论文部分内容阅读
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。
This paper presents an algorithm for automatic registration of LiDAR point cloud based on point-line and line-like invariance in urban area. Firstly, SIFT operator is used to extract point features and perform coarse registration. At the same time, linear features are extracted based on image and LiDAR point cloud respectively. Then, the similarity invariants of local features and line features are matched. 2D-3D tight registration model based on extended co-linear equations was used to achieve fine registration of aerial images with LiDAR point clouds. The method is characterized by adopting a coarse-to-fine registration strategy, combining the intensity-based registration algorithm and the registration algorithm based on the line features through point-line similarity invariance, and at a higher degree of automation Realize accurate registration of image and point cloud. Experiments show that compared with the auto-registration algorithm based on point cloud strength images, the proposed algorithm can achieve better registration results in urban areas.