Vision networks for RoboCup

Author: Nasrullah M Nafis

Nafis, Nasrullah M, 2021 Vision networks for RoboCup, 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 with the details.


To win against human soccer by 2050, RoboCup Federation has developed a platform for scientists to do research and develop intelligent multitask performing robots. This paper has been focused to develop a vision system for Robotis Op3 humanoid robots. The robot needs to be able to detect different f ield markings to participate in the RoboCup competition. The markings include straight lines, centre field circles and line intersections. This paper suggests a method to perform edge detection which eventually helps detect lines, circles and intersections . Two different methods of Hough Line transform have been tested on various data sets for detecting straight lines. This paper shows the use of Hough circle transforms to detect the field’s centre circle. A simple simultaneous linear equation has been used method for detecting straightline intersections. From the results we can see those realas a time conditions like luminance of light, camera angle plays a big role in the detection process.

Keywords: RoboCup, Line Detection, OpenCV, Hough Line Transform, Probabilistic Hough Transform, Circle Detection, Intersection Detection

Subject: Computer Science thesis

Thesis type: Masters
Completed: 2021
School: College of Science and Engineering
Supervisor: Prof. David M W Powers