ABSTRACT
This paper proposes a novel technique for clustering and classification of object trajectory-based video motion clips using spatiotemporal functional approximations. A Mahalanobis classifier is then used for the detection of anomalous trajectories. Motion trajectories are considered as time series and modeled using the leading Fourier coefficients obtained by a Discrete Fourier Transform. Trajectory clustering is then carried out in the Fourier coefficient feature space to discover patterns of similar object motions. The coefficients of the basis functions are used as input feature vectors to a Self-Organising Map which can learn similarities between object trajectories in an unsupervised manner. Encoding trajectories in this way leads to efficiency gains over existing approaches that use discrete point-based flow vectors to represent the whole trajectory. Experiments are performed on two different datasets -- synthetic and pedestrian object tracking - to demonstrate the effectiveness of our approach. Applications to motion data mining in video surveillance databases are envisaged.
- Chang, S.-F., Chen, W. Meng, H.J. Sundaram, H. and Zhong, D. A fully automated content-based video search engine supporting spatiotemporal queries. IEEE Trans. Circuits Syst. Video Technol., vol. 8, no. 5, pp. 602--615, Sept. 1998. Google ScholarDigital Library
- Jeannin, S. and Divakaran, A. MPEG-7 visual motion descriptors. IEEE Trans. Circuits Syst. Video Technol., vol. 11, no. 6, pp. 720--724, June 2001. Google ScholarDigital Library
- Dagtas, S., Ali-Khatib, W., Ghafor, A. and Kashyap, R.L. Models for motion-based video indexing and retrieval. IEEE Trans. Image Proc., vol. 9, no. 1, pp. 88--101, Jan 2000. Google ScholarDigital Library
- Aghbari, Z., Kaneko, K. and Makinouchi, A. Content-trajectory approach for searching video databases. IEEE Trans. Multimedia, vol. 5, no. 4, pp. 516--531, Dec. 2003. Google ScholarDigital Library
- Bashir, F., Khokhar, A. and Schonfeld, D. Segmented trajectory-based indexing and retrieval of video data. In Proc. IEEE Int. Conf. Image Processing, Spain, 2003, pp. 623--626.Google ScholarCross Ref
- Hsu, C.-T. and Teng, S.-J. Motion trajectory based video indexing and retrieval. In Proc. IEEE Int. Conf. Image Processing, pt 1, 2002, pp. 605--608.Google Scholar
- Bashir, F., Khokhar, A. and Schonfeld, D. A hybrid system for affine-invariant trajectory retrieval. In Proc. MIR'04, 2004, pp. 235--242. Google ScholarDigital Library
- Shim, C. and Chang, J. Content-based retrieval using trajectories of moving objects in video databases. In Proc.IEEE. 7th Int. Conf. Database Systems for Advanced Applications, 2001, pp. 169--170. Google ScholarDigital Library
- Shim, C. and Chang, J. Trajectory-based video retrieval for multimedia information systems. In Proc. ADVIS, LNCS 3261, 2004, pp. 372--382. Google ScholarDigital Library
- Jin, Y. and Mokhtarian, F. Efficient video retrieval by motion trajectory. In Proc. British Machine Visiom Conference'04, 2004.Google ScholarCross Ref
- Khalid, S. and Naftel, A. Evaluation of matching metrics for trajectory-based indexing and retrieval of video clips. In Proc.7th IEEE Workshop Applications Comp. Vision (WACV/MOTION'05), Colorado, USA, Jan. 2005, 242--249. Google ScholarDigital Library
- Wang, L., Hu, W. and Tan, T. Recent developments in human motion analysis. Pattern Recognition, vol. 36, no. 3, pp. 585--601, 2003.Google ScholarCross Ref
- Hu, W., Tan, T., Wang, L. and Maybank, S. A survey on visual surveillance of object motion and behaviors. IEEE Trans. Systems, Man & Cybernetic, Part C, vol.34, no.3, pp. 334--352, August 2004. Google ScholarDigital Library
- Johnson, N. and Hogg, D. Learning the distribution of object trajectories for event recognition. Image Vis. Comput., vol. 14, no. 8, pp. 609--615, 1996.Google ScholarCross Ref
- Owens, J. and Hunter, A. Application of the self-organising map to trajectory classification. In Proc. IEEE Int. Workshop Visual Surveillance, pp. 77--83, 2000. Google ScholarDigital Library
- Hu, W., Xiao, X., Xie, D., Tan, T. and Maybank, S. Traffic accident prediction using 3-D model-based vehicle tracking. IEEE Trans. Vehicular Tech., vol. 53, no. 3, pp. 677--694, May 2004.Google ScholarCross Ref
