Image or video quality enhancement method based on convolution neural networks

Abstract

The invention discloses an image or video quality enhancement method based on convolution neural networks, and belongs to the field of computer vision. The method comprises the steps of firstly, designing the two convolution neural networks for video quality enhancement, wherein the two networks have different computational complexities; secondly, selecting a plurality of training images or videos to train parameters in the two convolution neural networks; thirdly, selecting the convolution neural network with the more suitable computational complexity according to actual needs to input the images or videos with the quality to be enhanced into the selected network; finally, outputting the images or videos after quality enhancement through the network. According to the image or video quality enhancement method based on the convolution neural networks, the quality of the videos can be effectively enhanced; a user can select the specified convolution neural network with the more suitable computational complexity to enhance the quality of the images or videos according to the computational capacity or remaining power capacity of equipment.

Description

technical field [0001] The invention belongs to the field of computer vision, in particular to an image or video quality enhancement method based on a convolutional neural network. Background technique [0002] In the field of computer vision, video quality enhancement has an important impact on improving video (or image) quality and improving video (or image) visual effects; video (or image) quality enhancement generally refers to improving the quality of damaged video (or image) the quality of. In the current communication system, the problem of limited channel bandwidth exists widely, so the video (or image) transmission needs to go through the process of compression coding, in this process, the video (or image) quality will be lost; at the same time, the transmission channel often exists Noise, which will also lead to the loss of video (or image) quality after channel transmission; therefore, video (or image) quality enhancement has become a key issue in the field of co...

AI Technical Summary

Problems solved by technology

[0004] Aiming at the current problem that image or video quality needs to be adapted to multiple intelligent terminals, and image or video quality needs to be realized according to terminal conditions, the present invention provides an image or video quality enhancement method based on convolutional neural network. Currently, there is no such method based on volume Research on Video or Image Quality Enhancement by Productive Neural Network

Images