快速非均匀模糊图像的盲复原模型  

Fast blind deblurring models for restoration of non-uniform blur images

作　　者：程俊廷[1] 左旺孟[2] Cheng Junting Zuo Wangmeng(School of Mechanical Engineering, Heilongjiang University of Science ＆ Technology, Harbin 150022, China School of Computer Science ＆ Technology, Harbin Institute of Technology, Harbin 150001, China)

机构地区：[1]黑龙江科技大学机械工程学院,哈尔滨150022 [2]哈尔滨工业大学计算机科学与技术学院,哈尔滨150001

出　　处：《黑龙江科技大学学报》2017年第2期196-199,共4页Journal of Heilongjiang University of Science And Technology

摘　　要：针对目前非均匀模糊图像盲复原算法在复原效果、计算和存储复杂性以及自动化程度方面存在的问题和不足,研究基于广义可加卷积模型的非均匀模糊计算方法,建立改善估计准确度或增强复原图像视觉质量为目标的图像先验模型,并改进模型的复原效果和效率。结果表明:非均匀模糊图像盲复原算法可有效复原出模糊图像。与传统算法相比,该算法复原效率较高,具有一定的实用价值。This paper follows from the need for addressing the problems with the existing non-uniform image deblurring methods in terms of restoration quality, computational and memory complexity, and automatic parameter selection. The study provides a novel model for improving the restoration quality and reducing computational and memory complexity; enhancing the computational efficiency for non-uniform blurring, using the generalized additive convolution model; and developing novel image prior models for enhancing image visual quality. The results demonstrate that the non-uniform blurred image blind restoration algorithm enables a more effective restoration of the blurred image and provides a higher restoration efficiency than the traditional algorithm.

关 键 词：图像盲复原 去模糊 退化图像 稀疏建模 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]