基于PCA-LDA-SVM的多普勒雷达车型识别算法  

Vehicle Recognition Algorithm with Doppler Radar Based on PCA-LDA-SVM

作　　者：方菲菲[1,2] 余稳[1,3] 

机构地区：[1]中国科学院上海微系统与信息技术研究所,上海200050 [2]中国科学院研究生院,北京100039 [3]上海慧昌智能交通系统有限公司,上海200233

出　　处：《数据采集与处理》2012年第1期111-116,共6页Journal of Data Acquisition and Processing

基　　金：国家高新技术发展计划("八六三"计划)基金(2008AA11Z203)资助项目

摘　　要：车辆检测和车型识别是智能交通系统(Intelligent transportation system,ITS)中的一个重要方面,而目标识别是低分辨率雷达领域的一个难点。该文提出一种用多普勒雷达进行车型识别的方法,把车辆建模成包含多个散射中心的目标体,散射中心与雷达的距离与频谱能量有关,因此同一目标的频谱变化反映了该目标长高等轮廓特征。然后将有效的频谱特征结合主成分分析(Principal component and analysis,PCA)和线性判别分析(Linear discriminant analysis,LDA)进行降维,再利用支持向量机(Support vector machine,SVM)等分类器实现分型。文章对不同识别算法交叉验证的实验结果进行比较,表明基于PCA-LDA-SVM的车型识别算法效果理想,有广泛的应用前景。Vehicle detection and recognition is important to the development of intelligent transportation system（ITS）, but target recognition is a challenging problem for low-resolution radar. Hence, a vehicle recognition approach using Doppler radar is proposed, and the spec- trum variation of one vehicle reflects its outline. Then, the dimension of effective spectrum fea- ture can be reduced by the methods of principal component analysis （PCA） and linear discrimi- nant analysis （LDA）. Vehicles can be classified into three types by classifier algorithms such as support vector machine（SVM）, K-nearest neighbor（KNN）. Finally, experimental results of different algorithms are compared by cross validation, and it shows that the algorithm based on PCA-LDA-SVM can achieve an ideal result.

关 键 词：雷达目标识别 多普勒雷达 主成分分析 线性判别分析 支持向量机 

分 类 号：TN958.95[电子电信—信号与信息处理]