Towards Automatic Extraction and Modeling of Objects from Image Sequences

Abstract

. This paper exploits a simple but general technique to extract object models from arbitrary image sequences. Such object models can be used to structure and index the image sequence. The algorithm extracts and tracks homogenous regions, which may correspond to objects or object parts. By grouping similar moving regions the algorithm constructs models of potential objects. As such, the approach is model-free in the sense that it does not use a priori models to detect, track and segment objects. On the contrary, the ultimate goal of the approach is to build such models automatically from image sequences. 1 Introduction Many algorithms exist for the recognition of known and a priori modeled objects. A priori object models also facilitate the detection and segmentation of objects. However, it is unclear how to automatically extract such object models from a camera mounted on a robot moving around in a dynamic environment. This paper exploits the possibility to detect, segment, track and...