基于车牌识别数据的车辆特征分析研究

Vehicle Analysis Based on Vehicle License Plate Recognition Data

近几年来,由于智能卡口设备的广泛部署,全国各地都积累了海量的车牌识别数据,这些数据为智慧城市分析提供了有力保障。车辆特征的分析,可以为城市交通、城市犯罪和城市治理等方面做出决策依据。因此,本文根据车牌识别数据,采用数据挖掘k-means聚类分析方法,分析路网中的车辆行为特征,对车辆行为进行时空刻画。分析发现,一般情况下,地域、时间和车辆属性共同决定了部分车辆的行驶规律。除此之外,摄像头的安装位置也会对卡口记录的数据产生极大影响,摄像头区域设有停车区域,极其容易造成车辆停滞车辆产生很多重复数据的情况。更多的情况,家庭用通勤车在工作日表现出很明显的早出晚归特征,并且只在早晚高峰出现行车记录,且轨迹固定,车辆活动具有区域性。研究结果表明,从车辆的角度解析城市交通,从交通的角度剖析城市发展,对智慧城市,智慧交通的研究和政策制定具有重大意义。

Recently, large number of license plate recognition data have been accumulated throughout the country due to the extensive deployment of intelligent card port equipment. These data provide a powerful guarantee for the analysis of intelligent cities.The analysis of vehicle characteristics can provide decision-making basis for urban traffic, urban crime and urban governance. Therefore, according to the license plate recognition data, this paper uses the data mining K-means clustering analysis method to analyze the vehicle behavior characteristics in the road network, and describes the time and space of the vehicle behavior. The analysis shows that in general, the driving rules of some vehicles are decided by region, time and vehicle attributes.In addition, the installation position of the camera will have a great impact on the data of the recording of the card. The camera area has a parking area, which is extremely easy to cause a lot of duplication of data in the vehicle stagnant vehicle. More and more, the home use commuter car shows a clear feature of early arrival and evening return in the working day, and only in the morning and evening peak running record, and the track is fixed, the vehicle activity is regional. The research results show that the analysis of urban traffic from the angle of vehicles and the analysis of urban development from the perspective of traffic is of great significance to the research and policy making of intelligent cities, intelligent traffic.