关键词: 点云配准, 迭代邻近点算法, 点对约束, 距离约束, 参考点

Abstract: To improve the speed and accuracy of point cloud registration is the research emphasis. An iteration close point algorithm improved by distance constraint is proposed. With the registration point found by using close point criterion, the nearest principle is used to remove the point pairs which have the same points. Since the center of registration point is regarded as a reference point, after the wrong point pairs are removed by using point pairs distance constraint, point cloud registration is carried out. This improved algorithm is compared with another method which regards the center of point cloud as a reference point and uses iteration close point principle. Experiment results show that the speed and accuracy of registration for the algorithm are improved.

Key words: point cloud registration, iterative closest point algorithm, point pair constraint, distance constraint, reference point