Abstract
Computer vision is becoming more and more prevalent in recent years and is increasingly being used in various applications. One of these fields is RoboCup Soccer Small Size League (SSL). This report investigates solutions to recognize individual robots on the field-of-play and their limitations. The first of these classification methods is based on the Circle Hough Transform. This method utilizes edge detection in an image to determine the location and size of circles present in an observed image. The color within circles identified to belong to an individual robot is used to assign the robot with a unique ID. The second approach utilizes Principal Component Analysis to reconstruct a given image of a robot, comparing this reconstructed input image to the original. This approach is like that used commonly in facial recognition. The difference between the original and reconstructed image is used to assign an ID. Lastly, a Convolutional Neural Network is investigated. This method, which is increasingly being used in pattern recognition, utilizes massive data sets to train a classifier on various features. The performance of each classification method is obtained by attempting to classify a set of test images with variation in the height of image capture, the angle of image capture, and the lighting conditions- all of which are areas of concern in the implementation of a computer vision system in this application. The testing results in a clear distinction in robustness of the system, with the Convolutional Neural Network having a perfect performance in the testing data. Implementation in the actual system is discussed, reasoning that the Circle Hough Transform has an advantage in terms of simplicity.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Laws of the RoboCupSoccer small size league 2018. RoboCup (2020)
Hicklin, J.: Deep learning with MATLAB: transfer learning in 10 lines of MATLAB code video. https://www.mathworks.com/videos/deep-learning-with-matlab-transfer-learning-in-10-lines-of-matlab-code-1487714838381.html
Borovicka, J.: Circle detection using hough transforms documentation (2003)
Edge detection-Sobel. https://www.cs.auckland.ac.nz/courses/compsci373s1c/PatricesLectures/Edge%20detection-Sobel_2up.pdf
Technical University of Crete, Intelligent Systems Laboratory: Canny Edge Detector
Turk, M., Pentland, A.: Eigedces for recognition. J. Cogn. Neurosci. 3, 71–86 (1991)
What’s new in YOLO v3? https://towardsdatascience.com/yolo-v3-object-detection-53fb7d3bfe6b
Ommer, N.: RoboCup-SSL/ssl-vision. https://github.com/RoboCup-SSL/ssl-vision
Collins, R.: LoG and DoG filters (2007)
Fisher, R. et al.: Feature detectors - sobel edge detector. https://homepages.inf.ed.ac.uk/rbf/HIPR2/sobel.htm#:~:text=Brief%20Description,in%20an%20input%20grayscale%20image
Bounding box orientation, Stack Overflow. https://stackoverflow.com/questions/46154545/bounding-box-orientation. Accessed 12 Nov 2018
OpenCV: contour features. https://docs.opencv.org/3.4/dd/d49/tutorial_py_contour_features.html
Microsoft corporation, Microsoft LifeCam studio datasheet (2014). https://media.flixcar.com/f360cdn/Microsoft-303383578-Microsoft-TDS_LifeCamStudio.pdf. Accessed 12 Nov 2018
OpenCV: feature detection. https://docs.opencv.org/3.1.0/dd/d1a/group__imgproc__feature.html#ga47849c3be0d0406ad3ca45db65a25d2d
OpenCV: feature detection. https://docs.opencv.org/3.1.0/dd/d1a/group__imgproc__feature.html#ga47849c3be0d0406ad3ca45db65a25d2d.+
Get RGB value opencv python, Stack Overflow. https://stackoverflow.com/questions/12187354/get-rgb-value-opencv-python. Accessed 12 Nov 2018
Histogram equalization. https://www.math.uci.edu/icamp/courses/math77c/demos/hist_eq.pdf
Viorel Nicolau., Mihaela Andrei. On Image Compression for Mobile Robots Using Feed-Forward Neural Networks. Proceeding of 8th International workshop soft computing application, SOFA (2018), vol. II, pp 112–121
Park, N., De Luise, D.L., Rivera, D., Bustamante, L.M., Hemanth, J., Olariu, T., et al.: Multi-neural networks object identification. In: Proceeding of 8th International workshop soft computing application, SOFA, vol. II, pp 189–196 (2018)
Chua, S.-L., Foo, L.K., Juboor, S.S.S.: Towards real-time recognition of activities in smart homes, Int. J. Adv. Intell. Paradigms (IJAIP) 15(2), 146–164 (2020)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2023 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Cherfouh, K., Handerson, E.W., Gu, J., Scheme, E., Asad, M., Farooq, U. (2023). Robot Identification using Modern Pattern Recognition Techniques. In: Balas, V.E., Jain, L.C., Balas, M.M., Baleanu, D. (eds) Soft Computing Applications. SOFA 2020. Advances in Intelligent Systems and Computing, vol 1438. Springer, Cham. https://doi.org/10.1007/978-3-031-23636-5_3
Download citation
DOI: https://doi.org/10.1007/978-3-031-23636-5_3
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-23635-8
Online ISBN: 978-3-031-23636-5
eBook Packages: Intelligent Technologies and RoboticsIntelligent Technologies and Robotics (R0)