Author: Keith Man
Man, Keith, 2018 Maritime RobotX: A Ball Launcher and Target Tracking System, Flinders University, College of Science and Engineering
Terms of Use: This electronic version is (or will be) made publicly available by Flinders University in accordance with its open access policy for student theses. Copyright in this thesis remains with the author. You may use this material for uses permitted under the Copyright Act 1968. If you are the owner of any included third party copyright material and/or you believe that any material has been made available without permission of the copyright owner please contact copyright@flinders.edu.au with the details.
TopCat is a WAM-V unmanned surface vehicle (USV) outfitted to compete in the Maritime RobotX Challenge. The Detect and Deliver task in the 2018 Maritime RobotX Challenge requires a USV to detect a symbol on a floating dock and deliver four racquetballs into the dock. Currently, vision based cameras are used to identify a set of symbols within the cameras’ field of view as the USV approaches them. Symbol detection is carried out using a blob detection algorithm that reports the symbol identified. However, the current symbol detection algorithm is prone to misidentifying symbols in changeable lighting environments and does not report the position of symbols within its field of view. This thesis addresses this issue through the development of a target tracking system that uses blob detection algorithm with automatic parameter configuration in combination with region of interest data from the lidar and simultaneous localisation and mapping systems to identify a symbol’s location and pass that information to a pan-tilt servo system. The automatic configuration of internal parameters allows the algorithm to identify symbols in a wide range of lighting environments while region of interest data from lidar sensors is used to reduce the area in an image stream that needs to be searched for symbols. After identifying the position of a symbol within the cameras’ field of view, the required servo positions are then sent to a pan-tilt servo system acting as the targeting system of a ball launcher. The target tracking system developed will enable a ball launcher to be aimed at a target as part of the Detect and Deliver task of the 2018 Maritime RobotX Challenge.
Keywords: ball launcher, Maritime RobotX Challenge, OpenCV, real time image processing, region of interest, shape detection, target tracking
Subject: Engineering thesis
Thesis type: Masters
Completed: 2018
School: College of Science and Engineering
Supervisor: Professor Karl Sammut