控制与信息技术Issue(4)：59-66,8.DOI:10.13889/j.issn.2096-5427.2025.04.009
一种降低漏检率的视觉与毫米波雷达感知融合方法
A Vision and Millimeter-Wave Radar Perception Fusion Method to Reduce False Negative Rate
周艳辉 1彭之川 1朱田 1王全 1朱泽敏 1刘光伟1
作者信息
- 1. 长沙中车智驭新能源科技有限公司,湖南 长沙 410000
- 折叠
摘要
Abstract
In this paper,a vision and millimeter-wave radar perception fusion method that integrates a vehicle data tracking,weighting,and scoring mechanism is proposed,to address the detection challenges of important target omission or errors in target fusion scenarios between vision and millimeter-wave radar during intelligent driving.Initially,a dynamic matching box designed as a criterion functions to match vision and radar targets.Then,the matched targets are converted to tracked targets through position and speed weighting,which are scored based on vehicle operating conditions,operating areas,and hazard conditions.Furthermore,perception fusion targets are output,following the filtering of the tracked targets based on their scores and estimated paths.Tests based on real vehicles verify that the proposed method improves stability in target perception compared to the traditional fusion method based on Kalman Filter and reduces the false negative rate in the detection of fusion targets.关键词
智能驾驶/视觉检测/毫米波雷达/感知/融合Key words
intelligent driving/visual inspection/millimeter wave radar/perception/fusion分类
信息技术与安全科学引用本文复制引用
周艳辉,彭之川,朱田,王全,朱泽敏,刘光伟..一种降低漏检率的视觉与毫米波雷达感知融合方法[J].控制与信息技术,2025,(4):59-66,8.
