摘要
为实现全天候的高精度识别,提升山区公路交通的安全性,文中将目标检测与机器视觉技术相融合,设计了一种道路危险隐患识别与预警算法。该算法利用目标检测技术对各类路面缺陷和安全隐患进行动态多目标采集与识别,不仅提升了目标采集的效率,还实现了算法的可集成化和轻量化。通过在监控平台中引入机器视觉技术,有效提升了识别算法的抗扰性,使其在复杂气象干扰下仍能对特殊目标进行高精度识别。对所提算法进行的对比测试结果表明,其危险隐患的识别精准率可达96.6%,能够为山区公路交通安全提供有力的保障。
In order to achieve high⁃precision recognition 24/7 and improve the safety of mountain road traffic,this paper integrates object detection and machine vision technology to design a road hazard recognition and warning algorithm.This algorithm utilizes object detection technology to dynamically collect and identify various road surface defects and safety hazards,which not only improves the efficiency of object collection,but also achieves the algorithm’s integrability and lightweighting.By introducing machine vision technology into the monitoring platform,the anti⁃interference ability of the recognition algorithm has been effectively improved,enabling high⁃precision recognition of special targets even under complex meteorological interference.The comparative test results of the proposed algorithm show that its accuracy in identifying hidden dangers can reach 96.6%,which can provide strong guarantee for traffic safety on mountainous roads.
作者
徐振华
XU Zhenhua(School of Traffic&Transportation,Chongqing Jiaotong University,Chongqing 400064,China)
出处
《电子设计工程》
2025年第24期98-102,共5页
Electronic Design Engineering
基金
重庆市教委科学技术研究项目(KJQN202005201)。
关键词
危险隐患
山区公路
目标检测
机器视觉技术
识别精准率
dangerous hazards
mountain roads
object detection
machine vision technology
identific⁃ation accuracy
