视觉和雷达融合的桥梁防撞预警检测系统
刘彬 1王焕然1
作者信息
- 1. 宁波大学冲击与安全工程教育部重点实验室 浙江宁波 315211
- 折叠
摘要
为了获得更加精确、全面、实时的船舶航行信息,有效防止船舶偏离正常航道对桥梁产生碰撞,造成生命及财产安全损害,构建一种视觉和雷达融合的桥梁防撞预警检测系统.以YOLOv5算法为基础,对比不同船舶数据集和不同迭代次数下的船舶检测置信度结果,并且得到船舶类型、大小等信息.通过相控阵雷达,检测船舶的航行及距离信息.在桥区设置三个预警区域,当船舶航行进入预警区域时生成报警提示,并对不同类型的船舶进行声光报警或触发智能拦截装置,进而有效保障桥梁和船舶航行安全.
Abstract
In order to obtain more accurate,comprehensive,and real-time ship navigation information,effectively prevent ship from deviating from normal navigation route,colliding with bridge,and causing damage to life and property safety,a bridge collision warning and detection system that integrates vision and radar was constructed.Based on the YOLOv5 algorithm,The ship detection confidence results under different ship datasets and iteration times were compared,and the information such as ship type and size was obtained.The navigation and distance information of ship were detected through phased array radar.Three warning areas in the bridge area were set up,the alarm prompt was generated when ship entered the warning area,and the sound and light alarms were provided or the intelligent interception device was triggered for different types of ships,effectively ensuring the safety of bridge and ship navigation.
关键词
视觉/雷达/桥梁/防撞/检测Key words
Vision/Radar/Bridge/Anticollision/Detection引用本文复制引用
出版年
2025
NETL
NSTL
万方数据