一种基于图像显著结构的噪声模糊图像盲目反卷积方法
李波; 郝明国; 吕进锋; 孙士洁; 赵怀慈
2017-07-18
专利国别中国
专利号CN106960417A
专利类型发明
产权排序1
权利人中国科学院沈阳自动化研究所
其他题名Based on image it has obvious structure the noise of the fuzzy image the blind de-convolution method
中文摘要本发明公开一种基于图像显著结构的噪声模糊图像盲目反卷积方法。包括以下步骤:输入待去模糊的图像数据;对输入图像数据进行降噪预处理;对降噪预处理后的图像进行显著边缘提取;对显著边缘提取后的图像进行Shock滤波重建图像强边缘;利用图像强边缘计算用于模糊核估计的图像显著边缘;初始模糊核估计;利用所估计的初始模糊核进行粗略图像复原;对粗略图像复原后的基于ISD的模糊核修正;图像复原。本发明可有效地处理图像去模糊对噪声敏感问题,准确地估计出噪声模糊图像的模糊核,并给出高质量的复原图像。
是否PCT专利
英文摘要The invention claims a based on image it has obvious structure the noise of the fuzzy image the blind de-convolution method. Comprises the following steps: Input to remove fuzzy image data; To the input image data to reduce the noise pre-treatment; Can reduce noise in the picture after treatment is carried out it has obvious edge extraction; Is obvious the edge extraction after the images to shock wave filtering reconstruction image strong edge; By using the image calculating edge intensity and it is used for nuclear fuzzy evaluation image has obvious edge; Initial estimating blurred core; Using the estimated initial fuzzy core to rough image restoration; The rough image restoration after isd based on fuzzy correction core; Image restoration. The invention can effectively process fuzzy image removing the noise is sensitive to solve the problem of accurately estimating the noise of fuzzy image fuzzy core and the high quality of the restored image.
申请日期2016-01-08
语种中文
专利申请号CN201610010171.1
专利代理沈阳科苑专利商标代理有限公司 21002
内容类型专利
源URL[http://ir.sia.cn/handle/173321/20883]  
专题沈阳自动化研究所_光电信息技术研究室
推荐引用方式
GB/T 7714
李波,郝明国,吕进锋,等. 一种基于图像显著结构的噪声模糊图像盲目反卷积方法. CN106960417A. 2017-07-18.
个性服务
查看访问统计
相关权益政策
暂无数据
收藏/分享
所有评论 (0)
暂无评论
 

除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。


©版权所有 ©2017 CSpace - Powered by CSpace