Image denoising method based on non-local sparse model

Abstract

The invention discloses an image denoising method based on a non-local sparse model, and is mainly used for solving the problems that the details, textures and marginal region information are difficult to retain when a noise-containing image is subjected to sparse representation denoising in the prior art. The method comprises the implementation processes of: (1) solving similar set of each point neighborhood in the noise-containing image; (2) designing a sparse representation dictionary for the similar sets according to the sizes of the similar sets; (3) carrying out sparse decomposition and sparse reconstruction on the similar set data by using an SOMP algorithm utilizing the obtained dictionary, thus denoising the similar set data; and (4) summarizing the all denoising results of every point in the noise-containing image, and taking the average value as the final denoising result of the point, so as to further obtain the denoising result of the whole image. According to the invention, the image denoising effect and sparse representation efficiency of the image signals are improved, and the method can be used for target tracking and identifying.

Description

technical field [0001] The invention belongs to the technical field of image processing, and specifically is based on a non-local sparse model, uses similar set data to learn an accurate small dictionary, and achieves a method for efficiently sparsely reconstructing signals, which can be used for denoising of natural images. Background technique [0002] Due to the limitations of imaging equipment and imaging conditions in digital image processing, images are inevitably polluted by noise during the process of acquisition, conversion, and transportation. Therefore, image denoising occupies a pivotal position in the field of image processing and has become one of the most basic technologies in this field. Many practical noises in image processing can be approximated as Gaussian white noise, so removing Gaussian white noise in noisy images has become an important direction in the field of image denoising. [0003] Traditional denoising methods can be roughly divided into two c...

AI Technical Summary

Problems solved by technology

Since the training sample data is global and noisy, the trained dictionary also has noise, so when using the dictionary to sparsely represent the image signal, the details, textures and edge areas in the image cannot be accurately represented. Sparse reconstruction affects denoising effect

Images