Camera Motion Detection in the Rough Indexing Paradigm

This paper presents our camera motion detection method (pan, tilt and zoom) for TRECVID 2005. As input data, we only extract P-Frame motion compensation vectors directly from the MPEG compressed stream and we so achieve a performance of 3-4 times faster than real time. Our method is based on global camera motion estimation and a likelihood based signicance test of the camera parameters. The best run (RI-3) on the TRECVID 2005 test set provides 0.912 for mean precision and 0.737 for mean recall.

[1]  Sanjeev R. Kulkarni,et al.  A new method for camera motion parameter estimation , 1995, Proceedings., International Conference on Image Processing.

[2]  Patrick Bouthemy,et al.  A unified approach to shot change detection and camera motion characterization , 1999, IEEE Trans. Circuits Syst. Video Technol..

[3]  Chong-Wah Ngo,et al.  Motion characterization by temporal slices analysis , 2000, Proceedings IEEE Conference on Computer Vision and Pattern Recognition. CVPR 2000 (Cat. No.PR00662).

[4]  Xiang Cao,et al.  Video shot motion characterization based on hierarchical overlapped growing neural gas networks , 2003, Multimedia Systems.

[5]  Jinwoong Kim,et al.  Threshold‐Based Camera Motion Characterization of MPEG Video , 2004 .

[6]  Bernd Freisleben,et al.  Estimation of arbitrary camera motion in MPEG videos , 2004, Proceedings of the 17th International Conference on Pattern Recognition, 2004. ICPR 2004..

[7]  José Ignacio Benavides Benítez,et al.  Global motion estimation algorithm for video segmentation , 2003, Visual Communications and Image Processing.

[8]  Euripides G. M. Petrakis,et al.  Adaptive Methods for Motion Characterization and Segmentation of MPEG Compressed Frame Sequences , 2004, ICIAR.

[9]  Jenny Benois-Pineau,et al.  Shot Boundary Detection In The Framework of Rough Indexing Paradigm , 2004, TRECVID.

[10]  Jean-Marc Odobez,et al.  Robust Multiresolution Estimation of Parametric Motion Models , 1995, J. Vis. Commun. Image Represent..