IMPLEMENTATION OF OBJECT TRACKING AND GIMBAL POINTING IN MULTIROTOR DRONE USING OPENTLD ALGORITHM ON ROS (ROBOT OPERATING SYSTEM) FRAMEWORK

• Main Author: Hariz Dzulfikar , Fadhel
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/46968
• Institution: Institut Teknologi Bandung

Missile technology is part of Indonesia's defense and security industry development program. In missile technology, there is a main part that functions as the final guide towards the target of the missile system, that is Electro Optical Seeker (EOS) system that uses Computer Vision to perform digital image processing to track the target. In its development, we use UAV (Unmanned Aerial Vehicle), such as fixed wing UAV and UAV multirotor to model the actual missile system. The system developed will be implemented on a multirotor drone. UAV technology has been widely applied for environmental and security monitoring, meteorological surveillance, weather research, agriculture, exploitation and exploration of mineral materials even for military purposes. In the monitoring function, the UAV is expected to have an object tracking function so that the target can always be under UAV supervision. This system has function to do object tracking using a camera mounted on a gimbal mechanism on a multirotor drone. The operator has control over the gimbal mechanism by joystick, moving the orientation of the gimbal mechanism toward the selected object being the target. Then the system will do the object tracking and gimbal pointing functions so that the target does not leave the camera's field of view. By implementing object tracking algorithms on a UAV, the information from the object tracking algorithm can be utilized to control UAV system such as gimbal pointing and platform guidance. In this paper, we will discuss how to implement the OpenTLD object tracking algorithm and gimbal pointing on a multirotor drone by utilizing the results of the tracking process using the OpenTLD algorithm in ROS (Robot Operating System) environment. The author wants to implement the gimbal pointing function to compensate for the tracking function when the target object is almost out of the camera's viewpoint. The system testing was carried out using a multirotor UAV with a gimbal mechanism and a camera to do object tracking. Based on the test results, the system can do object tracking using the OpenTLD algorithm in ROS environement. In addition, the system is able to control the gimbal mechanism to maintain the target position so that it is always inside the field of view of the camera.