Image fusion denoising method and system based on U-Net and pyramid module

Abstract

The invention relates to an image fusion denoising method and system based on an U-Net and a pyramid module. The method comprises the following steps of: S1, preprocessing paired original noise images and noise-free images, and obtaining paired image blocks for training; S2, selecting an image denoising deep network model FFDNet, and loading trained model parameters; S3, constructing an image fusion denoising network based on the U-Net and the pyramid module; S4, inputting the paired image blocks into the FFDNet by taking the batch as the unit, inputting an obtained result into the image fusion denoising network, and training the image fusion denoising network according to a loss function; and S5, inputting the original noise images for testing into the FFDNet, and inputting an obtained result into the trained image fusion denoising network to obtain final denoised images. According to the method and the system, the denoising performance is improved, and more image details are reserved.

Description

technical field [0001] The invention belongs to the technical field of image and video processing, and in particular relates to an image fusion denoising method and system based on U-Net and a pyramid module. Background technique [0002] In recent years, with the continuous advancement of science and technology and the continuous upgrading of mobile digital devices, the way for people to obtain information has become easier and faster. Compared with text, images are one of the most intuitive ways for people to obtain information. In a clear image, the outline of the object is clear, the foreground and background of the image are clearly visible, and people can accurately classify the image content and perform operations such as target recognition. However, in the process of image acquisition, often due to the interference of some different types of external factors, such as camera shake, signal disturbance, etc., the quality of the generated image is degraded, and this deg...

AI Technical Summary

Problems solved by technology

Although the denoising results of deep learning methods are better than those of traditional methods, most of the current methods still have the problem of not being able to preserve image details well, and the denoising results obtained after denoising with different intensities on the same noisy image Images have different performances in texture-rich regions and texture-smooth regions, and effective image fusion can better preserve image details

Images