Optical Sky Scanner

Abstract

This thesis explores the process of designing a detection system which have the capability to detect Drones, helicopters and airplanes using machine learning and Image processing. The modern world has advanced in technology you can find security cameras of different capabilities having zooming and detection of objects but all of them can zoom the digital image of the object to a certain limit moreover the specifications of those cameras are fix and depends upon its manufacturing company. What if you have the choice to change the specification of the camera as per your own requirement? What if your zooming capability in much more than ordinary security cameras? In order to introduce a model in which users have the choice to change the specification of camera as required and to enhance the zooming capability of the camera plus providing Optical zoom which is far better than digital zoom we have designed a model which have the capability to scan the area up to 180 degrees in Azimuth and 180 degrees in elevation A dataset containing 10000 images of helicopters, drones and airplanes have been used to train the model in order to recognize and detect the target object.