大重叠度无人机遥感影像的三维建模方法研究  

Research on the 3-dimensional modeling method for remote sensing images of big degree of overlap acquired by unmaned aerial vehicle

作　　者：崔红霞[1] 林宗坚[2] 孙杰[2] 

机构地区：[1]武汉大学遥感信息工程学院 [2]中国测绘科学研究院,北京100039

出　　处：《测绘科学》2005年第2期36-38,共3页Science of Surveying and Mapping

基　　金：国家重点基础研究发展计划 ( 973 )项目 ( 2 0 0 2CCA0 4 70 0 )

摘　　要：探讨了无人机遥感系统获取的遥感影像的三维建模方法 ,应用相对定向直接进行无人机遥感影像相对定向 ,并通过实验分析了相邻影像间重叠度对相对定向精度的影响。基于高重叠度的无人机遥感影像 ,为提高同名点识别的精度 ,提出了多幅影像核线约束方法。In this paper,a 3-dimensional modeling method is discussed for remote sensing images acquired by UAVRS^().Direct solution algorithm for relative orientation is applied to deal with the relative orientation for remote sensing images of low altitude.Base on experiments,analysis is conducted on accuracy of relative orientation influenced by degree of overlap between adjacent images.With the remote sensing images of big degree of overlap,a method for identification of the conjugate points is put forward to increase the accuracy of matching points.The method is based on the principle of epipolar line constraint among multiple images with big degree of overlap between each other.Experiments indicate that the method for 3-dimensional modeling is applicable to deal with the remote sensing images acquired by unmaned aerial vehicle remote sensing system.

关 键 词：无人机 相对定向 影像匹配 核线 

分 类 号：TP79[自动化与计算机技术—检测技术与自动化装置]