【摘要】 论文探讨适用于激光雷达同时定位与地图构建系统的室外点云实时闭环检测算法，其任务是如何在大规模室外环境下快速准确完成位置识别，以便构建更加鲁棒的SLAM算法。论文提出了一种基于位置和改进的点云全局描述符相融合的闭环检测算法，并设计了新型的适用于该算法的点云帧间相似度匹配计算公式，最后使用一致性校验来尽可能避免误匹配。实验结果表明，论文算法相比于改进前的闭环检测算法无论是在时间还是准确率上都更具竞争力。更多还原

【Abstract】 This paper discusses the outdoor point cloud real-time closed-loop detection algorithm suitable for lidar simultaneous positioning and map construction system. Its task is to quickly and accurately complete location recognition in a large-scale outdoor environment in order to build a more robust SLAM algorithm. This paper proposes a closed-loop detection algorithm based on the fusion of location and improved point cloud global descriptors,and designs a new type of point cloud frame similarity matching calculation formula suitable for this algorithm,and finally uses a consistency check to do the best. The experimental results show that the proposed algorithm is more competitive in both time and accuracy than the closed-loop detection algorithm before the improvement.更多还原