Compression artifacts removing method of image based on deep learning

Abstract

The invention discloses a compression artifacts removing method of an image based on deep learning. Artifacts generated in highly compression of the image can be effectively removed. The method is characterized in that firstly, by use of the newest deep learning technology, a deep residual error network is used as a basic module to be applied to a network model, so a gradient diffusion problem of the depth network model is effectively relieved, and meanwhile, base layer characteristics and high layer characteristics obtained in network learning are fused through skipping connection, so quite enriched characteristic information is provided for reconstruction of an artifacts removing image and the artifacts removing performance of the model is further improved; and secondly, the invention further provides a model selection scheme, so artifacts removing operation can be properly selected for proper models with compression artifacts in different degrees. According to the invention, two sets of public data sets are tested, so the performance of the provided method is remarkably improved than that of the current best artifacts removing algorithm.

Description

technical field [0001] The invention relates to the field of image processing and deep learning technology, in particular to an image decompression artifact method based on deep learning. Background technique [0002] The access to image data occupies a large amount of traffic on news websites, various social platforms, and e-commerce platforms, and image compression technology can reduce the number of bytes downloaded from websites as much as possible, thereby improving the loading speed of web pages or the image browsing speed of social platforms. Lossy compression methods such as JPEG, WebP and other technologies are widely used in news websites, WeChat, Weibo and other platforms. These compression technologies not only improve the response speed of the client, but also save storage costs and bandwidth costs for the platform. However, the lossy compression of the image will cause a certain degree of distortion, and the decoded image will have many artifacts, which will b...

AI Technical Summary

Problems solved by technology

However, the patent's de-JPEG compression artifacts refer to Dong [1] The network model proposed by et al. forms a cyclic 4-layer convolutional neural network, which still uses a shallow network for de-artifact operations, and its de-artifact performance is limited.

Images