Deep Reinforcement Learning based Target Detection for Unmanned Aerial Vehicle

Considering the automation for delivering items using unmanned aerial vehicles has emerged with many research and studies. The research is proposed with one vital aspect to provide foremost necessary medical aids by reaching accurately to patients. This paper deals with developing an integrated camera-based positioning system with GPS for quadcopters to automate the landing process. In addition to the frontal-viewing camera, a downward-viewing camera is used for landing at the identified location based on a QR code. The efficient deep learning-based method for object detection is followed by a pixel-extraction-based method for the code recognition to identify the site accurately.