Video compression technology works by leaving out information. Only two classifications of information can be left out without visible loss: information one can predict about the scene and information that the viewer cannot see. The first category involves a study of scene statistics and the second requires a complete understanding of visual perception. The estimated compression ratio due to the latter is greater than 56. Further compression can be achieved based on scene predictability.
About 99% of the bits in an 8-bit encoded RGB moving image can be left out without noticeable difference in the image.
Dr. William E. Glenn, Director, Space Communications Technology Center, Florida Atlantic University. JPL SATCOM Quarterly, No. 7, October 1992, pp. 12-14.
|
The Columbus Optical SETI Observatory |
||
| Copyright ©, 1990-2006 | Personal
Web Site: www.stuartkingsley.com |
Last
modified:
10/28/06 Contact Info |