Fast DSP implementation of JPEG2000

Discrete wavelet transform (DWT) and the embedded block coding with optimized truncation (EBCOT) are two major parts of computational load in JPEG2000, the latest image-coding standard. In this paper, we propose two efficient methods to reduce the processing time of these parts. First, the overlapped block-based lifting scheme reduces the processing time of DWT by increasing a cache hit-rate to the utmost. The overlapped block-based lifting scheme partitions an entire image into blocks to fit into the cache size and reorders the sequence of the wavelet lifting. Second, we propose the parallel-pass architecture to improve tier 1 coding of EBCOT. This method processes the three coding passes of the same bit-plane in parallel.

[1]  Christopher Brooks,et al.  Cache-efficient wavelet lifting in JPEG 2000 , 2002, Proceedings. IEEE International Conference on Multimedia and Expo.

[2]  Michael W. Marcellin,et al.  JPEG2000 - image compression fundamentals, standards and practice , 2002, The Kluwer International Series in Engineering and Computer Science.

[3]  David S. Taubman,et al.  High performance scalable image compression with EBCOT , 1999, Proceedings 1999 International Conference on Image Processing (Cat. 99CH36348).

[4]  Andreas Uhl,et al.  Cache issues with JPEG2000 wavelet lifting , 2002, IS&T/SPIE Electronic Imaging.

[5]  Majid Rabbani,et al.  An overview of the JPEG 2000 still image compression standard , 2002, Signal Process. Image Commun..

[6]  I. Daubechies,et al.  Factoring wavelet transforms into lifting steps , 1998 .

[7]  Liang-Gee Chen,et al.  Analysis and architecture design of EBCOT for JPEG-2000 , 2001, ISCAS 2001. The 2001 IEEE International Symposium on Circuits and Systems (Cat. No.01CH37196).