Previous abstract Next abstract

Session 12 - Image Processing.
Oral session, Monday, June 30
Ballroom A, Chair: Stephen Kahler

[12.01] Image Compression Aboard the LASCO-EIT/SOHO Coronagraphs

R. A. Howard (Naval Research Lab), D. Wang (Interferometrics/Naval Research Lab)

The rate at which data from a space-borne instrument are transmitted to the ground is frequently much lower than the ability of an instrument to generate the data. Each of the four 1024 x 1024 pixel CCD cameras on the LASCO and EIT instruments on-board the SOHO satellite are capable of being read out to the main electronics box at the rate of 800,000 bits per second, whereas the rate at which these data are transmitted to the ground is only 5200 bits per second. At this rate, a full image is transmitted to the ground in about 54 minutes. Thus, the ability to reduce the amount of bits that are transmitted to the ground is essential to the nominal operation of LASCO/EIT which share a common electronics package. The LASCO/EIT software includes a number of data compression modules that can be commanded to perform various types of compression. These compression modules can be divided into two classes, lossless and lossy, depending upon whether or not the reconstructed image is an exact replication of the input image. The lossless schemes are no compression and the Rice algorithm, which belongs to the general class of entropy encoding schemes. With the Rice algorithm, the LASCO/EIT images can be compressed about a factor of 2 compared to no compression. The lossy schemes all involve some loss of image information and they can be divided into several classes: geometric, intensity scaling , summing or differencing of multiple images or transform encoding. The geometric schemes are to transmit pixels that are within the field of view or (2) are within a subblock or (3) transformation to polar coordinates. The intensity scaling includes "divide by 2" or square root compression. Summing and differencing of multiple images reduce the number of bits transmitted at the expense of processing time. The transform encoding schemes are an adaptive discrete cosine transform (ADCT) and a wavelet transform, both of which enable compressions of almost any factor, with a corresponding loss of information. The wavelet compression is a new capability for LASCO/EIT. The relative performance of the transform encoding schemes on the LASCO/EIT images will be discussed in more detail.

If you would like more information about this abstract, please follow the link to This link was provided by the author. When you follow it, you will leave the the Web space for this meeting; to return, you should use the Back button on your browser.

The author(s) of this abstract have provided an email address for comments about the abstract:

Program listing for Monday