Vehicle Face Recognition Algorithm Based on Fusion of Siamese Neural Network

Abstract

Because the license plate is easy to be altered, blocked and forged, the method of license plate recognition alone cannot accurately and quickly confirm the vehicle identity. This paper proposes a vehicle re-recognition method based on the fusion of Siamese deep neural network. The method is based on the huge differences in the face area of each car. First, the YOLOv3 algorithm is improved to detect the face area of the vehicle picture, and then the improved Siamese network algorithm recognition is used. Finally, the feature output of vehicle face image is mapped to Euclidean distance for vehicle face recognition. The YFSDNN (YOLOv3 Fusion Siamese Deep Neural Network) method is evaluated in data sets and experiments. The experimental results show that this method not only has high accuracy, but also greatly improves the detection and recognition speed, which can meet the real-time needs of vehicle face recognition.