基于Contourlet变换的SAR图像压缩算法

SAR Image Compress Algorithm Based on Contourlet Transform

针对小波变换不能充分表示图像的方向性信息,以及Contourlet变换过采样的问题,提出小波与Contourlet相结合的变换对SAR图像进行稀疏表示。分析了Contourlet变换系数的特点。采用固定阈值对小系数进行了剔除。最后结合游程编码对SAR图像进行压缩。实验表明,此方法较小波变换和Contourlet变换的SAR图像压缩方法有更好的效果。

Wavelet transform lack of directional information and Contourlet transform has the problem of over sampling. To overcome the deficiency above, a new algorithm which combines wavelet and Contourlet is proposed to represent the SAR image sparsely. Firstly the characteristics of coefficients are analyzed and then the run-length coding is applied to compress the SAR image after removing the low coefficients by hard threshold. The experiments demonstrate that the method performs much better than wavelet and Contourlet transform.