Optical Flow Tracking algorithm for Quadrotor Positioning Control

Abstract

We present a design of vision based positioning control system for unmanned vehicle, quadrotor. The proposed system is used to maintain horizontal position of quadrotor. It also can be used for specially autolanding purpose. The system must be able to detect any horizontal movement of quadrotor. A camera is used as the movement sensor. Camera is fixed in quadrotor
facing to the ground to capture ground images. Images is processed by a computer to detect horizontal movement of quadrotor. Quadrotor is controlled so its position is always stable above a specific location. The horizontal movement is detected by image processing using Lucas-Kanade optical flow algorithm.