Back to Volume
Paper: A Comparison of Lossless Image Compression Methods and the Effects of Noise
Volume: 411, Astronomical Data Analysis Software and Systems XVIII
Page: 25
Authors: Pence, W.D.; Seaman, R.; White, R.L.
Abstract: This paper compares two image compression methods and shows how they are affected by the amount of noise in the image. We use a simple procedure to measure the equivalent number of noise bits in an image, which sets an upper limit on the compression ratio. We compare the widely used GZIP compression program to a newer compression method that uses the Rice algorithm within the FITS tiled-image compression convention. When applied to a large sample of astronomical images, the Rice compression method produces 1.4 times better compression, has 2–3 times faster compression speed, and has about the same uncompression speed as GZIP. In order to promote wider use of this new compression method, our fpack and funpack image compression software is available from the HEASARC and NOAO websites.
Back to Volume