嵌入式零树小波在合成孔径声纳图像压缩中的应用

合成孔径声纳（SAS）图像具有分辨率高、数据量大特点,给数据存储和传输带来困难。由于合成孔径图像像素间具有一定的相关性,采用嵌入式零树小波（EZW）编码算法,利用小波系数的相关特性可以达到压缩图像的目的。文章研究了基于嵌入式零树小波的合成孔径声纳图像的压缩和解码算法,并用合成孔径声纳图像对算法进行了测试,采用峰值信噪比、均方误差、压缩比三项技术指标对图像压缩质量进行了评估。结果验证了基于嵌入式小波零树图像编码算法的合成孔径图像的压缩和解码算法可行性。

Application of Embedded Zero-tree Wavelet in SAS Image Compression

Synthetic aperture sonar （SAS） image is characterized as high resolution and large amount of data, which bring difficulty in data saving and transmission. Because the pixels of SAS image have certain correlation, embedded zero-tree wavelet algorithm can make good use of correlation among the wavelet coefficients to realize the image compression. This paper studies the algorithms of compression and decompression of SAS image based on EZW algorithm. The algorithms are tested by a SAS image, and the performances of the algorithms are analyzed by peak signal to noise ration （PSNR）, mean square error （MSE） and compression ratio （CR）. The result shows that the compression and decompression algorithms based on EZW algorithm are feasible for SAS image.