5th International Mobile Multimedia Communications Conference

Research Article

Location coding for mobile image retrieval

Download672 downloads
        author={Sam S.  Tsai and David  Chen and Gabriel  Takacs and Vijay  Chandrasekhar and Jatinder P.  Singh and Bernd  Girod},
        title={Location coding for mobile image retrieval},
        proceedings={5th International Mobile Multimedia Communications Conference},
  • Sam S. Tsai
    David Chen
    Gabriel Takacs
    Vijay Chandrasekhar
    Jatinder P. Singh
    Bernd Girod
    Year: 2010
    Location coding for mobile image retrieval
    DOI: 10.4108/ICST.MOBIMEDIA2009.7406
Sam S. Tsai1,*, David Chen1,*, Gabriel Takacs1,*, Vijay Chandrasekhar1,*, Jatinder P. Singh2,*, Bernd Girod1,*
  • 1: Information Systems Laboratory, Stanford University, Stanford, CA 94305, U.S.A.
  • 2: Deutsche Telekom Inc. R&D Lab, Los Altos, CA 94022, U.S.A.
*Contact email: sstsai@stanford.edu, dmchen@stanford.edu, gtakacs@stanford.edu, vijayc@stanford.edu, jatinder.singh@telekom.de, bgirod@stanford.edu


For mobile image retrieval, efficient data transmission can be achieved by sending only the query features. Each query feature is composed of a descriptor and a location in the image. The former is used to find candidate matching images using a "bag-of-words" approach while the latter is used in a geometric consistency check to map features in the query image to corresponding features in the database image.

We investigate how to compress the location information and how lossy compression affects the geometric consistency check. The location information is converted into a location histogram and a context-based arithmetic coding with location refinement method is then proposed to code the histogram. The effects of lossily compressing the location information are evaluated empirically in terms of the errors in corresponding features and the error of the estimated geometric transformation model. From our experiments, rates at ~5.1 bits per feature can achieve errors comparable to lossless coding. The proposed scheme achieves a 12.5x rate reduction compared to the floating point representation, and 2.8x rate reduction compared to a fixed point representation.