Variation in object shape is an important visual cue for deformable object recognition and classification. In this paper, we present an approach to model gradual changes in the 2-D shape of an object. We represent 2-D region shape in terms of the spatial frequency content of the region contour using Fourier coefficients. The temporal changes in these coefficients are used as the temporal signatures of the shape changes. Specifically, we use autoregressive model of the coefficient series. We demonstrate the efficacy of the model on several applications. First, we use the model parameters as discriminating features for object recognition and classification. Second, we show the use of the model for synthesis of dynamic shape using the model learned from a given image sequence. Third, we show that, with its capability of predicting shape, the model can be used to predict contours of moving regions which can be used as initial estimates for the contour based tracking methods.
|Original language||English (US)|
|Journal||Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition|
|State||Published - Oct 19 2004|
|Event||Proceedings of the 2004 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, CVPR 2004 - Washington, DC, United States|
Duration: Jun 27 2004 → Jul 2 2004
ASJC Scopus subject areas
- Computer Vision and Pattern Recognition