- Alon, J., Sclaroff, S., Kollios, G. and Pavlovic, V. Discovering clusters in motion time-series data. In Proc. IEEE CVPR, June 2004. Google ScholarDigital Library
- Hu, W., Xie, D., Tan, T. and Maybank, S. Learning activity patterns using fuzzy self-organizing neural networks. IEEE Trans. Systems, Man & Cybernetic, Pt. B, vol. 34, no. 3, pp. 1618--1626, June 2004. Google ScholarDigital Library
- Faloutsas, C. Ranganathan, M. and Manolopoulos, Y. Fast subsequence matching in time-series databases. In Proc. ACM SIGMOD Conf., 1994, pp. 419--429. Google ScholarDigital Library
- Chan, K. and Fu, A. Efficient time series matching by wavelets. In Proc. Int. Conf. Data Engineering, Sydney, March 1999, pp. 126--133. Google ScholarDigital Library
- Keogh, E., Chakrabarti, K., Pazzani, M. and Mehrota, S. Locally adaptive dimensionality reduction for indexing large time series databases. In Proc. ACM SIGMOD Conf., 2001, pp. 151--162. Google ScholarDigital Library
- Cui, Y. and Ng, R. Indexing spatio-temporal trajectories with Chebyshev polynomials. In Proc. ACM SIGMOD Conf., Paris, June 2004, pp. 599--610. Google ScholarDigital Library
- Kohonen, T. Self-Organizing Maps, 2 nd ed. New York: Springer-Verlag, 1997, vol. 30. Google ScholarDigital Library
- CAVIAR: Context aware vision using image-based active recognition. (2004, Jan. 10). {Online}. Available: http://homepages.inf.ed.ac.uk/rbf/CAVIARGoogle Scholar
- Buzan D., Sclaroff S., Kollios G. Extraction and clustering of motion trajectories in video. In Proc. International Conference on Pattern Recognition, 2004. Google ScholarDigital Library
- Vlachos, M., Kollios, G. and Gunopulos, D. Discovering similar multidimensional trajectories. In Proc. International Conference on Data Engineering, p. 673, 2002. Google ScholarDigital Library
- Yacoob, Y. and Black, M. J. Parameterized modeling and recognition of activities. Computer Vision and Image Understanding, Vol. 73 (2), pp. 232--247, Feb. 1999. Google ScholarDigital Library
- Bashir, F., Khokhar, A. and Schonfeld, D. HMM-based motion recognition system using segmented PCA, To appear in Proc. IEEE International Conference on Image Processing (ICIP 2005), Sept. 11 - Sept. 14, 2005. Genova, Italy.Google ScholarCross Ref
- Bashir, F., Ashfaq, A., Khokhar, A. and Schonfeld, D. Object trajectory-based motion modeling and classification using Hidden Markov Models, submitted to IEEE Transactions on Image Processing. Google ScholarDigital Library
- Rea, N., Dahyot, R. and Kokaram, A. Semantic event detection in sports through motion understanding. In Proc. of Conference on Image and Video Retrieval, Dublin, Ireland, July 21-23, 2004.Google ScholarCross Ref
- Ivo, F. and Sbalzarinii, J.T. Machine learning for biological classification applications. Technical Report Center for Turbulence Research, Proceedings of the Summer Program 2002.Google Scholar
- Johnson, R.A. and Wichern, D.W. Applied Multivariate Statistical Analysis, 4th Ed., Prentice-Hall, NJ, 1998. Google ScholarDigital Library
- Keogh, E. and Kasetty, S. On the Need for Time Series Data Mining Benchmarks: A Survey and Empirical Demonstration. Knowledge and Data Discovery, 2002, pp. 102--111. Google ScholarDigital Library
Index Terms
- Classifying spatiotemporal object trajectories using unsupervised learning of basis function coefficients
Recommendations
Classifying spatiotemporal object trajectories using unsupervised learning in the coefficient feature space
This paper proposes a novel technique for clustering and classification of object trajectory-based video motion clips using spatiotemporal function approximations. Assuming the clusters of trajectory points are distributed normally in the coefficient ...
Activity classification and anomaly detection using m-mediods based modelling of motion patterns
Techniques for video object motion analysis, behaviour recognition and event detection are becoming increasingly important with the rapid increase in demand for and deployment of video surveillance systems. Motion trajectories provide rich ...
Motion-based behaviour learning, profiling and classification in the presence of anomalies
Techniques for understanding video object motion activity are becoming increasingly important with the widespread adoption of CCTV surveillance systems. Motion trajectories provide rich spatiotemporal information about an object's activity. This paper ...
Comments