Minimize UAV Landing Error by Development of Precision Landing System Based on Object Detection

Abstract

Unmanned Aerial Vehicle (UAV) or commonly known as drones have undergone various kinds of technological developments, so that they are widely utilized in any field, in this paper focuses on developing a precision landing system for drones when completed in an RFID tag mission in a warehouse, utilizing technologies such as Pixhawk, Raspberry Pi, Intel Realsense T265 for navigation, and a webcam to help detect objects for precision landing of drones. The system of precision landing combines or communicates from Intel Realsense T265 and Vision Camera, which detects a marker where the drone is landing, then Pixhawk receives the command. The addition of this object detection-based precision landing system is to minimize the drone landing error if it only utilizes Intel RealSense T265, thus the development of this precision landing system minimizes the error from the drone when landing, so that the drone still lands in a safe area or predetermined place.