A JPEG 2000 Codec Primer, Part II

To meet the increasing demand for rapid image processing, the JPEG 2000 standard can provide a high-performance codec for easyintegration into software products.

By Stewart N. Taylor

Today's highly competitive business marketplace is placing increasing demands on fast and reliable image processing software and codec solutions for industrial and scientific imaging applications. Part I of this article (published in the January 2004 edition of Advanced Imaging) examined JPEG image processing codec implementations. Part II presents illustrative samples for implementing the high-performance JPEG 2000 standard, with emphasis on speed of development and ease of integration into software products.

Research on image compression has evolved steadily in the decade since the JPEG standard was introduced, and hardware constraints have loosened substantially during that period. Therefore, after a long process of competitive submissions, the Joint Photographic Experts Group (JPEG) has settled on a definition for the next generation of JPEG, called JPEG 2000. Despite its name, however, JPEG 2000 is not yet enjoying the same wide acceptance, or even the awareness, of its predecessor.

JPEG 2000 fundamentally follows the same coding paradigm of JPEG?transform, quantize and variable-length code?but nearly every element of its codec differs in some way. For example, the variable-length code is radically different and more closely resembles the bit-by-bit progressive JPEG method.