Visual-Inertial SLAM Algorithm for Low-Texture Subterranean Environments

作者：Hao Wang, Qian Sun, Junjing Zou, Wa Liu · 年份：2023 · DOI：10.1109/icmmt58241.2023.10277214 · 被引用次数：4 · 研究领域：Robotics and Sensor-Based Localization、3D Surveying and Cultural Heritage、Robotic Path Planning Algorithms

In order to solve the problem of poor performance of traditional visual-inertial SLAM algorithms in low-texture subterranean environments, a SLAM algorithm for low-texture subterranean environments was proposed. The algorithm adopts the visual-inertial SLAM scheme, and added the representation, detection, and matching of line features based on point features. The algorithm increased the available visual features, enhanced the robustness of the visual-inertial SLAM scheme, and solved the problem of positioning accuracy degradation caused by insufficient features. Experiments on the EuRoC and DARPA SubT datasets show that the proposed algorithm has stronger robustness and accuracy compared with VINS-Fusion in low-texture subterranean environments.