Image denoising method based on improved adaptive neural network

Abstract

The invention discloses an image denoising method based on an improved adaptive neural network, and the method comprises the steps: firstly collecting a noiseless image and a noise-polluted image to form a data set, and carrying out the preprocessing of the data set; secondly, building and training a Cycle-GAN network, connecting a U-shaped neural network behind the Cycle-GAN network, and naming the whole network as a Cycle-GAN-Unet network; training the constructed network; adding the loss function of the convolutional neural network and the loss function of the Cycle-GAN; and finally, performing final denoising on the output of the whole trained network by using a filter to obtain a denoised image. The images of the training data do not need to be labeled in a one-to-one correspondence manner, only one group of noiseless images and one group of noisy images are needed, the finally trained system can perform noise reduction on the newly input noisy images to output the noiseless images, and the noise reduction capability is better than that of a single Cycle-GAN network.

Description

technical field [0001] The patent of the present invention belongs to the field of computer vision, and specifically relates to an image denoising method based on an improved adaptive neural network. Background technique [0002] Image processing has always been a hot research topic in the computer field, and image denoising is a prerequisite. We need to obtain clear pictures to facilitate the next step. In recent years, the rise of neural networks has injected new vitality into the field of artificial intelligence. The powerful fitting ability of neural networks can fit almost any function that is too complicated to imagine. In the field of computer vision, the proposal of convolutional neural network is even more groundbreaking. Its ability to extract image features is currently unmatched by other methods. Compared with traditional image processing such as filter denoising methods, neural network The network has a stronger universality. A network can cope with various fil...

AI Technical Summary

Problems solved by technology

[0004] Purpose of the invention: In order to solve the problem that the traditional image denoising method cannot cope with complex noise, the present invention

Images