基于形态滤波的车辆检测算法研究

Research on Vehicle Detection Algorithm Based on Morphological Filtering

近年来,越来越多的智能停车管理系统采用无线磁阻传感器网络技术实现车辆检测。但该技术面临磁干扰问题,在干扰环境下,已有的车辆检测算法精度急剧下降。提出一种基于形态滤波的车辆检测算法,采用级联形态开-闭和形态闭-开运算设计滤波器,可有效过滤噪声干扰,且不会造成车辆磁信号的偏倚现象。实验结果表明:与已有车辆检测算法相比,本文算法提高了车辆检测准确率,检测精度达到98.5%。

In recent years,more and more intelligent parking management systems use wireless magnetic sensor network technology to realize vehicle detection.However,this technology is facing the problem of magnetic interference.Under the interference environment,the vehicle detection accuracy of the existing algorithm drops sharply.This paper proposes a vehicle detection algorithm based on morphological filtering,which uses cascading morphological open-close and close-off operations to design filters.This filter can effectively filter noise without causing bias of the vehicle magnetic signal.The experimental results show that,compared with the existing vehicle detection algorithms,the algorithm in this paper improves the vehicle detection accuracy rate,and the detection accuracy reaches 98.5%.