A Wavelet-Based Preprocessing for Moving Object Segmentation in Video Sequences
A simple preprocessing method for extracting boundary regions of moving objects in a video sequence is presented. We use Chui’s overssampled shift-invariant wavelet transform and the multiresolution motion estimation and compensation in the wavelet domain. Dominant prediction errors often appear along the boundary of a moving object. Our algorithm is developed to detect boundary regions at a coarse scale by utilizing the prediction error information provided in all subband images at the coarse resolution. This is taken as our first step toward the video object segmentation for use in the wavelet-based MPEG-4.
KeywordsPrediction Error Video Sequence Motion Estimation Object Boundary Current Frame
Unable to display preview. Download preview PDF.
- 5.C. Chui, X. Shi, and A. Chan “An oversampled frame algorithm for real-time implementation and applications” Proc. SPIE Conf. On Wavelet Applications”, Orland, FL, April 1994, Vol. 2242, pp 272–301.Google Scholar
- 6.L. Zheng, J. C. Liu, A.K. Chan, W. Smith “Object-Based Image Segmentation Using DWT/RDWT Multiresolution Markov Random Field” Proc. IEEE International Conf. On Acoustics, Speech, and Signal Processing, Phoenix, AZ, March 1999, Vol. 6, pp. 3485–3488.Google Scholar
- 7.I. Kompatsiaris, and M. G. Strintzis “Spatiotemporal Segmentation and Tracking of Objects for Visualization of Videoconference Image Sequences” IEEE Trans. On Circuits and Systems for Video Technology, Vol. 10, pp. 1388–1402, Dec, 2000.Google Scholar
- 9.M. Bagci, I. Yilmaz, M.H. Karci, T. Kolcak, U. Orguner, Y. Yardimci, M. Demirekler, and A.E. Cetin “Moving Object Detection and Tracking in Video Based on Higher Order Statistics and Kalman Filtering” Proc. (CDROM) 2001 IEEE-EURASIP workshop on Nonlinear Signal and Image Processing, Baltimore, MD, June 2001.Google Scholar