越野环境下履带车辆的障碍识别与可通行性分析

To address the limitations of existing traversability analysis methods,which often suffer from incomplete obstacle recognition and poor generalization in complex off-road environments, this paper proposes an obstacle recognition method based on open-vocabulary semantic segmentation. The method extracts the semantic labels of obstacles and terrain around the vehicle,enabling effective identification of previously unseen obstacles in unstructured environments. The method is validated on the datasets in real-world experiments,demonstrating its stability and comprehensive recognition capability. On this basis,a multi-layer 2. 5D map is constructed by integrating semantic labels with 3D point clouds. The traversability level of a terrain is preliminarily classified according to semantic labels. And then the terrain smoothness is quantified based on ground elevation,and the geometric parameters of special environmental features (e. g.,vertical wall) are measured. Furthermore,the driving posture of vehicle is predicted by incorporating the geometric configuration of a tracked vehicle, thereby quantifying the coupling relationship among static slope stability, semantic terrain categories and geometric attributes. A cost function is then designed to jointly assess the traversal risk and cost of vehicle,ultimately generating a vehicle-centric traversability map. The effectiveness and reliability of the proposed method are verified by comparing it with similar methods,which enhances data support for decision-making,planning,and control of unmanned tracked platforms.