首页 | 本学科首页   官方微博 | 高级检索  
     

基于小波变换的近平面影像数字配准方法研究
引用本文:苏毛弟,晏磊,赵学军,卢宾宾. 基于小波变换的近平面影像数字配准方法研究[J]. 影像技术, 2008, 20(1): 19-23
作者姓名:苏毛弟  晏磊  赵学军  卢宾宾
作者单位:1. 北京大学空间信息集成与3S工程应用实验室,北京,100871;中国矿业大学(北京)机电与信息工程学院,北京,100083
2. 北京大学空间信息集成与3S工程应用实验室,北京,100871
3. 中国矿业大学(北京)机电与信息工程学院,北京,100083
摘    要:本文将小波变换技术与相似度检测算法、直线拟合方法等相结合,用于近平面影像数字配准。首先,对含噪遥感影像进行多尺度小波变换,以提取不同尺度上的广义特征点,并剔除图像噪声;其次,将序贯相似度检测算法和多分辨率分析相结合进行控制点匹配;最后,利用基于最小二乘法的直线拟合,建立配准变换关系。实际遥感影像的相对配准实验验证了方法的有效性。

关 键 词:特征点  小波变换  配准
文章编号:1001-0270(2008)01-0019-05
收稿时间:2007-12-07
修稿时间:2007-12-07

Research and Applications of Digital Registration and A Feature Point Based on Wavelet Analysis in Near Plane Image
SU Mao-di,YAN Lei,ZHAO Xue-jun,LU Bin-bin. Research and Applications of Digital Registration and A Feature Point Based on Wavelet Analysis in Near Plane Image[J]. Image Technology, 2008, 20(1): 19-23
Authors:SU Mao-di  YAN Lei  ZHAO Xue-jun  LU Bin-bin
Affiliation:SU Mao-di,YAN Lei, ZHAO Xue-jun,LU Bin-bin (1. Beijing Key Lab. of Spatial Information Integration & Applications, PKU, Beijing 100871; 2. Institute of Mechanical Electronic and Information Engineering, China University of Mining and Technology, Beijing 100083)
Abstract:This article aims to an image registration method which combines the wavelet transform technology with the SSDA (Sequential Similarity Detection Algorithms) and least-square method. First, the muhiscale wavelet transform is used in a remote sensing image with noise, to extract generalized feature points in different levels and eliminate noise in the image. Secondly, the SSDA is incorporated with the multi-resolving analysis for the RCPs (Registration Control Points) match. Finally, the search of RCPs can be used to establish the function relationship by the least-square method. Experiment confirmed that the method can validly register the remote sensing image
Keywords:feature point   wavelet analysis    registration
本文献已被 维普 万方数据 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