计算机工程与应用 ›› 2019, Vol. 55 ›› Issue (7): 30-39.DOI: 10.3778/j.issn.1002-8331.1812-0148

• 热点与综述 • 上一篇    下一篇

基于RGB-D相机的SLAM技术研究综述

陈世浪,吴俊君   

  1. 佛山科学技术学院 机电工程学院,广东 佛山 528225
  • 出版日期:2019-04-01 发布日期:2019-04-15

RGB-D SLAM: a Survey

CHEN Shilang, WU Junjun   

  1. School of Mechatronics Engineering, Foshan University, Foshan, Guangdong 528225, China
  • Online:2019-04-01 Published:2019-04-15

摘要: 视觉同时定位与地图构建(Simultaneous Localization and Mapping,SLAM)是自主移动机器人、自动驾驶车和无人机关键技术,是当前机器人和计算机视觉领域的研究热点。对当前主流的SLAM技术发展历程及研究现状进行概述。重点围绕RGB-D SLAM研究展开讨论,包括:对各类视觉传感器进行性能分析和对比;对当前具有代表性的几类RGB-D SLAM系统的原理、方法和技术性能进行详细的分析;对RGB-D SLAM的关键问题、测试数据集及评估指标进行综合性归纳。最后,对视觉SLAM的发展趋势进行展望与总结。

关键词: 自主移动机器人, 同时定位与地图构建(SLAM), 深度相机, 图像语义, 机器学习

Abstract: RGB-D SLAM(Simultaneous Localization and Mapping) is a  key technology of autonomous mobile robot, autonomous vehicles and Unmanned Aerial Vehicle. It is a research hotspot in robotics and Compute Vision(CV). Firstly the paper summarizes the development history and current status of the prevailing SLAM technology. Then RGB-D SLAM research, including analysis and comparison of the performance of related sensors are followed. Principle, method and technical performance of state-of-art RGB-D SLAM system are analyzed. Key issues, datasets and evaluation indicators are summarized. Finally, the development trend of RGB-D SLAM is summarized and prospected.

Key words: autonomous mobile robot, Simultaneous Localization and Mapping (SLAM), RGB-D camera, image semantic, machine learning