Monocular Vision Based UAV Target Detection and Ranging System Implemented on OpenCV and Tensor Flow

In recent years, UAV(Unmanned Aerial Vehicle) becomes more and more popular and it is put into operation in various fields, such as agricultural irrigation, geological exploration and so on. In this paper, an unmanned aerial vehicle target detection and ranging system based on OpenCV and Tensor Flow is discussed, which can detect and avoid obstacles in time, improving the service life and intelligence level of the unmanned aerial vehicle. Firstly, gray-scale need to be put on the image captured by the camera. After Gaussian denoising, Sobel operator is applied for edge detection. Then, target recognition is carried out by using OpenCV and Tensor Flow, following by which is distance measurement, on the basis of OpenCV and video stream. Finally, the measurement results can be obtained.