计算机应用 ›› 2020, Vol. 40 ›› Issue (8): 2428-2433.DOI: 10.11772/j.issn.1001-9081.2019122227

• 应用前沿、交叉与综合 • 上一篇    下一篇

基于深度学习的道路障碍物检测方法

彭育辉, 郑玮鸿, 张剑锋   

  1. 福州大学 机械工程及自动化学院, 福州 350116
  • 收稿日期:2020-01-05 修回日期:2020-02-25 出版日期:2020-08-10 发布日期:2020-05-13
  • 通讯作者: 彭育辉(1975-),男,福建莆田人,教授,博士,主要研究方向:汽车无人驾驶、计算机辅助图形图像,pengyuhui@fzu.edu.cn
  • 作者简介:郑玮鸿(1994-),男,福建莆田人,硕士研究生,主要研究方向:点云数据处理、深度学习、三维目标检测;张剑锋(1995-),男,福建泉州人,硕士研究生,主要研究方向:同时定位与地图构建、点云数据处理。
  • 基金资助:
    福建省科技厅产学合作重大项目(2017H6007)。

Deep learning-based on-road obstacle detection method

PENG Yuhui, ZHENG Weihong, ZHANG Jianfeng   

  1. School of Mechanical Engineering and Automation, Fuzhou University, Fuzhou Fujian 350116, China
  • Received:2020-01-05 Revised:2020-02-25 Online:2020-08-10 Published:2020-05-13
  • Supported by:
    This work is partially supported by Fujian Science and Technology Department Major Program of Industry-University Collaboration (2017H6007).

摘要: 针对基于激光雷达(LiDAR)的三维点云数据处理及道路障碍物检测的问题,提出一种基于深度学习的路障碍物检测方法。首先,采用统计滤波算法对原始点云进行离群点的剔除处理;其次,提出一种端到端的深度神经网络VNMax,利用最大池化对区域候选网络(RPN)架构进行优化,构建改进的目标检测层;最后,在KITTI数据集上进行了训练及测试实验。结果显示,经过滤波处理,点云中各点之间的平均距离得到有效减少。通过对在KITTI数据集的简单、中等和困难任务的车辆定位处理结果比较得出,所提方法的平均精度比VoxelNet(Unofficial)分别提高了11.3个百分点、6.02个百分点和3.89个百分点。实验测试结果表明,统计滤波算法仍是有效的三维点云数据处理手段,最大池化模块可以提高深度神经网络的学习性能和目标定位能力。

关键词: 无人驾驶, 深度学习, 激光雷达, 目标检测, 三维点云

Abstract: Concerning the problems of 3D point cloud data processing and on-road obstacle detection based on Light Detection And Ranging (LiDAR), a deep learning-based on-road obstacle detection method was proposed. First, the statistical filtering algorithm was applied to eliminate the outliers from the original point cloud, improving the roughness of point clouds. Then, an end-to-end deep neural network named VNMax was proposed, the max pooling was used to optimize the structure of Region Proposal Network (RPN), and an improved target detection layer was built. Finally, training and testing experiments were performed on KITTI dataset. The results show that, by filtering, the average distance between the points in point cloud is reduced effectively. For the car location processing results of easy, medium difficult and hard detection tasks in KITTI dataset, it can be seen that the average precisions of the proposed method are improved by 11.30 percentage points, 6.02 percentage points and 3.89 percentage points, respectively, compared with those of the VoxelNet. Experimental results show that the statistical filtering algorithm is still an effective 3D point cloud data processing method, and the max pooling module can improve the learning performance and object location ability of the deep neural network.

Key words: autonomous driving, deep learning, Light Detection And Ranging (LiDAR), object detection, 3D-point cloud

中图分类号: