Design and Implementation of Drone Using Deep Learning

In this modern world of technology, there are multiple applications of robots which are using deep learning to interact with the human world, instead of using working humans. So, developing a system that would help us detect and recognize a face has become necessary to find out criminals and track for any unwanted objects flying nearby. The use of such unmanned air vehicles (UAV) to detect criminals and also alert for any moving objects that try to attack can help in both surveillance and infiltration will lead us to a future where no person will be required to carry out such missions physically, as they will be monitoring from a safe distance. The use of OpenCV library in python to detect and recognize a face made it possible to implement this system. In this research, Haar cascades are used to detect and recognize a face, and the code is implemented onto a Raspberry Pi. Using deep learning, the pictures of the face stored during detection can be used to recognize the person and hence successfully implementing the core software for this project. To track unwanted objects that may attack, an IR sensor is attached to the drone which when detects an obstacle and takes a picture on the Pi Camera sending it directly via email to the user.