Image storage optimization and feature calculation on Netezza Database system

Abstract

The storage of images in databases has long been a delicate matter but comes now more and more in use. This kind of data is in most DBMS limited in storage size to 64kb, making necessary to divide large objects into fragments; moreover, programming languages also present limitations in data retrieval. In this paper, we experimented the retrieval, reconstitution, feature calculation of pictures using different fragment size and determined the optimal one. In this work, the server used was a Netezza model from the N2001 line. In this work, we present the architecture of the database, our research and bring recommendation about the optimal fragmentation size to store images into a database.