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

基于像方特征点的多视影像最小二乘匹配
引用本文:陈晓辉,马丽霞,王晓光.基于像方特征点的多视影像最小二乘匹配[J].测绘与空间地理信息,2013(10):248-254.
作者姓名:陈晓辉  马丽霞  王晓光
作者单位:吉林建筑大学城建学院土木工程系,吉林长春,130111
摘    要:为了能得到更精确的影像初始值,进行高精度的影像匹配,本文分别选取了两种物方多视匹配模型和不同匹配策略、不同约束条件下的影像粗匹配结果作为初始值进行最小二乘精匹配。实验结果表明物方匹配模型可以提供很精确的初始值,而金字塔、视差和共线方程共同约束下的最小二乘匹配精度更高,可靠性、鲁棒性及稳定性更好,验证了结合合理的匹配策略和约束方法的最小二乘优化可以将匹配精度达到1/10000精度。

关 键 词:UCD影像  匹配策略  多视匹配模型  最小二乘匹配

The Least-Square Matching of Multi-View Images Based on Photographical Feature Points
CHEN Xiao-hui , MA Li-xia , WANG Xiao-guang.The Least-Square Matching of Multi-View Images Based on Photographical Feature Points[J].Geomatics & Spatial Information Technology,2013(10):248-254.
Authors:CHEN Xiao-hui  MA Li-xia  WANG Xiao-guang
Institution:CHEN Xiao - hui, MA Li - xia, WANG Xiao - guang ( The City College of Jilin Architectural and Civil Engineering Institute, Changchun 130111, China)
Abstract:In order to get a more accurate image initial value to match image in high - precision, this paper selects two multi - images matching in object model and different matching strategies, under different constraints coarse matching results as initial values to match by the least square matching. Experimental results show that matching in object models can provide very accurate initial value. What' s more, pyramids, the common parallax and collinear equations under the constraint of least - square matching can get the results with higher accuracy, reliability, robustness, better stability and so on. With a reasonable method of least - square optimization and con- straints matching can get the results as the accuracy of 1 / 10 000.
Keywords:UCD image  matching strategy  multi-view matching model  least-square matching
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