Compressed Beef Marbling Image Database with Browser-based Retrieval System

The system we are proposing in this paper demonstrates the capability of a robust multi-user access database management system by which compressed beef carcass rib-eye images, together with other related information, can be easily accessed on the Internet through web browsers in a client-server distributed fashion. This framework, we believe, can pioneer a new direction in information reuse and sharing among different users of the Japanese livestock industry that is non-existent at present. The compression technique we employ has its basis in a class of functions which we termed Fluency functions that enables highly efficient storage and retrieval of image data over the Internet with limited bandwidth. Eventually, we envision that the characteristics of the compressed data could provide insights into creating efficient image matching and retrieval algorithms that constitute a layer above the functions being provided by current database management systems. Such research, we believe, could contribute to a wider range of application domains involving images such as cartographic data and medical images, etc.

[1]  Massimo Maresca,et al.  Efficiency of Standard Software Architectures for JAVA-Based Access to Remote Databases , 1998, HPCN Europe.

[2]  Nobuyuki Otsu,et al.  Feature extraction algorithm for beef marbling , 1999, 1999 IEEE Pacific Rim Conference on Communications, Computers and Signal Processing (PACRIM 1999). Conference Proceedings (Cat. No.99CH36368).

[3]  Koji Nakamura,et al.  Practical compactly supported sampling functions of degree 2 , 1999, 1999 IEEE Pacific Rim Conference on Communications, Computers and Signal Processing (PACRIM 1999). Conference Proceedings (Cat. No.99CH36368).

[4]  N. Otsu A threshold selection method from gray level histograms , 1979 .