Abstract: The system adopted the method  of dual positioning of GPS and image recognition, used the PID flight attitude control of unmanned aerial vehicle (UAV), and the open source computer vision library (OpenCV) was used to process the images. The cascade classifier was applied to solve the problem of target accurate detection. The positioning and navigation control system designed by integrating various technologies can quickly complete the transmission and processing of information, and enhance the system function. After GPS rough positioning, the image capture, preprocessing, fitting and discrimination of the target graphics are accurately positioned to ensure the real-time, efficient and accurate navigation  control system of UAV.

Key words: unmanned aerial vehicle (UAV), GPS positioning, image processing, target detection, computer vision technology