Blind image blurring by Gaussian filtering extreme channels prior
Xue Long Hu, Chengquan Huang, Run Feng, Lihua Zhou, Lan Zheng · Proceedings of the 2021 5th International Conference on Electronic Information Technology and Computer Engineering · 2021
Previous extreme channels prior deblurring methods still cannot solve noise and avoid ringing artifacts in restored image problems well. In order to solve these caused, we propose a Gaussian filtering extreme channels prior (GFECP) to plug the Gaussian filtering into extreme channels prior work architecture for effective blind image deblurring. First, we computed image structure based on the Gaussian filtering model and extreme channels prior. Second, the half quadratic splitting technique is used to solve the non-convex problem of the model and estimate the clear image. Finally, we use an iterative multi-scale blind deconvolution method as regularization to restore the blur kernel. Experimental results clearly demonstrate that the proposed method outperforms state-of-the-art methods in terms of robustness, subjective visual effects and objective evaluation indexes.