Automatic Event Detection and Classification Based on Ball Trajectory in Broadcast Tennis Video Using SVM and HMM

An identifying event in sports video has many efforts of sports applications. In this paper, proposed a system for automatic detection of key events in Broadcast Tennis Video (BTV). The ultimate goal is to detect the events of complete tennis match. The detected tennis events are fault, rally and net approach, there are also other events in BTV, they all are considered as secondary one. To detect the events of tennis by analyzing the player`s position and ball tracking. The experiments done in different tennis tournament, which has the events (fault, rally and net approach), the some of the visual features are extracted from MHI (Motion History Image) and modelled by Support Vector Machines (SVM) and Hidden Markov Model (HMM) for recognizing tennis events. In result HMM gives a higher accuracy rate of 96.66% when compared to SVM rate of 86.42%.

[1]  Patrick Gros,et al.  Classification-oriented structure learning in Bayesian networks for multimodal event detection in videos , 2012, Multimedia Tools and Applications.

[2]  Wen Gao,et al.  Action Recognition in Broadcast Tennis Video Using Optical Flow and Support Vector Machine , 2006, ECCV Workshop on HCI.

[3]  Shiqiang Yang,et al.  Motion based event recognition using HMM , 2002, Object recognition supported by user interaction for service robots.

[4]  Walter A. Kosters,et al.  Tennis Patterns : Player , Match and Beyond , 2010 .

[5]  David Windridge,et al.  Ball event recognition using hmm for automatic tennis annotation , 2010, 2010 IEEE International Conference on Image Processing.

[6]  Changsheng Xu,et al.  Using Webcast Text for Semantic Event Detection in Broadcast Sports Video , 2008, IEEE Transactions on Multimedia.

[7]  Lihi Zelnik-Manor,et al.  Event-based analysis of video , 2001, Proceedings of the 2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. CVPR 2001.

[8]  Wei-Ta Chu,et al.  Event detection in tennis matches based on video data mining , 2008, 2008 IEEE International Conference on Multimedia and Expo.

[9]  Mubarak Shah,et al.  High-level event recognition in unconstrained videos , 2013, International Journal of Multimedia Information Retrieval.

[10]  A. Agah,et al.  Recognition of Marker-less Human Actions in Videos Using Hidden Markov Models , 2011 .

[11]  Chung-Ming Kuo,et al.  Event detection for broadcast tennis videos based on trajectory analysis , 2012, 2012 2nd International Conference on Consumer Electronics, Communications and Networks (CECNet).

[12]  Willem Jonker,et al.  Recognizing Strokes in Tennis Videos using Hidden Markov Models , 2001, VIIP.

[13]  Noel E. O'Connor,et al.  Real-time event classification in field sport videos , 2015, Signal Process. Image Commun..

[14]  Patrick Gros,et al.  HMM based structuring of tennis videos using visual and audio cues , 2003, 2003 International Conference on Multimedia and Expo. ICME '03. Proceedings (Cat. No.03TH8698).