TY - GEN
T1 - Lossless compression using joint predictor for astronomical images
AU - Wu, Bo Zong
AU - Tang, Angela Chih Wei
PY - 2009
Y1 - 2009
N2 - Downloading astronomical images through Internet is a slow operation due to their huge size. Although several lossless image coding standards that have good performance have been developed in the past years, none of them are specifically designed for astronomical data. Motivated by this, this paper proposes a lossless coding scheme for astronomical image compressions. We design a joint predictor which combines the interpolation predictor and partial MMSE predictor. Such strategy benefits from its high compression ratio and low computation complexity. Moreover, the scalable and embedding functions can be further supported. The interpolation predictor is realized by upsampling the downsampled input image using bi-cubic interpolation, while the partial minimum mean square error (MMSE) predictor predicts the background and foreground (i.e., stars) separately. Finally, we design a simplified Tier-1 coder from JPEG2000 for entropy coding. Our experimental results show that the proposed encoder can achieve higher compression ratio than JPEG2000 and JPEG-LS.
AB - Downloading astronomical images through Internet is a slow operation due to their huge size. Although several lossless image coding standards that have good performance have been developed in the past years, none of them are specifically designed for astronomical data. Motivated by this, this paper proposes a lossless coding scheme for astronomical image compressions. We design a joint predictor which combines the interpolation predictor and partial MMSE predictor. Such strategy benefits from its high compression ratio and low computation complexity. Moreover, the scalable and embedding functions can be further supported. The interpolation predictor is realized by upsampling the downsampled input image using bi-cubic interpolation, while the partial minimum mean square error (MMSE) predictor predicts the background and foreground (i.e., stars) separately. Finally, we design a simplified Tier-1 coder from JPEG2000 for entropy coding. Our experimental results show that the proposed encoder can achieve higher compression ratio than JPEG2000 and JPEG-LS.
UR - http://www.scopus.com/inward/record.url?scp=72449208637&partnerID=8YFLogxK
U2 - 10.1007/978-3-642-10520-3_25
DO - 10.1007/978-3-642-10520-3_25
M3 - 會議論文篇章
AN - SCOPUS:72449208637
SN - 364210519X
SN - 9783642105197
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 274
EP - 282
BT - Advances in Visual Computing - 5th International Symposium, ISVC 2009, Proceedings
T2 - 5th International Symposium on Advances in Visual Computing, ISVC 2009
Y2 - 30 November 2009 through 2 December 2009
ER -