NEXT | BACK | INDEX

WHY HUFFMAN ENCODING ?
  • No Patent Problems: Dictionary compression and arithmetic compression are both areas of legal problems over patents.

  • Fast: Static Huffman compression is much faster for both coding and decoding than arithmetic compression. (See "Lossless Compression for Text and Images", A Moffat, T C Bell, and I H Witten, in "Signal Compression: Coding of Speech, Audio, Text, Image, and Video", published by World Scientific Publishing, page 195-196, 1997.)

  • Practical: The compression algorithm comparisons shows that the Huffman based algorithms give good compression in a reason time.

  • Very Close to Optimum: Whilst Huffman encoding must assign an integer number of bits to every symbol, this is not a limitation to compression ratios in almost all practical cases.

  • Static Algorithm: The coding table can be produced once and used for many images with the same statistical characteristics, to avoid the need of calculating image statistics and symbol codes.

  • Entry Look-Up Table: Since very pixel is coded by 1 of more bits, if predictor 1 is used, a simple look-up table may be defined between pixel position in the array and bit position in the compressed data.