Friday, September 19, 2008

week 5

Wikipedia article on Data Compression
-lossless compression vs lossy compression (when the file remains the same vs when the file is simplified to use even less space)
-LZR (zip), DEFLATE (PNG), LZW (gif)

data compression basics
-this article discusses the problems of data compression in terms of nonrepeated characters
-there is a minor problem with having a character that both signifies a beginning/end to compression AND appears in the data itself
-RLE requires three characters in a row to be useful => useful in images, used in faxes
-LZ works by referencing back to something already used (i.e. if it's used once, it isn't used again)
-entropy coding - the words get numbers, so a common word will get a low number (e.g. the = 1)
-lossy compression preserves the interpretation of data if not the exact data itself (i.e. taking This article a smooth grayscale and making it less smooth)
-the loss refers to mathematical data rather than perceivable data, and the process isn't reversable

Imaging Pittsburgh
-This article describes the difficulties and some of the logistics of setting up a digital collections of photos of Pittsburgh.

YouTube and libraries
-Gives a basic intro to youtube to someone who has never heard of it/used it. Then details the ways in which youtube could be used in a library (e.g. taping classes and putting them online so people could learn on their own time)

No comments: