现代信息科技2025，Vol.9Issue(22)：194-198,5.DOI:10.19850/j.cnki.2096-4706.2025.22.036
基于激光雷达与视觉融合的机器人SLAM技术
Robot SLAM Technology Based on the Fusion of LiDAR and Vision
摘要
Abstract
In response to the limitations of a single sensing mode in complex application scenarios,this study,based on the multi-sensor fusion theoretical framework,integrates the high-precision spatial modeling advantages of LiDAR and the rich environmental feature extraction capabilities of visual sensors to construct a SLAM system for heterogeneous sensing data fusion.By comparing and analyzing the mapping effects of pure laser SLAM system and multi-sensor fusion system based on Karto and Gmapping algorithms,experimental data shows that the fusion system exhibits significant improvements in key indicators such as map integrity,environmental feature fidelity,and spatial topology accuracy.This hybrid SLAM architecture effectively solves the problem of mapping failure of a single sensor in texture-poor areas or dynamic scenes through a multimodal data complementarity mechanism,providing a centimeter-level positioning accuracy and high-reliability environmental cognition solution for robot autonomous navigation systems in complex environments.关键词
移动机器人/激光雷达/视觉传感器/SLAM技术Key words
mobile robot/LiDAR/visual sensor/SLAM technology分类
电子信息工程引用本文复制引用
杨红莉,梁远生,于浩,纪晓杏,曾宪阳..基于激光雷达与视觉融合的机器人SLAM技术[J].现代信息科技,2025,9(22):194-198,5.基金项目
国家自然科学基金项目(11701274) （11701274）
江苏省自然科学基金项目(BK20170760) （BK20170760）
江苏省研究生科研与实践创新计划项目(SJCX24_1297,SJCX24_1291) （SJCX24_1297,SJCX24_1291）
南京工程学院研究生教育教学改革课题(2024YJYJG26,2024YJYJG28) （2024YJYJG26,2024YJYJG28）
南京工程学院创新基金重大项目(CKJA202206) （CKJA202206）
