An object-oriented framework for static and dynamic object recognition

A framework based on the object-oriented paradigm is presented. In the framework, object form and motion are treated in a uniform way and integrated in the reasoning scheme for the object recognition tasks. A modified object model is proposed to support the conceptual focus in the recognition task and to allow efficient information retrieval. Both spatial data and alphanumeric information are integrated in a consistent organization in order to facilitate object recognition. As each model is an individual entity which can convey its ability to match a spatial entity which was extracted from the image data, several of them could be active in the same time.<
>