Vehicle Brand Recognition by Deep Neural Networks

Abstract

Vehicle brand recognition aims to identify the brand of different vehicles in traffic video or images. The recognition result, that is, the brand of a vehicle, is an important information for a certain vehicle, which could be applied in an intelligent transport system (ITS). In this paper, we proposed a novel framework for vehicle brand recognition using deep neural network, that is, YOLO-V3. The network learned how to locate the front face of a vehicle and then recognize its brand automatically. After training, the network can recognize vehicle brand in video or images in a relative high accuracy. According to the experimental results, the proposed method shows a good performance on a traffic image dataset.