A novel algorithm for automatic registration of urban aerial images with airborne light detection and ranging (LiDAR) data sets was proposed .First of all ,a coarse registration step was applied by the SIFT operator and in the meanwhile ,straight line features were extracted from images and LiDAR point clouds respectively .And then , matched points were utilized to search matched lines via line-point similarity invariants in local areas .Finally , direct registration function model of 2D images and 3D LiDAR points based on line features was applied to achieve accurate registration .The main advantage of the proposed algorithm was that the coarse-to-fine matching strategy was adopted ,which effectively combines traditional registration method based on LiDAR intensity images and the algorithm on the basis of line feature registration model , and thus an accurate registration result under a high degree of automation was implemented .The experimental result shows that this algorithm can achieve more accurate results in urban areas comparing with the algorithms on the basis of LiDAR intensity images .%提出一种基于点、线相似不变性的城区航空影像与机载LiDAR点云自动配准算法。首先通过SIFT算子提取点特征并进行粗配准，同时分别基于影像和LiDAR点云提取直线特征；然后利用局部区域点特征与线特征的相似不变性，通过匹配点对搜索匹配直线对；最后采用基于扩展共线方程的2D-3D严密配准模型实现航空影像与LiDAR点云的精配准。本方法的特点是：采取了由粗到精的配准策略，通过点、线相似不变性，将基于强度的配准算法和基于线特征的配准算法有机结合，在较高的自动化程度下实现了影像与点云的精确配准。试验证明，与基于点云强度影像的自动配准算法相比，本文的算法在城市地区能够取得较好的配准结果。
展开▼
