A LIDAR SLAM Based on Point-Line Features for Underground Mining Vehicle

This paper presents an original 2D LIDAR SLAM based on point and line features to the underground tunnel environment. In our algorithm, the point features are extracted by using Curvature detector with range information weights. The line features detection algorithm contains segmentation by two hierarchical levels and line fitting. We present the approach used in our prototype of underground mining truck which achieves real-time localization and mapping at a 5cm resolution. We provide experimental results and comparisons to other approaches. The results show that this proposed method has good robustness and accuracy.