An Image Deblurring Algorithm Based on Edge Selection
Liu Tingting, Kang Kai, Wang Tian-yun, Zhu Guoquan, Zhou Jian-xin · 2019
In the image deblurring problem, as the blurred images usually have less gradient total contrast than the clear images, the maximum a posterior estimation method prefers to the non-blur interpretation. Hence, this paper focuses on how edge selection can affect blurring kernel estimation, and an image deblurring algorithm based on edge selection is proposed, which can be used to estimate blurring kernel and latent clear image simultaneously under the framework of maximum a posterior probability estimation. Experimental result shows that the proposed algorithm can estimate the optimal blurring kernel, and the restored latent clear image can get better visual experience.