Project basketball

Video Surveillance of Basketball Matchesand Goal Detection : 

By Akhilesh K. Sinha Nishant Singh Supervised by Prof. Amitabha Mukerjee Video Surveillance of Basketball Matchesand Goal Detection Indian Institute of Technology, Kanpur

Motivation : 

Motivation Unsupervised Surveillance (90‘s) Composite Event Discovery [3] Video Understanding Framework [6] Importance of Scoring (another word for counting by some A-PRIORI based method)[6] Automatic Video Interpretation

Relevant Work : 

Relevant Work Object Detection [1][2]. Declarative representation of scenarios [9], using spatio-temporal and logical constraints. Use of game pauses, noises and some event sequences. Cinematic and object features [10]. Scene cuts and camera motion parameters [11]. Voice [12]. Three main categories of approaches are used to recognize scenarios [8]. : 1. Probabilistic/neural network combining potentially recognized scenario 2. Symbolic network that Stores Totally Recognized Scenarios. 3. Symbolic network that Stores Partially Recognized Scenarios.

Our Approach : Overview : 

Our Approach : Overview HAAR Classification Viola-Jones Object Detection Prune Errors Tracked Sequence Goal Detection Basketball Video FG ‘+‘ Img Lost ball

Results: intermediate steps : 

Results: intermediate steps

Goal Detection : 

Goal Detection Camera View not enough. Learning by training, a possibility. External parameters, position of the ball, size of the ball, net interference etc. Our 4 point location based goal detection works for many situation, but, as obvious it is error prone in the final step of deciding goal.

Results & Conclusion : 

Results & Conclusion FG+Image to remove Detection errors. Goals decision, to an extent correctly. Learning alorithms should improve result. Conclusion Goal ≠ Goal ? Goal = Automatic Video Interpretation √ Difficult with a single Camera.

References : 

References Paul Viola, Michael Jones, "Rapid Object Detection using a Boosted Cascade of Simple Features," CVPR, p. 511,  2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'01) - Volume 1,  2001 P. Viola and M. Jones. “Robust real-time object detection”. Technical Report 2001/01, Compaq CRL, February 2001. 8 http://citeseer.ist.psu.edu/viola01robust.html Alexander Toshev, Francois Bremond, Monique Thonnat, "An APRIORI-based Method for Frequent Composite Event Discovery in Videos,“ , p.10-18,  Fourth IEEE International Conference on Computer Vision Systems (ICVS'06),  2006 Hannah M. Dee  and Sergio A. Velastin, “How close are we to solving the problem of automated visual surveillance?” Journal of Machine Vision and Applications, Springer Berlin / Heidelberg, ISSN-0932-8092 (Print) 1432-1769, 2006 G. Médioni, I. Cohen, F. Brémond, S. Hongeng et R. Nevatia, “Event Detection and Analysis from Video Streams”, in IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 23. No. 8, pp. 873-889, Aug 01. F. Brémond, M. Thonnat and M. Zuniga, “Video Understanding Framework For Automatic Behavior Recognition”. In Behavior Research Methods , 38(3), 416-426, 2006. Robert G. Knodell et al., “Formulation and application of a numerical scoring system for assessing histological activity in asymptomatic chronic active hepatitis”, Hepatology, Volume 1, Issue 5, Pages 431 – 435, 2006.

References contd. : 

References contd. A.J. Howell and H. Buxton. “Active vision techniques for visually mediated interaction. “Image and Vision Computing, 0(12):861-871, October 2002. N. Rota and M. Thonnat. “Activity recognition from video sequences using declarative models”. In Proceedings of the 14th European Conference on Articfiial Intelligence (ECAI00), Berlin, Germany, August 2000. Ekin, A. Tekalp, A.M. Mehrotra, “Automatic soccer video analysis and summarization “, Image Processing, IEEE Transactions , Volume: 12, Issue: 7, On page(s): 796- 807, July 2003. Y. Rui, A. Gupta, and A. Acero, “Automatically extracting highlights for TV baseball programs,” in Proceedings of ACM Multimedia, 2000.