| Summary: | Huffman encoding and arithmetic coding algorithms have shown great potential in the field of image compression. These algorithms are the origin of current image compression techniques. Nevertheless, there are some deficiencies in both algorithms that use the frequencies of the characters in the data. They aim to represent the symbols used in the data in the shortest bit sequence. However, they represent data that has a low frequency of use with very long bit sequences. The arithmetic coding algorithm was developed to address the shortcomings of the Huffman encoding algorithm. This paper proposes an efficient, alternative encoding algorithm that uses the Huffman encoding algorithm. The main objective of the proposed algorithm is to reduce the number of bits that are symbolized with long bit codewords by the Huffman encoding algorithm. Initially, the Huffman encoding algorithm is applied to the data. The characters that are represented by the short bit sequence from the Huffman encoding algorithm are ignored. Flag bits are then added according to whether the successive symbols are on the same leaf. If the next character is not on the same leaf, flag bit “0„ is added, otherwise flag bit “1„ is added between the characters. In other words, the key significance of this algorithm is that it uses the effective aspects of the Huffman encoding algorithm, and it also proposes a solution to long bit sequences that cannot be efficiently represented. Most importantly, the validity of the algorithm is meticulously evaluated with three different groups of images. Randomly selected images from the USC-SIPI and STARE databases, and randomly selected standard images on internet, are used. The algorithm encodes compressing operations for images successfully. Some images that have a balanced tree structure have yielded close results compared to other algorithms. However, when the total results are inspected, the proposed encoding algorithm achieved excellent results.
|
|---|
