图优化SLAM的嵌入式处理技术

Graph-based SLAM Embedded Processing Technology

基于图优化的同时定位与地图构建(simultaneous localization and mapping,SLAM),可以让移动机器人在未知环境下构建周围的地图同时确定机器人在地图中的位置。复杂环境下其精确度和计算效率均优于基于滤波的SLAM,现已成为主流的方法。图优化SLAM用图来表示和解决SLAM问题,对计算资源要求很高,这极大限制了其在嵌入式系统上实时应用。对此提出了一种基于NVIDIA TX2的图优化SLAM处理技术,根据其架构特点优化SLAM算法,加速处理,实现图优化的实时处理。实验证明,通过高性能嵌入式处理硬件与算法优化的有机结合,可有效提升图优化SLAM在嵌入式系统中的处理性能。

Graph-based simultaneous localization and mapping allows the mobile robot to construct the surrounding map in an unknown environment while determining the robot’s position in the map.The accuracy and computational efficiency of Graph-based SLAM are better than those of the filter-based SLAM in complex environment,which has become the mainstream method.Graph-based SLAM uses the graph to represent and solve the SLAM problem,which requires high computational resources and greatly limits its application in real-time on embedded systems.This paper proposes an optimized SLAM processing technology based on NVIDIA TX2 and optimizes the SLAM algorithm according to its architecture characteristics to speed up processing and achieve real-time processing of graph optimization.The experiments have proved that the organic combination of high-performance embedded processing hardware and algorithm optimization can effectively improve the processing performance of Graph-based SLAM in embedded systems.