Content-based Video Clustering and Querying

The recent proliferation of multimedia resources has created a substantial demand for video editing, searching and browsing tools. In this project, we have developed a user-friendly tool for parsing and querying video. Our toolbox is integrated into the resolution independent video processing language Rivl, and builds on an algorithm that computes a short sequence of key frames (a so-called video skim) from a MPEG video. With our toolbox, the user can select one of the key frames in the skimmed sequence and query for similar key frames or ask for a condensed summary that eliminates redundant frames from a video skim. A second goal of this project was to develop techniques that allow the user to query for all key frames that contain a particular object. Our toolbox uses recently developed image comparison techniques like Color Coherence Vectors and motion tracking algorithms.

People

Materials

Comments? Interested in a demo ? Send us mail!

CS631 home page