Integrating Loam-Based Lidar Odometry, Slope Detection, and 3D Reconstruction for Robust Slam Fusion

Reliable Light Detection and Ranging (LiDAR)-based odometry and terrain perception are essential for autonomous systems operating in complex outdoor environments. However, conventional LiDAR odometry methods may suffer from reduced robustness when facing sparse point clouds, uneven terrain, or challenging geometric structures. In this paper, we present an improved LiDAR-Inertial Odometry (LIO) system based on geometric modeling and curvature analysis. The proposed system enhances feature extraction and point cloud matching by incorporating curvature-based measurements, and integrates LiDAR and inertial information through factor graph optimization to achieve stable and accurate state estimation. In addition, a slope fitting strategy is employed to support terrain analysis under different observation distances. Extensive experiments are conducted to evaluate the performance of the proposed system in terms of accuracy, robustness, and computational efficiency. The results demonstrate that the proposed approach achieves consistent improvements over representative LiDAR odometry baselines, while maintaining realtime performance, indicating its effectiveness for terrain perception and autonomous navigation applications.