VST-model: a uniform topological modeling of the visual-spatio-temporal video features

Recently, two approaches investigated indexing and retrieving video data. One approach utilized the visual features, and the other approach exploited the spatiotemporal features. In this paper, we present a new Visual-Spatio-Temporal model (VST model), to represent videos. The visual features and the spatio-temporal features of video objects are integrated and represented uniformally by a topological model. As a result, rich set of new relationships are defined to support and simplify the formulation of more semantical queries. The expressiveness of the VST video model is validated by some query examples