Content area
Full Text
A little experience wandering the World Wide Web reveals that storing a digitized .image, especially a high-resolution one, requires lots of memory. Viewing some homepages, for instance, is a painfully slow process even with a high-speed connection: While image files pour avalanches of bytes into your computer's memory, the page grows line by line on your monitor. Image storage and reconstruction may prove even more troublesome in the CD-ROM business, where someone might want to put thousands of images on a disk. Microsoft's Encarta CD-ROM encyclopedia, in fact, contains about 7,000 photographs. Getting them all on one disk required a technique called fractal image compression.
Fractal image compression is one of the newer ways of squeezing an image file by systematically analyzing it and saving-instead of a description of every pixel-a smaller set of instructions that can be used to precisely reconstruct it. Yuval Fisher, visiting researcher at the Institute for Nonlinear Science at the University of California at San Diego, says, "Image-compression methods eliminate redundant information from a representation of an image.
Roughly speaking, fractal compression finds selfsimilarity at different scales, and eliminates repeated descriptions. For example, the tip of an elbow may closely resemble the tip of a nose, and it is sufficient to specify just one of them along...