中文摘要：

针对空中三角测量影像连接点提取中存在的误匹配、点位多而分布不均和点位定位精度低等问题,提出了一种影像连接点均衡化高精度自动提取方法。首先采用分块SIFT技术进行特征的提取与匹配,并利用并查集数据结构进行特征点的高效多视追踪;然后采用提出的物方分块点位筛选算法对点位进行了相对均衡化的择优挑选;最后采用最小二乘匹配技术对得到的SIFT连接点坐标位置进行精化。选取中国嵩山遥感定标场的有人机影像和沙漠地区无人机影像作为试验数据,通过目视检查、像方反投影误差和检查点精度等3个指标进行了分析,结果表明本文方法有效克服了弱纹理和重复纹理导致的连接点提取与匹配困难,并改善了连接点分布的均匀性和提高了连接点的定位精度。

英文摘要：

To solve the problem of matching errors, redundant points with bad distribution,and low points position precision in photogrammetry aerial triangulation, an automatic tie points extraction with uniform distribution and high precision is presented. Firstly, Block SIFT technology is conducted on every image, then an unordered feature tracking method based on union-find set is adopted to detect the multi-view correspondences. Secondly, by using the proposed algorithm of points selection in object block space, the relatively balanced points are obtained. Finally, the precision of the image coordinates are improved by least square matching, in the experiment part, Songshan images and desert images are used to test the proposed method. Through the examination and analysis of three indicators, the visual inspection, back projection errors in image space, and the precision of check points, the results show that the proposed method can effectively overcome the difficulties in extracting and matching tie points caused by weak texture and repeated texture, enhance the uniformity of tie points distribution, and improve the position precision of tie points.