基于双多线激光雷达的非结构化环境负障碍感知技术

负障碍感知是非结构化环境下的难点问题，本文针对该问题提出一种新的基于双多线激光雷达（Light detection and ranging，LiDAR）的感知方法.采用分布嵌入式架构对双激光雷达数据进行同步采集与实时处理，将雷达点云映射到多尺度栅格，统计栅格的点云密度与相对高度等特征并标记，从点云数据提取负障碍几何特征，通过将栅格的统计特征与负障碍的几何特征做多特征关联找到关键特征点对，将特征点对聚类并过滤，识别出负障碍.方法不受地面平整度影响，已成功应用在无人驾驶车上.使用表明该方法具有较高的实时性和可靠性，在非结构化环境下具有良好的感知效果.

Negative Obstacle Perception in Unstructured Environment With Double Multi-beam LiDAR

Negative obstacle perception is a difficult problem in unstructured environment, and a new negative obstacle perception algorithm in unstructured environment with double multi-beam light detection and ranging (LiDAR) is proposed. Firstly, a distributed embedded architecture for LiDAR data acquisition and processing is designed. Secondly, LiDAR points are projected to multi-scale gird maps and points density as well as relative height of each cell is computed, with each cell marked according to the feature. Then the geometric feature of negative obstacles is extracted from point cloud, the key points in pair are searched with both statistical characteristics and geometric features. Finally, clustering algorithm is used to recognize negative obstacles. The algorithm does not depend on the flatness of the ground and has been successfully applied to an unmanned ground vehicle. The application shows that the algorithm is real-time, reliable and has good detection ability.