A Fast Spatial Clustering Method for Sparse LiDAR Point Clouds Using GPU Programming.

Jeonghoon Kwak,Yifei Tian,Wei Song,Su Sun,Long Chen,Yunsick Sung

doi:10.3390/s20082309

Abstract

Fast and accurate obstacle detection is essential for accurate perception of mobile vehicles’ environment. Because point clouds sensed by light detection and ranging (LiDAR) sensors are sparse and unstructured, traditional obstacle clustering on raw point clouds are inaccurate and time consuming. Thus, to achieve fast obstacle clustering in an unknown terrain, this paper proposes an elevation-reference connected component labeling (ER-CCL) algorithm using graphic processing unit (GPU) programing. LiDAR points are first projected onto a rasterized x–z plane so that sparse points are mapped into a series of regularly arranged small cells. Based on the height distribution of the LiDAR point, the ground cells are filtered out and a flag map is generated. Next, the ER-CCL algorithm is implemented on the label map generated from the flag map to mark individual clusters with unique labels. Finally, obstacle labeling results are inverse transformed from the x–z plane to 3D points to provide clustering results. For real-time 3D point cloud clustering, ER-CCL is accelerated by running it in parallel with the aid of GPU programming technology.

Highlights

IntroductionTo achieve fast obstacle clustering in an unknown terrain, this paper proposes an elevation-reference connected component labeling (ER-CCL) algorithm using graphic processing unit (GPU) programing
We propose an obstacle flag map generation method to filter out ground cells and identify non-ground cells on a 2D horizontal plane
We have proposed a graphic processing unit (GPU)-based fast spatial clustering method to label dispersed light detection and ranging (LiDAR) point clouds into individual groups

Summary

Introduction

To achieve fast obstacle clustering in an unknown terrain, this paper proposes an elevation-reference connected component labeling (ER-CCL) algorithm using graphic processing unit (GPU) programing. Real-time obstacle clustering algorithm research involves 3D point clouds sensed by a stereo camera and video sequences mounted on UGVs [8,9]. Central processing unit (CPU)-based computation methods always implement the neighboring points search process point by point, which is difficult to achieve as a real-time approach [17]. To solve these problems, a graphic processing unit (GPU)-based 3D obstacle labeling method is proposed to realize real-time obstacle clustering in LiDAR point clouds. From the non-ground obstacle cells, a GPU-based elevation-reference connected component labeling (ER-CCL)

Methods

Results

Conclusion