A Beam Guardrail Detection Algorithm Using Lidar for Intelligent Vehicle

摘要

In the paper, a beam guardrail detection algorithm using Lidar for intelligent vehicle is proposed. The proposed algorithm consists of feature extraction and clustering steps. The layered feature of Lidar scanning data is used to segment the point cloud into lines. Then, two novel features, which are corner feature and height feature, are developed to extract the beam guardrail points. Therefore, a clustering algorithm is used to remove some noise extracted points. To evaluate the performance of the proposed algorithm, a performance measure is defined. Computer simulations demonstrate that the proposed detection algorithm has improved performance compared to conventional methods and the proposed performance measure is meaningful and explicit.