重庆邮电大学学报（自然科学版）Issue(1)：87-94,8.DOI:10.3979/j.issn.1673-825X.2016.01.013
显著边缘直接引导的动态模糊图像盲复原方法
Image blind restoration algorithm based on obvious edge feature
摘要
Abstract
For the problem of the imaging blurring caused by motion of the vehicle or vibration of the active vision plat in vision measurement,a new algorithm of image blind restoration is proposed by obvious edge feature.Firstly,an algorithm for searching and enhancing image edges is designed to overcome the drawbacks of the Canny algorithm.Further,a method of optimum estimation for the PSF is proposed only based on the neighborhood area of the obvious edges.Secondly,a recur-rence residual algorithm is proposed for image deblurring,which can restrain the ringing of the restored image caused by ob-vious edges.Lastly,a multi-layering with iteration flow path is designed for guaranteeing the accuracy and reliability of the blind restoration.Experiments demonstrate that the method can reconstruct images and suppress the ringing effectively.The experiment under un-ideal illumination condition also shows the good adaptability against environment.This image restora-tion new algorithm can be efficient in practical application.关键词
盲复原/显著边缘/点扩散函数(PSF)/递归残差/分层迭代Key words
blind restoration/obvious edges/point spread function(PSF)/recurrence residual/multi-layering with itera-tion分类
信息技术与安全科学引用本文复制引用
付强,孙秀霞,刘树光,徐嵩,彭轲..显著边缘直接引导的动态模糊图像盲复原方法[J].重庆邮电大学学报（自然科学版）,2016,(1):87-94,8.基金项目
中国博士后科学基金（2014M562629）；陕西省自然科学基金（2014JM8332）Foundation ItemsThe China Postdoctoral Science Foundation （2014M562629）；The Natural Science Foundation Project of Shaanxi （）
