Video Clustering Using SuperHistograms in Large Archives

Methods for characterizing video segments and allowing fast search in large archives are becoming essential in the video information flood. In this paper, we present a method for characterizing and clustering video segments using cumulative color histogram. The underlying assumption is that a video segment has a consistent color palette, which can be derived as a family of merged individual shot histograms. These merged histograms (SuperHistograms) are clustered using a Nearest Neighbor-clustering algorithm. Given a query video, in order to find similar videos, the SuperHistogram of the video will be generated and compared to the centers of the Nearest Neighbor clusters. The video clips in the cluster with center nearest to the query, can be searched to find video clips most similar to the query video. This method can be used in a variety of applications that need video classification and retrieval methods such as video editing, video archival, digital libraries, consumer products, and web crawling.

[1]  Mohamed Abdel-Mottaleb,et al.  Hierarchical clustering algorithm for fast image retrieval , 1998, Electronic Imaging.

[2]  Mohamed Abdel-Mottaleb,et al.  CONIVAS: content-based image and video access system , 1997, MULTIMEDIA '96.

[3]  Belur V. Dasarathy,et al.  Nearest neighbor (NN) norms: NN pattern classification techniques , 1991 .

[4]  Michael T. Orchard,et al.  Color quantization of images , 1991, IEEE Trans. Signal Process..

[5]  Akio Nagasaka,et al.  Automatic Video Indexing and Full-Video Search for Object Appearances , 1991, VDB.

[6]  Takeo Kanade,et al.  Informedia Digital Video Library , 1995, CACM.

[7]  S. Panchanathan,et al.  Image Indexing Using Moments and Wavelets , 1996, 1996. Digest of Technical Papers., International Conference on Consumer Electronics.

[8]  Nevenka Dimitrova,et al.  Color superhistograms for video representation , 1999, Proceedings 1999 International Conference on Image Processing (Cat. 99CH36348).

[9]  Shih-Fu Chang,et al.  VideoQ: an automated content based video search system using visual cues , 1997, MULTIMEDIA '97.

[10]  Eli Upfal,et al.  Updates to the QBIC system , 1997, Electronic Imaging.