Image Background Search: Combining Object
Detection Techniques into ContentBased
Similarity Image Retrieval(CBSIR) Systems

Rohini Srihari, Zhongfei Zhang, Aibing Rao
CEDAR, SUNY At Buffalo
rohini, zhongfei, arao@cedar.buffalo.edu
April 15, 1999

Abstract A framework of combining object detection techniques with a CBSIR
system is discussed. As an example of the framework, a special CBSIR system which
decides the similarity of images based on background features is presented and back
ground subimages can be extracted by using a face detector. This background search
can be used as a classification tool for pictures of home albums.

References
[1] http://quark.ifp.uiuc.edu:8080/index.html.
[2] http://www.ctr.columbia.edu/ jrsmith.
[3] http://www.excalib.com/demos.html.
[4] http://wwwqbic.almaden.ibm.com.
[5] http://www.virage.com.
[6] J. Hafner W. Niblack R. Barber, M. Flickner and D. Petkovic. Efficient and effective
querying by image content. Journal of Intelligent Information Systems, 3:231--262, 1994.
[7] J. R. Smith. Integrated Spatial and Feature Image Systems: Retrieval, Analysis a nd
Compression. PhD thesis, Columbia Univ., 1997.
[8] Rohini K. Srihari and Zhongfei Zhang. Finding pictures in context. Proc. of IAPR In
ternational Workshop on Multimedia Information Analysis & Retrieval, SpringerVerlag
Press, pages 109 -- 123, 1998.
