POLE-SHAPED OBJECT DETECTION USING MOBILE LIDAR DATA IN RURAL ROAD ENVIRONMENTS

Abstract. Pole-shaped objects (PSOs) located along a road play key role in road safety and planning. Automation is required for calculating the numbers of trees need to be removed and utility poles need to be relocated during rural road widening. Road-side poles are among the most frequently struck road-side objects during road-side accidents. An automatic method is therefore proposed for detecting PSOs using LiDAR point cloud captured along the roadway using Mobile LiDAR system. The proposed method is tested on the point cloud data of rural road environment in India. Dataset of study area having text file size of 1.22 GB is processed in 13 minutes resulting in completeness of 88.63 % and correctness of 95.12 % in identifying PSOs within 10m of the road boundary. In data of across road coverage of 5m of the road boundary, the completeness of 93.10 % and correctness of 100% are achieved. Poles attached with other objects, tilted poles and the poles occluded by tree branches and shrubs are detected by the proposed method.