基于朴素贝叶斯分类模型的车型识别方法

Naive-Bayesian-Classifier-Based Method for Classifying Vehicles

提出一种基于朴素贝叶斯分类模型的车辆分类方法,采用车辆的实际特征数据长度和宽度作为训练样本,离线训练朴素贝叶斯分类模型,同时利于CCD摄像机采集道路车辆图像,提取车辆轮廓曲线外接矩形的长度和宽度作为测试样本,通过离线训练获得的分类器,对车辆类型进行识别.仿真试验证明,朴素贝叶斯分类模型具有较高的分类性能,在同等训练和测试条件下,可以获得比BP神经网络分类器优越的分类效果.

This paper proposes a vehicle classifying method based on the Naive Bayesian classifier theory. Firstly, the actual vehicle characteristic data of length and width are used as training samples for NBClassifier off-line. The standard vehicle images are captured by CCD cameras, with the vehicle characteristie data of length and width obtained by using edge detecting algorithm as tested samples. Then the vehicle type can be judged by means of the trained NBClassifier according to the method proposed in ibis paper, Experiments show that it has higher accuracy compared with the BP network under the same test conditions.