江苏大学学报（自然科学版）2026，Vol.47Issue(2)：125-133,9.DOI:10.3969/j.issn.1671-7775.2026.02.001
基于改进PointPillars的自动驾驶障碍物点云检测算法
Obstacle point cloud detection algorithm for automatic driving based on improved PointPillars
摘要
Abstract
To solve the problems of high false detection rate of interferential point clouds and high missing detection rate of distant sparse point clouds in automatic driving scene,the obstacle point cloud detection algorithm based on improved PointPillars was proposed.The point cloud in pillars was encoded by the aggregation module and shared multi-layer perceptron(MLP).The salient and detailed features were mapped into pseudo-image features by stacking the max-pooling and average-pooling.To solve the problem of pseudo-image feature with insufficient attention and utilization,the deep and shallow feature maps were fused by attention and residual second block(ARSB)module to optimize the gradient and enhance the coordinate attention(CA)to effective targets.The results show that the improved algorithm has high detection accuracy for global point clouds.The detection precision of the improved algorithm is better than those of the classical 3D detection algorithms of PointPillars and STD methods,especially for the detection of car category.The detection speed is fast,which meets the requirements of real-time.关键词
障碍物点云/深度学习/点云目标检测/点云柱体编码/伪图特征提取模块Key words
obstacle point cloud/deep learning/point cloud detection/point pillar encode/attention and residual second block分类
信息技术与安全科学引用本文复制引用
沈跃,沈卓凡,刘慧,周昊,曾潇..基于改进PointPillars的自动驾驶障碍物点云检测算法[J].江苏大学学报（自然科学版）,2026,47(2):125-133,9.基金项目
国家自然科学基金资助项目(32171908) （32171908）
