首页 | 本学科首页   官方微博 | 高级检索  
     检索      

基于先验知识的模糊影像运动模糊核估计研究
引用本文:喜文飞,史正涛,钱堂慧.基于先验知识的模糊影像运动模糊核估计研究[J].测绘工程,2020(2):1-5.
作者姓名:喜文飞  史正涛  钱堂慧
作者单位:云南师范大学旅游与地理科学学院;云南海钜地理信息技术有限公司
基金项目:云南省科技厅基金项目(2017FB078);云南师范大学教改项目(YNJG201845)。
摘    要:无人机模糊影像恢复的效果取决于模糊核的精确度,高质量的模糊核是模糊影像恢复的关键。文中提出了一种利用先验知识预估模糊核的算法。根据无人机相邻影像,结合随机模糊核生成模糊影像的样本数据库,利用模糊影像和样本影像之间的残差,拟合多项式,计算模糊核。结合实验进行验证,对获取的无人机模糊影像进行验证,运动模糊影像水平方向模糊核的预估准确度达到100%,结合维纳滤波算法进行影像恢复,该方法可以很好的恢复运动模糊影像。

关 键 词:无人机模糊影像  模糊核  维纳滤波  影像恢复

Research on motion fuzzy kernel estimation of fuzzy image based on prior knowledge
XI Wenfei,SHI Zhengtao,QIAN Tanghui.Research on motion fuzzy kernel estimation of fuzzy image based on prior knowledge[J].Engineering of Surveying and Mapping,2020(2):1-5.
Authors:XI Wenfei  SHI Zhengtao  QIAN Tanghui
Institution:(College of Tourism and Geographic Sciences,Yunnan Normal University,Kunming 650500,China;Yunnan Haiju Geographic Information Technology Co.,Ltd,Kunming 650000,China)
Abstract:The effect of UAV fuzzy image restoration depends on the accuracy of fuzzy kernel,and high quality fuzzy kernel is the key to fuzzy image restoration.In this paper,an algorithm for predicting fuzzy kernels using prior knowledge is proposed.According to the adjacent images of UAV and the random fuzzy kernel,the sample database of fuzzy image is generated.The residual error between fuzzy image and sample image is used to fit the multinomial,and the fuzzy kernel is calculated.Combined with the experimental verification,the obtained UAV fuzzy image is verified,and the prediction accuracy of the horizontal direction blur kernel of the motion blur image is 100%.Combined with Wiener filtering algorithm for image restoration,this method can recover the motion blur image very well.
Keywords:UAV fuzzy image  blur kernel  Wiener filtering  image restoration
本文献已被 CNKI 维普 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号