An Image Denoising Method Based on Generative Adversarial Network

Abstract

An image denoising method based on a generative confrontation network disclosed by the present invention is characterized in that it comprises the following steps: Step 1: Select an experimental data set; Step 2: Select Gaussian additive white noise as a noise model; Step 3: Build and generate Network model, train the generator network G for denoising; Step 4: Build a discriminative network model, and the discriminator D is used to classify the authenticity of the input image; Step 5: Build a joint loss function model; Step 6: Train the generative confrontation Network; Step 7: Image Denoising Quality Evaluation. The invention is an image denoising method based on generating confrontation network, which can achieve the denoising effect of retaining more texture details and edge features.

Description

technical field [0001] The invention belongs to the technical fields of image analysis, deep learning and computer vision, and in particular relates to an image denoising method based on a generative confrontation network. Background technique [0002] As a similar description of objective objects, images have an incomparable amount of information and intuition that cannot be compared with text and other media, but images are inevitably polluted by noise during the process of collection, storage, transmission, and use. The existence of noise makes the quality of the image uncontrollably degrade, even loses important information of the image, changes the pixel value of the original image, brings a great negative impact on computer vision processing, and directly affects the subsequent image processing. Therefore, how to reduce the noise pollution of the image, and at the same time remove the noise and restore the original information from the polluted image, has been a hot is...

AI Technical Summary

Problems solved by technology

However, most of these traditional denoising methods have defects such as blurred edges and unclear features due to the loss and neglect of detailed information such as structure, texture and edges of image features; on the other hand, with the improvement of computer hardware level, deep learning neural network It has entered a period of rapid development, and many scholars have turned their research attention to the application of deep learning technology in image processing, and have achieved certain results

Images