The International Archives of the Photogrammetry. Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part B4. Beijing 2008 
to use less space or to reduce transfer delays any more, but to or 
ganize the valuable information differently. The multiresolution 
structure on the new compression standards enables the fast an 
efficient generation of low resolution products for a limited com 
plexity. This new vision of compression advantages should lead 
to an increase use of compression directly within image process 
ing algorithms. 
REFERENCES 
GML in JPEG 2000 for Geographic Imagery (GMLJP2) Encod 
ing Specification, 2006. 
Taubman, D. S. and Marcellin, M. W., 2002. JPEG2000 Im 
age Compression Fundamentals, Standards and Practice. Kluwer 
Academic Publishers, Boston, MA. 
Thiebaut, C., Christophe, E., Lebedeff, D. and Latry, C., 2007. 
CNES studies of on-board compression for multispectral and hy- 
perspectral images. In: Satellite Data Compression, Communi 
cations, and Archiving III, Vol. 6683, SPIE. 
Binary stream 
Figure 3: Example of bitstream organization for the LRCP order. To obtain the first quality layer, one just has to read the beginning of 
the bitstream. However, to decode only one resolution level (Rq for example), or just a specific area of the image (Pi for example), the 
decoder have to retrieve information in different part of the bitstream. 
(a) Quality (b) Memory 
Figure 6: Impact of tile sizes on quality is neglectable (for realistic values). On the oposite, using tiles can greatly reduce the required 
memory for processing. 
1286