声频特征在车型自动分类识别中的应用研究

Applied Research Based on Vehicle Acoustic Feature in Automatic Vehicle Classification and Recognition

为满足智能化交通的需要,利用车辆声频信号特征进行车型自动识别,对车辆声频信号进行建模,以线性预测编码倒谱(LPCC)为特征,采用聚类分析算法划分车型,并使用Fisher判别法对车型进行识别。实验结果表明此方法可准确和有效的实现车型的自动分类与识别。

In order to meet the intelligent transportation requirements, vehicle acoustic feature was used to achieve the automatic vehicle classification and recognition. The vehicle acoustic signal is modeled and the linear predictive coding cepstrum is involved as a feature parameter. The study introduce cluster analysis and discriminant analysis to achieve the automatic vehicle classification and recognition. Data test proves that the method achieve the accurate and effective automatic vehicle classification and recognition.