摘要
A new fractal image compression algorithm based on high frequency energy (HFE) partitioning andmatched domain block searching is presented to code synthetic aperture radar (SAR) imagery. In the hybridcoding algorithm, the original SAR image is decomposed to low frequency components and high frequencycomponents by wavelet transform (WT). Then the coder uses HFE of block to partition and searchthe matched domain block for each range block to code the low frequency components. For the high frequencycomponents, a modified embedded zero-tree wavelet coding algorithm is applied. Experiment resultsshow that the proposed coder obtains about 0. 3dB gain when compared to the traditional fractal coderbased on the quadtree partition. Moreover, the subjective visual quality of the reconstructed SAR imageof the proposed coder outperforms that of the traditional fractal coders in the same compression ratio(CR).
A new fractal image compression algorithm based on high frequency energy （HFE） partitioning and matched domain block searching is presented to code synthetic aperture radar （SAR） imagery. In the hybrid coding algorithm, the original SAR image is decomposed to low frequency components and high frequency components by wavelet transform （WT）. Then the coder uses HFE of block to partition and search the matched domain block for each range block to code the low frequency components. For the high frequency components, a modified embedded zero-tree wavelet coding algorithm is applied. Experiment results show that the proposed coder obtains about 0.3dB gain when compared to the traditional fractal coder based on the quadtree partition. Moreover, the subjective visual quality of the reconstructed SAR image of the proposed coder outperforms that of the traditional fractal coders in the same compression ratio （CR）.
基金
Supported by the National Natural Science Foundation of China (No. 90304003)
the President Fund of GUCAS (No. O85101HM03).
