next up previous index
Next: P4.2 An application of Up: Session P4. Data Analysis Previous: Session P4. Data Analysis   Author Index


P4.1 Adaptive, Lossless Compression of 2-D Astronomical Images

Chris N. Sabbey (Yale University)

We describe a lossless compression program developed for incorporation into the data acquisition system of a 16 CCD driftscan survey. The program is based on switched linear prediction with Rice coding of the high order bits of the prediction residuals. Several notable features of the program are its speed (> 5 MB/sec on a 300 MHz Pentium II), adaptivity to a variety of image types without any user-tuned parameters, and suitability for online work (it is a one-pass algorithm with minimal buffering requirements and a steady output stream).

The compression ratios and run times of the program are compared to other well-known compression programs using a variety of test images.



adass@ncsa.uiuc.edu