Use of Grid Computing Technology and JPEG2000 Standard for Lossless Gigabyte Image Compression

Image processing domain requires high calculation capacities and storage facilities especially when we manipulating huge images like ultrasonic and satellite ones. In fact, manipulating such images should be done without losing any kind of information. For this purpose, those images are generally used and manipulated without compression because actual compression algorithms do not offer the possibility of giving a good lossless compression ratio to override their slowliness. Lately, a new compression Standard, namely JPEG2000, was introduced. This standard offers, in addition to many interesting characteristics, the possibility of lossless image compression, under the same core and with extremely competitive compression ratio. In this paper we will describe the gridification and implementation of the JEPG2000 standard on the EUMEDGRID infrastructure. This gridified version of the Jasper implementation of the JPEG2000 standard follows three steps: each Worker Node extracts, remotely from the Storage Element (SE) based on Grid File Access Library routines (GFAL), a portion of the considered image, compresses it and saves the result back into the SE. A merge task is performed after the gathering of all image portions. We will demonstrate that, this gridified algorithm, not only, compresses gigabyte images successfully but also efficiently and in a highly competitive processing time.