應用棒球運動視訊之棒球軌跡研究

A Study of Baseball Trajectory Tracking using Baseball Sport Videos

林家緯(Chia-Wei Lin)

指導教授 ： 張元翔

中原大學/工學院/資訊工程研究所/碩士(2013年)

https://doi.org/10.6840/cycu201300687

摘要

隨著多媒體技術發展，運動視訊的即時分析可以提昇人們在觀賞運動視訊時的娛樂性。本研究旨在發展一套「棒球運動視訊之棒球軌跡追蹤系統」，可以自動化重現投手投球過程中之棒球運動軌跡。研究方法包括：(1) 影像前置處理；(2) 棒球偵測；及 (3) 棒球軌跡追蹤。系統主要是利用影像及視訊處裡技術，於續影格中擷取棒球候選區域，同時利用位置、尺寸、連顏色及形狀等特徵偵測棒球位置，最後採用卡爾曼濾波器進行棒球軌跡追蹤。初步研究結果顯示，本方法可以達到82%之平均棒球偵測率及91%之平均棒球軌跡追蹤準確率。總結而言，本研究提出一套有效的自動化運動分析系統，可以重現棒球軌跡，進一步帶給觀賞者更豐富的比賽內容。

關鍵字

運動視訊分析 ； 物件追蹤 ； 影像處理 ； 卡爾曼濾波器

並列摘要

With the development of multimedia technology, real-time analysis of sport videos can enhance the entertainment when people are watching sport videos. This study aims to develop a baseball trajectory tracking system using sport videos, which can automatically reproduce the baseball trajectory when a pitcher is throwing the baseball. Research methods include: (1) image preprocessing; (2) baseball detection; and (3) baseball trajectory tracking. Our system mainly uses image and video processing technology to extract baseball candidate regions in consecutive frames. Meanwhile, the baseball’s location is detected using position, size, color, and shape features. Finally, the Kalman filter is used for the baseball trajectory tracking. Preliminary results show that this method could achieve the average baseball detection rate of 82% and the average baseball trajectory tracking accuracy of 91%. In conclusion, this study presents an effective automated motion analysis system that could reproduce the baseball trajectory, further bringing viewers a richer game content.

並列關鍵字

image processing ； Kalman filter ； object tracking ； sport video analysis

參考文獻

[2] André Guéziec, “Tracking pitches for broadcast television,” IEEE International Conference on Computer, Vol. 35, No. 3, pp. 38-43, 2002.

[3] D. Liang, Y. Liu, Q. Huang, and W. Gao, “A Scheme for Ball Detection and Tracking in Broadcast Soccer Video,” Springer-Verlag Berlin Heidelberg, pp. 864 – 875, 2005.

[5] Xinguo Yu, Nianjuan Jiang, and Ee Luang Ang, “Trajectory-Based Ball Detection and Tracking with Aid of Homography in Broadcast Tennis Video,” Visual Communications and Image Processing, SPIE Vol. 6508, 650809.

[6] K. C. P. Wong and L. S. Dooley, “High-Motion Table Tennis Ball Tracking for Umpiring Applications,” IEEE International Conference on Signal Processing, pp. 2460-2463, 2010.

[7] B. Chakraborty and S. Meher, “A Trajectory-Based Ball Detection and Tracking System with Applications to Shot-type Identification in Volleyball Videos,” IEEE International Conference on Signal Processing and Communication, 2012.