Research on Laser-Vision-Inertial Multi-Sensor Tight Coupling SLAM Method for Mines
Abstract：The complex geological conditions and lighting changes in mine environments pose challenges for robot localization and mapping. Traditional single-sensor SLAM (laser SLAM or visual SLAM) has limitations in such environments. This paper proposes a laser-vision-inertial multi-sensor tight coupling SLAM method, which combines the advantages of LiDAR, visual camera, and IMU to solve the problem of tunnel structure laser degradation and visual tracking failure in dark environments. Through external parameter calibration, multi-sensor data alignment is achieved, and an adaptive visual weight mechanism is designed to dynamically assess the reliability of laser and visual data. Experiments show that this method reduces positioning error to 1.2% (reducing by more than 30% compared to single-sensor SLAM) and improves mapping accuracy by 25%. Especially in dark scenarios, it still maintains stable positioning performance, enhancing the system's adaptability and robustness.
Machine-generated Keywords：
Publication Date：2025-11-30
Online Publishing Date：2025-12-10(First online date of this platform, not the publication date of the document)
Pages：5( 113-117 )