Variational JPEG artifacts suppression based on high-order MRFs

作者：

Highlights：

• A variational model combining a Fields of Experts prior and the quantization constraint set is propose for JPEG deblocking.

• The exploited QCS is usually lacking in traditional JPEG deblocking approaches.

• The resulting non-convex optimization problem is efficiently solved by a forward-backward splitting algorithm.

• Our model leads to higher PSNR-B results and visually comparable performance to state-of-the-art deblocking methods.

摘要

Highlights•A variational model combining a Fields of Experts prior and the quantization constraint set is propose for JPEG deblocking.•The exploited QCS is usually lacking in traditional JPEG deblocking approaches.•The resulting non-convex optimization problem is efficiently solved by a forward-backward splitting algorithm.•Our model leads to higher PSNR-B results and visually comparable performance to state-of-the-art deblocking methods.

论文关键词：JPEG deblocking,Fields of Experts,Non-convex optimization,MRFs

论文评审过程：Received 2 February 2016, Revised 21 December 2016, Accepted 21 December 2016, Available online 24 December 2016, Version of Record 30 December 2016.

论文官网地址：https://doi.org/10.1016/j.image.2016.12.006