惯性/视觉/激光雷达SLAM技术综述CSCD

同时定位与建图(SLAM)技术已广泛应用于各类自主移动平台中,其中视觉SLAM和激光雷达SLAM是两种主要的SLAM技术方案。然而,视觉SLAM系统易受视觉环境变化的影响,而激光雷达SLAM系统则在结构单一等环境中会出现精度退化甚至失效的情况。随着智能移动平台应用场景的不断拓展,对SLAM系统的精度和鲁棒性等提出了更高要求,将多种具有互补性的传感器进行融合是提升SLAM系统性能的有效途径。据此,聚焦惯性/视觉/激光雷达多传感器融合SLAM技术,从多传感器标定和多源数据融合两个主要方面进行综述,最后对多传感器融合SLAM技术的发展趋势进行了展望。

A Review of Simultaneous Localization and Mapping Based on Inertial-Visual-Lidar Fusion

Simultaneous Localization and Mapping (SLAM) has been widely used in autonomous mobile robots and it can be broadly categorized as vision-based and lidar-based SLAM. However, visual SLAM is subject to visual condition variations and lidar SLAM degrades or even fails in structure-less environments. As different applications of autonomous robots continue to appear, the demand for accurate and robust SLAM is increasing, and fusing complementary sensors is a promising solution. Therefore, this paper focuses on the SLAM systems based on inertial, visual and lidar fusion, and reviews the techniques of multi-sensor calibration and data fusion. The development trend of multi-sensor fusion SLAM is prospected.