Video analysis and classification for MPEG-7 applications

Current information and communication technologies provide the infrastructure to send bits anywhere, but do not presume to handle information at the semantic level. Consequently, the huge amount of multimedia data imposes people a heavy burden of manipulating, including searching/choosing, interpreting, skimming, and integrating information. Thus automatic video content analysis, including video semantic category classification is necessary for various MPEG-7 applications, such as video content-based retrieval databases, on-line video indexing and filtering, video summaries for browsing systems and so on. We implemented and applied a set of simple visual/motional feature descriptors to our proposed video semantic category classification system for various applications.