Vehicle Types Recognition Based on Neural Network

As a key technology of intelligent transportation system, higher identification accuracy, robustness and real-time are needed in vehicle recognition. Therefore, in view of the features of vehicle types, this paper proposes a BP neural network car types classifier method based on fuzzy C-means clustering. First, on the basis of the pretreatment of the images of the vehicle, we abstract the features of car types from images and classify the massive data by fuzzy C- means clustering algorithm. Then, design the BP neural networks to train and test the classified data. Finally it is carried on compressive judgment by the computer. Experiments prove the validity of the classifier. It can recognize the highway vehicle types rapidly.