An Improved Drone Detection Method Using Deep Learning for Augmentation Detection Speed
Mohammad Bahrami, Seyyed Amir Asghari, Mohammadreza Binesh Marvasti, Sajjad Ansarian · 2022
In recent years, remotely piloted birds (drones) have become significantly more accessible to the general public. Affordable prices economical, being equipped with advanced technologies, small sizes, easy portability and setup, create many concerns. For example, drones can be used for destructive activities, spying on private properties, monitoring vital places, carrying dangerous objects which is a big threat to the society. For this reason, the identification of drones is considered important has been in order to solve the above challenges that the university and the industry have provided several solutions in recent years. In this paper, an improved method is introduced to detect drones based on deep learning. This system is designed based on camera recognition. Based on the camera images, the system determines the location of the drone on the image by dragging the box around it. For our methods OpenCV library and YOLO algorithm are used. The simulation results show that the drone can be detected in 1 7 milliseconds and the detection is done with 85% accuracy.