on an adaptive weight full variation model. Background technique [0002] The process of acquisition, storage and transmission of digital images will inevitably be polluted by noise, and denoising is needed to improve the quality. In the past two decades, partial differential equation (PDE) methods have been widely used in edge-preserving image denoising, especially the Rudin-Osher-Fatemi (ROF) model, which includes TV model normalized images and noise-preserving images. The process of seeking energy minimization between true items, the model seeks an image very close to the original image, and achieves a good denoising effect. [0003] However, the Rudin-Osher-Fatemi (ROF) model has a significant disadvantage that it is prone to step effect, that is, the flat area of ​​​​the image produces false boundaries. Aiming at the shortcoming of to...

AI Technical Summary

Problems solved by technology

[0003] However, the Rudin-Osher-Fatemi (ROF) model has a significant disadvantage that it is prone to step effect, that is, the flat area of ​​​​the image produces false boundaries.

Aiming at the shortcoming of total variation (ROF), Song proposed a generalized total variation denoising model based on the L1+p norm. This model can overcome the generation of false edges, but it is difficult to objectively select different images p Greatly affects the image denoising effect

Zhang Hongying and others introduced the gradient information of the image into the total variation denoising model. The model automatically selects the TV model with better edge preservation in the edge area of ​​the image, and automatically selects the L2 model with better smoothness in the smooth area. It is very sensitive and does not have the ability of anti-diffusion, and the model diffuses smoothly along the vertical direction of the pixel gradient, losing some important texture information, and in many applications that rely on images, any loss of information may have serious implications, especially in medical applications where human health and life are affected

Images