Monday, March 24, 2003
( 8:37 AM ) Matt
So, I was a little misleading when I said that storing all of the wavelet moments of a 1MP image would take up 100 times as much space as the uncompressed image. The number of coefficients needed to represent an image is the same as the number of values needed by an uncompressed image. So, you're probably wondering where that factor of 100 came from. It's just that adjacent wavelet moments share a lot of coefficients.
What we need is some very efficient storage method. I'm afraid that standard compression techniques like LZ77 and Huffman coding won't be effective on this database, though it is worth a try.# -
Comments: Post a Comment