Compression Algorithm Based on Multiresolution Wavelet in Embedded Systems

Embedded systematic information memory is restricted by space that the image information has to be compressed. In the traditional compressional algorithm of wavelet transform, the image information is produced by Multi-resolution Analysis (MRA), and rebuilt directly after removing one or several layers of high-frequency information. The algorithm is easy-going and convenient, but it is easy to lose some important high-frequency information. This paper describes the embedded zero tree wavelet algorithm (EZW) based on the traditional algorithm. We initialize the fluctuating threshold which is regarded as the criterion of discarding the wavelet coefficients and keeping the important high frequency components, and keep high quality of compression ratio of the information for carrying out the image information compression.