视觉和激光SLAM发展综述与展望

打开文本图片集

中图分类号：TB9;TB3 文献标志码：A 文章编号：1674-5124（2025）06-0001-16

Overview prospect vision lidar SLAM

CHENG Liang1²， LUO Zhoul²， HE Yunze³， CHEN Yongcan³

（1. School ， ， 2. ， 519085， 3. ， University， Changsha 41oo82， China）

Abstract： Simultaneous localization mapping （SLAM） technology has a wide range applications in unmanned equipment， which can achieve indoor or outdoor autonomous localization mapping tasks. In this paper， the basic framework vision SLAM laser SLAM is introduced， the functions algorithms front-end odometer，back-end optimization，loop detection map construction are described indetail. After this，theclasic algorithms in the development visual/laser SLAMare summarized their advantages disadvantages areanalyzedas wellas the excelent improvement schemes after this.Inaddition，the typical application scenarios SLAM technology in life are listed， the important role SLAM technology in the fields automatic driving unmanned equipment is shown. Finally， the curent development trends research progress SLAM systems are discussed，as well as challenges issues to be considered in future applications， including multi-type sensor fusion， integration with deep learning technologies， the key role interdisciplinary collaboration. Through the comprehensive analysis discussion SLAM technology， it provides pround theoretical guidance practical reference for further promoting the development application SLAM technology.

Keywords：SLAM; camera; lidar;multi-sensor fusion

0 引言

近些年来随着人工智能的迅速发展，人们对便捷的生活以及智能化生产的需求越来越高，在这种需求的驱使下，机器人行业快速发展，而移动机器人是众多机器人种类中的一种，是许多企业与高校的研究热点[1]。（剩余32860字）