基于激光里程计和点云描述符匹配的自动驾驶车辆定位方法研究

Research on Automatic Vehicle Localization Method Based on Lidar Odometry and Point Cloud Descriptor Matching

近年来,随着信息技术的发展,汽车逐渐走向了智能化、无人化,自动驾驶是其中的一个重要特征。在自动驾驶系统中,精确的定位结果是实现合理路径规划和车辆安全行驶的前提。但当前传统基于概率特征的匹配定位方法,在环境发生变化时匹配精度降低,并且在定位初始化条件较差的情况下容易陷入局部最优,而无法得到全局的最优匹配。针对这些问题,本文提出将实时的激光里程计定位方法和基于匹配定位的方法结合,构建了一种基于三维点云描述符的匹配方法,提高了定位系统的鲁棒性和定位精度。

In recent years,with the development of information technology,automobiles have gradually become intelligent and unmanned,and autonomous driving is one of the important features.In the autonomous driving system,accurate localization results are a prerequisite for achieving reasonable path planning and safe vehicle driving.However,the current traditional matching localization method based on probabilistic features decreases the matching accuracy when the environment changes,and it is easy to fall into the local optimum without getting the global optimum matching under the poor initialization conditions of localization.To address these problems,this paper proposes to combine the real-time laser odometry localization method and the matching-based localization method,and constructs a matching method based on 3D point cloud descriptors to improve the robustness and localization accuracy of the localization system.