DefinePK

Abstract

Fractal audio compression (FAC) is based on the concept of partitioning audio data, and then exploiting the self-similarity present in audio signal in order to eliminate the redundancy may exist in audio signal. This elimination plays a significant role in audio compression. The key step for fractal coding is to pick up the domain blocks which are similar to the original blocks (i.e., range blocks), and these range blocks are represented as a set of affine transformations [1]. Fractal compression has a significant draw back especially while coding large range blocks since the level of error produced is very high and unaccepted. In order to improve performance in terms of compression gain and fidelity error of the fractal coding especially in case of using long range blocks, the wavelet method is utilized. The tests results indicate that a significant improvement in the PSNR values occurs; making a significant decrease in the compression error level after the addition of wavelet method. Moreover, the attained results of the conducted tests when using the wavelet transform for multiple passes and rounds implies a significant increase in the compression ratio values as the block length increase, while the generated PSNR carries good and acceptable values which reaches 30db.