Light-SLAM: A Robust Deep-Learning Visual SLAM System Based on LightGlue under Challenging Lighting Conditions
This work addresses robustness issues in visual SLAM for autonomous systems like driving and robotics, but it is incremental as it combines existing deep learning and geometry-based approaches.
The authors tackled the problem of visual SLAM robustness in challenging lighting conditions by proposing a hybrid system based on LightGlue, which improved accuracy and robustness compared to traditional and deep learning methods, achieving real-time GPU performance.
Simultaneous Localization and Mapping (SLAM) has become a critical technology for intelligent transportation systems and autonomous robots and is widely used in autonomous driving. However, traditional manual feature-based methods in challenging lighting environments make it difficult to ensure robustness and accuracy. Some deep learning-based methods show potential but still have significant drawbacks. To address this problem, we propose a novel hybrid system for visual SLAM based on the LightGlue deep learning network. It uses deep local feature descriptors to replace traditional hand-crafted features and a more efficient and accurate deep network to achieve fast and precise feature matching. Thus, we use the robustness of deep learning to improve the whole system. We have combined traditional geometry-based approaches to introduce a complete visual SLAM system for monocular, binocular, and RGB-D sensors. We thoroughly tested the proposed system on four public datasets: KITTI, EuRoC, TUM, and 4Season, as well as on actual campus scenes. The experimental results show that the proposed method exhibits better accuracy and robustness in adapting to low-light and strongly light-varying environments than traditional manual features and deep learning-based methods. It can also run on GPU in real time.