Blind image deblurring based on the sparsity of patch minimum information
作者：
Highlights：
• We propose a strongly imposed zero patch minimum constraint for the latent clear image, which helps alleviate the ill-posedness of the inverse problem for blind image deblurring.
• We retrieve important fine details by assigning the patch minimum information obtained from the blurred image back to the latent image to further enhance its structure.
• We introduce an adaptive regularizer which was shown to have significantly better edge-preserving property than the total variation regularizer for the final image restoration of blurred images.
摘要
•We propose a strongly imposed zero patch minimum constraint for the latent clear image, which helps alleviate the ill-posedness of the inverse problem for blind image deblurring.•We retrieve important fine details by assigning the patch minimum information obtained from the blurred image back to the latent image to further enhance its structure.•We introduce an adaptive regularizer which was shown to have significantly better edge-preserving property than the total variation regularizer for the final image restoration of blurred images.
论文关键词：Sparsity,Patch minimum information,Blind image deblurring,Kernel estimation,Adaptive regularization,Variational model
论文评审过程：Received 5 April 2020, Revised 27 July 2020, Accepted 16 August 2020, Available online 17 August 2020, Version of Record 20 August 2020.
论文官网地址：https://doi.org/10.1016/j.patcog.2020.107597
