Vehicle Identification and Classification System Based on Decision Tree

Abstract

. The vehicle identification and classification system based on Decision tree was designed in this paper. Firstly, the laser distance measuring sensor and radar sensor were used for data acquisition in order to get the information of the vehicle side-view profile. Secondly, the length, the height, the length ratio of the top and the variance of the top height were extracted as the feature of the vehicle. Thirdly, a large number of statistical samples were used to estimating the gross distribution of each motorcycle type, and through the decision tree, a machine for vehicle identification and classification can be built to achieve the vehicle auto-classification. The experimental results show that the system improved the accuracy of vehicle identification and achieved good results in practical operation.