一种无人机影像空三加密点抽稀算法  

Aerial Triangulation Points Thinning Algorithm for UAV Image

作　　者：王刊生[1] 郑亮[1] WANG Kansheng,ZHENG Liang(CCCC Second Highway Consultants Co.Ltd., Wuhan 430056, Chin)

机构地区：[1]中交第二公路勘察设计研究院有限公司,湖北武汉430056

出　　处：《测绘通报》2018年第3期108-112,共5页Bulletin of Surveying and Mapping

基　　金：中交第二公路勘察设计研究院科技研发项目(KJFZ-2015-075)

摘　　要：在无人机空三加密中,特征点匹配多采用基于计算机视觉的匹配算法进行处理,对稀少的空中三角测量控制点进行控制点位的补充。由于无人机影像分辨率高,细节丰富,使得匹配特征点数量庞大,给后续的区域网平差带来困难。本文提供的空三加密点抽稀算法,可大幅降低加密点的数量,优先保留优质点位,并保证被保留点位均匀分布。In UAV aerial triangulation, feature point matching is mostly processed by matching algorithm based on computer vision, to supplement the number of ground control points. Because of the high resolution and rich details of the UAV image, the number of matehing feature points is huge, which brings difficulties to the subsequent block adjustment.The points thinning algorithm is provided in this paper, whieh greatly reduces the number of matching feature points,preserves the high quality points first, and ensures the uniform distribution of the reserved points.

关 键 词：无人机 空三加密 影像匹配 点抽稀算法 

分 类 号：P237[天文地球—摄影测量与遥感]