《电子技术应用》
您所在的位置:首页 > 人工智能 > 设计应用 > 基于点云补全的三维目标检测
基于点云补全的三维目标检测
2023年电子技术应用第8期
陈辉,王帅杰,蔡晗
(桂林电子科技大学 信息与通信学院, 广西 桂林 541004)
摘要: LiDAR技术的发展为自动驾驶提供了丰富的3D数据。然而,由于遮挡和某些反射材料的原因引起信号丢失,LiDAR点云实际上是不完整的2.5D数据,这对 3D 感知提出了根本性挑战。针对这一问题,提出对原始数据进行三维补全的方法。根据大多数物体形状对称且重复率高的特点,通过学习先验对象形状的方法估计点云中遮挡部分的完整形状。该方法首先识别被遮挡和信号缺失影响的区域,在这些区域中预测区域所包含对象形状的占用概率。针对物体间遮挡的情况,通过形状的占用概率和共享同类形状形态进行三维补全。对自身遮挡的物体,通过自身镜像进行恢复。最后通过点云目标检测网络进行学习。结果表明,通过该方法能有效地提高生成点云3D边框的mAP(mean Average Precision)。
中图分类号:TP389.1
文献标志码:A
DOI: 10.16157/j.issn.0258-7998.223624
中文引用格式: 陈辉,王帅杰,蔡晗. 基于点云补全的三维目标检测[J]. 电子技术应用,2023,49(8):1-6.
英文引用格式: Chen Hui,Wang Shuaijie,Cai Han. 3D object detection based on point cloud completion[J]. Application of Electronic Technique,2023,49(8):1-6.
3D object detection based on point cloud completion
Chen Hui,Wang Shuaijie,Cai Han
(School of lnformation and Communication, Guilin University of Electronic Technology, Guilin 541004, China)
Abstract: The development of LiDAR technology provides abundant 3D data for autonomous driving. However, LIDAR point cloud is actually incomplete 2.5D data due to signal loss caused by occlusion and some reflective materials, which poses a fundamental challenge to 3D perception. To solve this problem, this paper proposes a method for 3D completion of the original data. According to the symmetric shape and high repetition rate of most objects, the complete shape of the occluded part in the point cloud is estimated by learning the prior object shape. The method first identifies regions affected by occlusions and signal loss, and in these regions, predicts the occupancy probability of the shapes of objects contained in the regions. For the case of occlusion between objects, 3D completion is performed through the occupancy probability of the shape and the morphologies that share the same shape. The objects occluded by themselves are restored by mirroring themselves. Finally, it is learned through the point cloud target detection network. The results show that this method can effectively improve the mAP for generating point cloud 3D borders.
Key words : LiDAR;point cloud;3D completion;target detection

0 引言

3D目标检测作为自动驾驶感知系统的核心基础之一,可以广泛应用于路径规划、运动预测、碰撞避免等功能。通常,带有相应3D激光雷传感器的汽车已经成自动驾驶领域的标准配置,由此能够提供准确的深度信息,点云数据的处理也越来越普遍、越来越重要。尽管已有很多进展,但由于点云本质上的高度稀疏性和不规则的特性,使得传统的卷积神经网络无法对点云数据进行准确的学习,而且由于相机视图和激光雷达鸟瞰视图之间的不对齐而导致的导致模态协同和远距离尺度变化等原因,三维点云的处理远比二维图像要难得多。因此,在三维点云上的目标检测目前仍处于初级阶段。



本文详细内容请下载:https://www.chinaaet.com/resource/share/2000005474




作者信息:

陈辉,王帅杰,蔡晗

(桂林电子科技大学 信息与通信学院, 广西 桂林 541004)

微信图片_20210517164139.jpg

此内容为AET网站原创,未经授权禁止转载。