Abstract: In view of the low efficiency of traditional vehicle license tag copying, a method of automatic vehicle license tag copying is proposed based on quad-rotor UAV helicopter. UAV is used to collect a large number of vehicle overlook maps of vehicle. The overlook maps are preprocessed and lower sampled. Then  the gradient histogram feature of the vehicle image is extracted. The image features are input into the convolution neural network for training. Vehicle identification model is obtained after training. The model is used to recognize vehicles. According to the shape feature of vehicle, vehicle attitude is estimated. The position and angle of the camera on the UAV are calculated according to the position and orientation of the vehicle. UAV experimental platform is built to test system of automatic vehicle license tag copying. The experimental results show that the UAV can independently capture the clear license tag image of vehicle. Automatization of vehicle license tag copying is realized.

Key words: quad-rotor UAV helicopter, automatization of vehicle license tag copying, recognition of vehicle