Stabilzation Position of Quadcopter Using Vision-Based Corner Detector from Top-Down Footage of Camera

Abstract

Quadcopter is a kind of robot which is popularly used in both academic and industrial environment. In this paper, we present and implement a method to stabilize a quadcopter prototype’s position using feature extraction and tracking from camera footage. The quadcopter's position and linear velocity are determined from images which are captured by a downward-facing camera - Logitech C270. First, Shi-Tomasi technique is used to detect corners in the images and from this method, displacement of the quadcopter is yielded. Linear velocity is then calculated by using the quadcopter’s displacement. Once the linear velocity of the quadcopter has been estimated, the cascade PID controller is proposed to stabilize the hovering quadcopter’s position. Simulation results prove the ability of controlller on Matlab/Simulink. Then, a real quadcopter prototype is built to evaluate the proposed method and the experimental results recording in approximately 70 seconds show that the quadcopter remained its position with minimal error.