Data Compression


Criteria

Survey Formats

Basics

Compression Methods

Data Formats


Huffman Code

Example

Characteristics

Variants

Dynamic Huffman Code

Adaptive Huffman Code

Initialization

Standard Distribution

Uniform Distribution

Extension for New Symbols

Pros and Cons

Algorithm

Example


Glossary

Index


Download


www.BinaryEssence.com

Extension for New Symbols


For the initialization with standard or uniform distribution the entire set of symbols must be contained in the code tree even if these are not used within the original data. The introduction of a special code offers a solution. The code identifies a new symbol immediately following. Thus it is guaranteed that only symbols already encoded are part of the code tree.


Disadvantage of this variant is the increased coding effort for the first appearance of a symbol. This consists of the uncoded symbol and the Huffman code for the control character. Caused by the structure of the Huffman tree the control character requires a proportionally large code length.


Initially the Huffman tree only consists of one single node which forms the root of the tree and represents the control character. With each additional symbol the number of the leaf nodes grows at one.


Example:



 <   ^   > 

Initialization Initialization with Uniform Distribution Pros and Cons