Research on Railway Obstacle Detection Method Based on Developed Euclidean Clustering

To prevent the problem of safety accidents caused by intrusion obstacles into railway clearance, this paper proposes an obstacle detection method based on Light Detection and Ranging (LiDAR) to obtain process rich three-dimensional (3D) information depth scene. The first preprocesses point cloud scenario collected LiDAR divide a basic area containing rails. Then, divides roadbed plane fits rails with random sample consensus (RANSAC) algorithm, dividing according position address issue over or under-segmentation in traditional Euclidean clustering method, which is due sparser clouds farther object from LiDAR, improves conventional clustering. It introduces adaptive distance threshold categorize obstacles. Finally, compared clustering, K-means density-based spatial applications noise (DBSCAN) improved cluster has achieved better results terms computing time segmentation accuracy. Experimental show ability detect successfully.