Vehicle Detection and Classification Using Neural Network

Nowadays, intelligent transportation system has become more important in traffic surveillance. The proposed implementation deals with vehicle detection and classification using the neural network. The current work is carried out in MATLAB. The images are preprocessed and then Segmentation is carried out.  Morphological operations are applied to get a vehicle in the image. The features of vehicle images are extracted to detect a class of vehicle. Additional 8 histogram count generated from the vehicle shape is used to characterize the vehicles in the images. These features are trained with the Artificial Neural Network (ANN). So that vehicle images are classified as image with bike or car based on the input i.e. testing and trained features. The average accuracy of the vehicle detected and classified is achieved with percentage of 80%.