Spatial attention image denoising method integrating local information and global information

Abstract

The invention belongs to the field of image processing, and particularly relates to a spatial attention image denoising method integrating local information and global information, and the method comprises the steps: obtaining an image in real time, inputting the obtained image into a trained image denoising model, and obtaining a denoised image; wherein the image denoising model comprises an image shallow feature extraction module, an image feature enhancement group and an image reconstruction module. According to the image denoising model designed by the invention, a non-local module, a spatial attention residual block and expansion convolution are combined by adopting a series connection mode and a parallel connection mode at the same time to capture local features and global features of an image at the same time, so that the representation capability of a network is greatly enhanced; noise removal and image structure protection can be well considered at the same time.

Description

technical field [0001] The invention belongs to the field of image processing, and in particular relates to a spatial attention image denoising method integrating local information and global information. Background technique [0002] Image denoising is a long-standing ill-conditioned inverse problem, the purpose is to remove the noise in the image and restore the underlying clear image. The image denoising method based on deep learning is to use the neural network to learn the mapping relationship between the noise image and the clear image, so as to achieve the purpose of effectively removing noise. [0003] The current image denoising algorithms are mainly divided into two categories: optimization-based methods and learning-based methods. For the optimization-based method, this type of method is based on the statistical characteristics of the image to model the image denoising and iteratively optimize the solution. Therefore, it is inevitable to introduce heuristic param...

AI Technical Summary

Problems solved by technology

For the optimization-based method, this type of method is based on the statistical characteristics of the image to model the image denoising and iteratively optimize the solution. Therefore, it is inevitable to introduce heuristic parameter adjustment and very time-consuming The restoration process; for learning-based methods, the existing methods either only consider the global features of the image, or only consider the local features of the image, so they cannot effectively take into account the removal of noise and the protection of the local structure of the image

Images