Abstract: Aiming at the demands of automatic classification of vehicles, by introducing the description method in spatial morphology of Geography Information System(GIS), this paper presents geometric parameters, calculation methods of the side view of the vehicles and geometric feature parameters, and combines feature geometric parameters for vehicle-type classification. With the geometrical features of the side view, it gives a method to extract vehicle axle parameters automatically, and designs a vehicle classifier based on geometric feature parameters. This method is proved to be effective by experiment.

Key words: video image, vehicle geometric parameters, vehicle-type classification, axle parameters

摘要： 针对车辆自动分类需求，通过引入地理信息系统空间形态的描述方法，给出车辆侧面图像的几何信息描述参数和计算方法，设计用于车辆分类的几何特征参数及其组合参数，基于侧视轮廓的几何特征，提出车辆轮轴参数的自动提取方法，设计一个基于几何形态参数的车型自动分类器。实验证明该方案的有效性。

关键词: 视频图像, 车辆几何参数, 车型分类, 轮轴参数

CLC Number: 

• TP391.41