计算机科学与探索

• 学术研究 •    

激光点云的三维目标检测研究进展

周燕, 蒲磊, 林良熙, 刘翔宇, 曾凡智, 周月霞   

  1. 佛山科学技术学院 计算机系, 广东 佛山 528000

Research Progress on 3D Object Detection of LiDAR Point Cloud

ZHOU Yan, PU Lei, LIN Liangxi, LIU Xiangyu, ZENG Fanzhi, ZHOU Yuexia   

  1. Department of Computer Science, Foshan University, Foshan, Guangdong 528000, China

摘要: 三维目标检测是近年来新兴的研究方向,其主要任务是对空间中的目标进行定位与识别。现有使用单目或双目视觉的方法来完成三维目标检测任务,容易受物体遮挡、视点变化和尺度变化的影响,导致检测精度不佳及鲁棒性差等问题。由于激光点云能描述三维场景的信息,在激光点云数据的基础上使用深度学习的方法完成三维目标检测任务,已成为三维视觉领域中研究的热点。针对激光点云的三维目标检测,梳理了近年来相关的研究工作。首先根据输入网络的数据形式将基于激光点云的三维目标检测方法分为基于原始点云、基于点云投影、基于点云体素化及基于多模态融合的三维目标检测方法,并对各类最具有代表性的方法进行了详细阐述。然后介绍了当前常用的开源数据集及其评价指标,并在数据集上对各类方法进行了性能对比,从多个方面讨论了各类方法的优势及局限性。最后指出当前激光点云的三维目标检测研究存在的不足和难点,并对其未来的发展趋势进行了总结与展望。

关键词: 三维目标检测, 激光点云, 深度学习

Abstract: 3D object detection is a new research direction in recent years and its main task is the location and recognization of targets in space. The existing methods for 3D object detection using monocular or binocular stereo vision are easily affected by object occlusion, viewpoint changing and scale changing in 3D scene, there will be problems such as poor detection accuracy and robustness. LiDAR point cloud can provide 3D scene information, so using deep learning method to complete 3D object detection based on LiDAR point cloud has become a research hotspot in the field of 3D vision. Aiming at the 3D object detection based on LiDAR point cloud, the relevant research in recent years is reviewed. Firstly, the 3D object detection methods based on LiDAR point cloud are divided into point cloud based, view based, voxel based and multi-modal fusion based 3D object detection methods according to the form of network input, and the most representative methods in each category are described in detail. Then common dataset are introduced, and performance of representative methods are evaluated, and the advantages and limitations of each method are discussed from several aspects. Finally, the shortcomings and difficulties are given. It also discusses and looks forward the future development direction.

Key words: 3D object detection, LiDAR Point Cloud, Deep Learning