激光雷达与RGB-D相机融合的区域图像算法SLAMOA
Region-based Image Algorithm SLAM with Fusion of LiDAR and RGB-D Camera
为高效辅助救援团队在火灾后迅速定位被困人员,本文使用搭载树莓派 4B 控制器、激光雷达、RGB-D 相机、IMU 传感器和轮式里程计的移动机器人来构建火灾后现场室内环境的高精度地图.针对单线激光雷达数据密度较低的问题,对 Gmapping 算法的建议分布进行重新设计,并优化其重采样方法.此外,提出了一种区域图像算法,将二维激光雷达数据与深度相机数据融合.仿真结果表明,改进后的 Gmapping 算法在粒子收敛速度上有显著提升,融合算法在建图精度上优于单激光雷达和单 RGB-D 相机建图.
In order to efficiently assist the rescue team in quickly locating trapped people after the fire disaster,this paper uses a mobile robot equipped with Raspberry Pi 4B controller,lidar,RGB-D camera,IMU(Inertial measurement unit)sensor and wheeled odometer to construct a high-precision map of the indoor environment after the fire.Aiming at the problem of low data density of single-line lidar,the suggested distribution of Gmapping algorithm is redesigned,and its resampling method is optimized.Furthermore,a region image algorithm is proposed to fuse 2D lidar data with depth camera data.Simulation results indicate that the improved Gmapping algorithm significantly enhances particle convergence speed,and the fusion algorithm outperforms single LiDAR and single RGB-D camera mapping in terms of mapping accuracy.
罗舒涵;汤奥斐;王琪龙;王宁;弓泽晨;师碧波
西安理工大学 机械与精密仪器工程学院,西安 710048西安理工大学 机械与精密仪器工程学院,西安 710048西安理工大学 机械与精密仪器工程学院,西安 710048西安理工大学 机械与精密仪器工程学院,西安 710048西安理工大学 机械与精密仪器工程学院,西安 710048西安理工大学 机械与精密仪器工程学院,西安 710048
信息技术与安全科学
移动机器人多传感器融合地图构建激光雷达RGB-D相机
mobile robotmulti-sensor fusionmap constructionLiDARRGB-D cameras
《机械科学与技术》 2026 (8)
1307-1314,8
评论