视觉激光融合SLAM分类简介——V-LOAM、VELO、LVI-SAM、LIMO

slamer入坑

已于 2023-11-13 21:47:44 修改

阅读量900

点赞数 5

CC 4.0 BY-SA版权

文章标签： github 机器人无人机车载系统计算机视觉开源

于 2023-11-13 21:14:54 首次发布

本文链接：https://blog.csdn.net/m0_58605508/article/details/134371399

本文介绍了VLOAM和LIMO两种视觉和激光结合的SLAM算法，VLOAM用于激光SLAM，ICRA2015提出但无开源代码；LIMO侧重视觉SLAM，2018年IROS发布并开源。文章还列举了实验设备和相关论文源。

摘要生成于 C知道，由 DeepSeek-R1 满血版支持，前往体验 >

VLOAM和LIMO算法就是属于视觉和激光结合的算法，其中VLOAM算法更偏向于是一种激光SLAM算法，而LIMO算法更偏向于是一种视觉SLAM算法

实验设备和论文来源:

VLOAM算法是在2015年ICRA上提出的，原论文名为《Visual-lidar Odometry and Mapping: Low-drift, Robust, and Fast》，作者并没有对代码进行开源，复现的版本有两个，分别是YukunXia/VLOAM-CMU-16833和Jinqiang/demo_lidar

The camera is a uEye monochrome camera configured at 60Hz frame rate. The 3D lidar is based on a Hokuyo UTM-30LX laser scanner. The laser scanner has 180° field of view and 0.25° resolution with 40 lines/sec scanning rate. A motor actuates the laser scanner for rotational motion to realize 3D scan. The motor is controlled to rotate at 180° / s angular speed between −90° and 90° with the horizontal orientation of the laser scanner as zero. An encoder measures the motor rotation angle with 0.25° resolution.

LIMO算法2018年发表在IROS会议，原论文名为《LIMO: LiDAR-Monocular Visual Odometry》，该论文是有开源的johannes-graeter/limo

全球定位及惯性导航系统(GPS/IMU): OXTS RT 3003 ×1 (open sky localization errors < 5 cm)
3D 64线激光雷达: Velodyne HDL-64E ×1 (10 Hz, 64 laser beams,range: 100 m)
灰度摄像机: Point Grey Flea 2 (FL2-14S3M-C) ×2 (10 Hz,resolution: 1392×512 pixels, opening: 90◦ ×35◦),
彩色摄像机: Point Grey Flea 2 (FL2-14S3C-C) ×2 (10 Hz,resolution: 1392×512 pixels, opening: 90◦×35◦),
光学镜头(4-8 mm): Edmund Optics NT59-917 ×4

LVI-SAM算法2021年ICRA上提出的，原论文名为《LVI-SAM: Tightly-coupled Lidar-Visual-Inertial Odometry via Smoothing and Mapping》，该论文开源代码https://git.io/lvi-sam

Velodyne VLP-16 lidar

FLIR BFS-U3-04S2M-CS camera

MicroStrain 3DM-GX5-25 IMU

Reach RS+ GPS (for ground truth)