Vision based control using parrot AR drone quadrotor

• Main Author: Muhammad Ruzaini A. Rasip.
• Other Authors: Wang Jianliang
• Format: Final Year Project
• Language: English
• Published: 2013
• Subjects: DRNTU::Engineering::Electrical and electronic engineering::Control and instrumentation::Robotics
• Online Access: http://hdl.handle.net/10356/55133
• Institution: Nanyang Technological University

This paper, is the author's Final Year Project report represent the research on Vision Based Control of Quad-rotor and experimental of dynamic stability as well as vision tracking. It is to evaluate the development of a quad-rotor having the ability to localize, navigate & tracking autonomously in unknown and GPS denied environment. Through performance & surrounding’s data, the function techniques uses on-board monocular camera and quad-rotor navigates independently from any artificial markers or external sensors. The approach comprises systems; 1) Monocular key frame based on simultaneous localization and mapping (SLAM) system for the position estimation and projector path; depicts through GUI. 2) Installation of Kalman Filter and PID controller which allows control dynamics to fuse, synchronize and governs the orientation position of drone to its projector flight path & scale estimation. 3) Installation of the selective tracking system into the Drone’s view, using continuous adaptive mean shift algorithm (Camshift). With regards to time-frame and planning, construction of the Drone is not a requirement. From embedded sensory installed, data captured will thus generate a visual map of 3D environment. It is based through mathematical statistical formulation of Maximum Likelihood (ML) approach method. Derivation of a closed loop system, for the ‘maximum likelihood estimator of the scale’ is needed.GUI will be created for the environment and captured data display in real-time. Captured data would link to excel for documentation. Final phase of this project would be the implementation of the tracking system through camshaft technique. The subject will be tracked on selective basis from the input of the user.