基于4D Radar增益的Lidar目标检测研究  

Research on Lidar Target Detection Based on 4D Radar Gain

作　　者：梁晓阳 赵巍[1] 张以成 张宇[1] LIANG Xiaoyang;ZHAO Wei;ZHANG Yicheng;ZHANG Yu(School of Mechanical Engineering,Tianjin University of Technology and Education,Tianjin 300222,China)

机构地区：[1]天津职业技术师范大学机械工程学院,天津300222

出　　处：《自动化与仪表》2025年第3期96-101,共6页Automation & Instrumentation

摘　　要：针对传统Lidar目标检测模型对于雨雪等天气下的低鲁棒性和对遮挡及多路径反射的难以识别等问题,该研究提出了一种新颖的数据融合框架,该研究聚焦于将4D毫米波雷达数据整合到现有的Lidar目标检测模型中,旨在提高目标检测的准确性和鲁棒性。通过将4D毫米波雷达独有的多普勒和速度特征与在恶劣天气下的鲁棒性与传统Lidar相结合,提升了传统Lidar在各种场景下对目标的适应性和可靠性。此外,通过定量和定性分析验证了整合4D雷达数据后检测模型在多个评估指标上的显著提升。通过消融实验分析了RCS和速度等特征对不同目标检测性能的影响。This study addresses the low robustness of traditional Lidar target detection models under adverse weather conditions such as rain and snow,as well as difficulties in recognizing occlusions and multipath reflections.Propose a novel data fusion framework that focuses on integrating 4D millimeter-wave radar data into existing Lidar target de-tection models,aiming to improve the accuracy and robustness of target detection.By combining the unique Doppler and velocity features of 4D millimeter-wave radar with the robustness of traditional Lidar in harsh weather,enhance the adaptability and reliability of traditional Lidar in various scenarios.Furthermore,quantitative and qualitative anal-yses validate the significant improvement of the detection model in multiple evaluation metrics after integrating 4D radar data.Ablation experiments analyze the impact of features such as RCs and velocity on the performance of dif-ferent target detection scenarios.

关 键 词：4D Radar 目标检测 LIDAR 传感器 

分 类 号：TP212.9[自动化与计算机技术—检测技术与自动化装置]