MPEG-4 Video Retrieval using Video-Objects and Edge Potential Functions

A novel video retrieval tool based on MPEG-4 video object (VO) representation is presented. The algorithm extends the concept of edge potential functions (EPF), already used in shape-based image retrieval, tailored to work on shapes extracted from VO planes defined in MPEG-4 syntax. First, key frames are selected from the VO sequence by detecting significant object deformations. Then, object boundaries are extracted from each key frame by a simple manipulation of the relevant object plane bitmap, and normalized. Finally a shape-EPF (S-EPF) is calculated from the normalized boundaries and used to perform the matching with user’s query. Experimental results demonstrate that the proposed algorithm is efficient and fast in indexing a video sequence according to the presence of specific video objects.