Towards automatic annotation of video documents

This paper presents a video indexing tool for sequences which contain moving persons, using a model-based dynamic scene analysis. The authors show how scene-specific knowledge can be used to guide the analysis process. The authors´ approach permits one to avoid some problems inherent in the segmentation of primitives. A model-based post-processing step is used to correct a posteriori the shape of the extracted objects. A scenario, describing the sequence in terms of basic events is proposed. These events constitute a first level of annotation and experiments show a good robustness regarding occlusions problems