Semantic Video Indexing

The Semantics-Based Video Indexing and Retrieval (SBVIR) software tools help users of video surveillance systems to easily conduct video-based forensics and image intelligence (IMINT) from surveillance video database. Based on advanced video analytics, SBVIR automatically generates metadata tags for archived surveillance video so that the user can rapidly execute intuitive semantic queries on large-scale video database.


Major capabilities of SBVIR system include:

  • Target detection
  • Multi-camera multi-target tracking
  • Hierarchical target classification
  • Automated activities recognition
  • Video query-by-keyword and query-by-example
  • Support relevance feedback