Motion Segmentation and Tracking by Embedding Global Model Within a Contextual Relaxation Scheme

In this paper, we integrate the model-based tracking and local contexture (temporal and spatial) relaxation scheme into a MAP framework to track non-rigid objects such as human hands or faces. By combining them together, our algorithm is much more accurate and robust. The global shape model enables us to represent the dynamic and kinematic constraints. It also helps us to get a better initial segmentation and hence greatly reduce the computation of contextual relaxation. The local contextual constraints help us to get a more accurate support map by considering the local information. Hence it will reduce the significance of the error in the global model. Our algorithm can work autonomously and no need to initialize it. It can detect new moving objects and track them with new blobs. It can also keep track of a stalled object because we also utilize the spatial-temporal constraints. Some promising experiment results are shown.

[1]  Nebojsa Jojic,et al.  Tracking articulated objects in dense disparity maps , 1999 .

[2]  H. Vincent Poor,et al.  An Introduction to Signal Detection and Estimation , 1994, Springer Texts in Electrical Engineering.

[3]  Nuno Vasconcelos,et al.  Empirical Bayesian EM-based motion segmentation , 1997, Proceedings of IEEE Computer Society Conference on Computer Vision and Pattern Recognition.

[4]  Alex Pentland,et al.  Pfinder: Real-Time Tracking of the Human Body , 1997, IEEE Trans. Pattern Anal. Mach. Intell..

[5]  Christoph Bregler,et al.  Learning and recognizing human dynamics in video sequences , 1997, Proceedings of IEEE Computer Society Conference on Computer Vision and Pattern Recognition.

[6]  P. R. Giaccone,et al.  Segmenting image sequences by embedding motion and colour cues within a contextual relaxation scheme , 1999 .

[7]  Ying Wu,et al.  Color tracking by transductive learning , 2000, Proceedings IEEE Conference on Computer Vision and Pattern Recognition. CVPR 2000 (Cat. No.PR00662).

[8]  Richard Szeliski,et al.  An integrated Bayesian approach to layer extraction from image sequences , 1999, Proceedings of the Seventh IEEE International Conference on Computer Vision.

[9]  C. Stiller,et al.  Estimating motion in image sequences , 1999, IEEE Signal Process. Mag..

[10]  J. Besag.  On the Statistical Analysis of Dirty Pictures , 1986 .

[11]  C. Stiller,et al.  Estimating Motion in Image Sequences A tutorial on modeling and computation of 2 D motion , 2022 .