Vehicle Recognition Based on Multi-feature Extraction and SVM Parameter Optimization
CHENG Shu-hong1,2,GAO Xu1,ZHOU Bin1
1. College of Electrical Engineering, Yanshan University, Qinhuangdao, Hebei 066004, China
2. Postdoctoral Workstation of CITIC Dicastal Co., Ltd, Qinhuangdao, Hebei 066004, China
Abstract A kind of vehicle recognition method which was based on multi-feature extraction and support vector machines(SVM) parameter optimization is proposeal. Many kinds of problems that used the single-feature can be influenced by those factors such as light, weather and shadow, etc. Those problems could be solved by our method. In addition, our method can identify the moving vehicle model. At first, the samples of vehicle are collected and begin the process of image preprocessing, a variety of features will be extracted, including geometric features, texture features and histogram of gradient features. The second, combining and testing the various features, then the results with the results of single-feature testing are compared. At last, preparing for the recognition of the vehicle by SVM which was optimized by Particle Swarm Optimization(PSO). The experimental results show that the method which is put forward can achieve a good recognition results. The recognition rate can reach more than 90%.
［1］刘昕, 王殿海, 王新颖, 等. 基于IPv6的智能交通信息采集与处理方法［J］. 吉林大学学报(工学版), 2010, 40(5):1226-1229.
［2］Qin Y, Yuan B J, Pi S. Research on framework and key technologies of urban rail intelligent transportation system ［C］//Proceedings of the 2015 International Conference on Electrical and Information Technologies for Rail Transportation (2nd). Zhuzhou:Springer Berlin Heidelberg, 2016:729-736.
［3］Ludanek H. Ludanek on ICT and intelligent transportation systems［J］. Ericsson Review, 2016, 93(1): 30-39.
［4］Moral-Munoz J A, Cobo M J, Chiclana F, et al. Analyzing Highly Cited Papers in Intelligent Transportation Systems［J］.IEEE Transactions on Intelligent Transportation Systems, 2016, 17(4):993-1001.
［5］Zhao J, Liu Z J. Research on Vehicle Type Recognition based on Data surveying［J］.Science and Technology, 2016, 9(2):319-330.
［6］Wang Y, Zhou J L. Vehicle type recognition in WSN based on ITESP algorithm［C］//10th IEEE International Conference on Ubiquitous Intelligence and Computing (UIC 2013) and 10th IEEE International Conference on Autonomic and Trusted Computing (ATC 2013).Vietrisul Mare, Italy, 2013:668-671.
［7］Rao Y B. Automatic vehicle recognition in multiple cameras for video surveillance［J］. Visual Computer,2015, 31(3):271-280.
［8］王瑞, 王康晏, 冯玉田, 等. 复杂场景下声频传感器网络核稀疏表示车辆识别［J］. 西安电子科技大学学报(自然科学版), 2015, 42(4):115-120.
［9］Wang T T, Xiu C B, Cheng Y. Vehicle Recognition Based on Saliency Detection and Color Histogram［C］//27th Chinese Control and Decision Conference (CCDC 2015), Qingdao, 2015.
［10］Liu S M, Huang Y P, Zhang R J. On-Road Vehicle Recognition Using the Symmetry Property and Snake Models［J］. International Journal of Advanced Robotic Systems,2013,10:1-9.
［11］PriyadharshiniR A, Arivazhagan S, Sangeetha L. Vehicle Recognition based on Gabor and Log-Gabor Transforms［C］//2014 IEEE International Conference on Advanced Communication, Control and Computing Technologies(ICACCCT 2014). Tamil Nadu, India, 2015.
［12］Zhu G Y, Huang D, Zhang P, et al. E-Proximal support vector machine for binary classification and its application in vehicle recognition［J］. Neurocomputing, 2015, 161: 260-266.
［13］王恒奎, 边耐欣, 王文, 等. 基于Trimmed NURBS 曲面几何特征的数字化自适应采样［J］. 计量学报, 2002, 23(4):272-275.
［14］王崇文, 李见为, 林国清. 一种新的复合指纹匹配法［J］. 计量学报, 2004, 25(1):77-80.
［15］Li B, Huo G. Face recognition using locality sensitive histograms of oriented gradients［J］. Optics ,2016, 127(6):3489-3494.
［16］孟宗,季艳,谷伟明, 等. 基于支持向量机和窗函数的DEMD端点效应抑制方法［J］. 计量学报, 2016, 37(2):180-184.
［17］刘明. 支持向量机中Sigmoid核函数的研究［D］. 西安:西安电子科技大学,2009.
2018, Vol. 39 
