基于图像处理的车辆类型识别的研究  

Research on vehicle type recognition based on image processing

作　　者：朱金荣[1] 张广杰[1] 夏长权[1] ZHU Jinrong;ZHANG Guangjie;XIA Changquan(College of Physical Science and Technology,Yangzhou University,Yangzhou 225002,China)

机构地区：[1]扬州大学物理科学与技术学院,江苏扬州225002

出　　处：《现代电子技术》2017年第4期16-18,23,共4页Modern Electronics Technique

基　　金：国家自然科学基金(11304271);江苏省基础研究计划自然科学基金项目(BK20130434)

摘　　要：为了完善车型检测识别的算法,首先利用傅里叶描述子的方法建立车辆图形库,对摄像头采集到的车辆侧面图像进行形态学操作,得到车辆轮廓封闭曲线并与图形库模板匹配,确定车型;然后利用基于积分的方法处理车辆顶面图像,计算出相应特征参数,确定车辆大小信息,解决车型检测的多检、错检现象。该文对图像处理算法做了具体推导并给出了实现流程,通过仿真实验表明,该文的算法能够充分地提取车辆的形状与几何尺寸信息,车型识别正确率很高,且计算量小。In order to optimize the algorithm of vehicle type detection,the graphic base of vehicles was established with the method of Fourier descriptor. The vehicle side images acquired by camera are treated by means of morphological operation to get the closed curves of vehicle contour and match them with the templates in the graphic base to determine the vehicle type.The vehicle top images are processed with the integral method to calculate the corresponding characteristic parameters to ensure the size information of a vehicle,so as to eliminate the phenomenon of the wrong detection. The image processing algorithm is derived and its implementation flow process is given in this paper. The simulation experiment result shows that the algorithm is able to adequately extract the information of vehicle shape and size,the recognition accuracy is high,and the calculated amount small.

关 键 词：智能交通 车型检测 傅里叶描述子 AOI区域划分 

分 类 号：TN911.73-34[电子电信—通信与信息系统]