Huffman coding

Definition: Data compression technology in which the amount of data used to code a symbol or piece of information varies with the probability of its occurence. * The algorithm reduces frequently occurring strings of data to shorter codes in order to save on memory, hence also known as Code Length Optimisation. For example, character strings like '-ping' and '-ation' occur commonly in English; space savings are made by replacing them with codes e.g. 'pg' and 'tn'. * Information with a high probability of appearing are coded with fewer bits of data, information with low probability are coded with more bits of data. The result is more efficient coding of data than if all codes were the same length. * Described by D A Huffmann in 1952. * Used in JPEG compression.

Previous Term: hue  Next Term: Hurter-Driffield curve