Deteksi Bola Pada Konversi Poin Permainan Tenis Meja Berbasis Video Processing
DOI:
https://doi.org/10.34818/eoe.v5i3.7661Abstract
Abstrak Pendeteksian pergerakan bola pada tenis meja dengan metode background substraction and estimation adalah satu dari sekian banyak metode yang dipakai. Penentuan kondisi bola dengan Kalman filter, yaitu proses deteksi dan koreksi yang digunakan sangat cocok pada metode ini. Deteksi dan tracking bola dengan background substraction and estimation based on Kalman filter, membuat penelitian tugas akhir ini memiliki keunikan dari metode yang lain. Pada tugas akhir ini dirancang suatu program pengolahan video untuk mendeteksi pergerakan dan menentukan kondisi bola dari permainan tenis meja. Metode background subtraction and estimation digunakan sebagai pemisah antara background dan foreground, sehingga didapatkan objek yang akan di deteksi. Pendeteksian dan tracking bola dengan Kalman filter, sehingga sistem dapat menentukan bola saat melewati net dan garis. Pada deteksi bola berwarna oranye, saat jarak 1 meter dengan luas area 250 pixels menghasilkan akurasi sebesar 100%. Deteksi bola melewati garis, terdapat pada luas area 500 pixels dan 250 pixels dengan akurasi 100%. Sedangkan deteksi bola melewati net, terdapat pada luas area 500 pixels dengan akurasi 81,81%. Kata kunci : pengolahan video, deteksi gerakan, tenis meja, background subtraction, Kalman filter. Abstract Detection of moving ball on table tennis with background substraction and estimation method is one of the many methods used. Determine ball conditions with the Kalman filter, the detection and correction process is very suitable on this method. Detection and tracking of balls with substraction background and estimation based on Kalman filter, making this final research has unique way than the other methods. In this final project designed a video processing program to detect movement and determine the condition of the ball from table tennis game. Background subtraction and estimation method is used to separate between the background and foreground, so that the object will be detected. Detection and tracking of the ball with a filter calm, allowing the system to determine the ball as it passes through the net and the line. In the detection of orange ball, at 1 meter distance and area 250 pixels get 100% accuracy. Ball detection when passing line, could be at area 250 pixels and 500 pixels with accuracy 100%. Ball detection when passing the net at area 500 pixels with accuracy 81,81%. Keywords: video processing, detection, table tennis, background subtraction, Kalman filter.Downloads
References
[1] FIFA. 2012. “FIFA Quality Programme for Goal-Line Technology”. International Football Association Board (IFAB).
[2] Esma Bilgin. 2009. “Kalman filter for dummies” http://bilgin.esme.org/BitsAndBytes/KalmanFilterforDummies.
[3] Singla Nishu. 2014. ”Motion Detection Based on Frame Difference Method”, International Journal of Information & Computation Technology. Volume 4, Number 15, pp. 1559-1565.
[4] Mao Jinzi. 2006. “Tracking A Tennis Ball Using Image Processing Techniques”, University of Saskatchewan.
[5] Yu, X. and Farin, D. 2005. “Current and emerging topics in sports video processing”, IEEE International Conference on Multimedia and Expo.
[6] Cornell University. 2015. “Table tennis game tracker”. http://people.ece.cornell.edu/land/courses/ece5760/FinalProjects/s2015.
[7] Jong-Yun Kim, Tae-Yong Kim. 2009. “Soccer Ball Tracking using Dynamic Kalman filter with Velocity Control”, Chung-Ang University.
[8] Pal sumit kumar, ghorai sohan. 2013 “Moving Object Tracking System In Video With Kalman filter”, IJERT.
[9] Greg Welch, Gary Bishop. 2006. “An Introduction to the Kalman filter”, University of North Carolina.
[10] M.S.Grewal, A.P. Andrews. 2001 "Kalman filtering - Theory and Practice Using MATLAB", Wiley.
[11] Scott Jesse, Pusateri Michael A, Cornish Duane. 2013 “Kalman filter Based Video Background Estimation”, Penn State University.



