基于S+P变换和EZW编码的图像压缩算法

Image Compression Algorithm Based on S + P Transform and EZW Coding

嵌入式零树小波是根据标准小波变换的空间多分辨率特征将重要节点与零树根进行编码来完成图像压缩的编码算法,但标准小波变换对图像的边缘、轮廓以及纹理等特征信息的表达不具有稀疏性,不能够较好地表达图像的细节信息。针对标准小波变换的缺陷,新算法根据S+P变换能够实现整数到整数的变换,将EZW和S+P变换结合起来,提出了一种基于S+P变换改进的EZW编码算法,即使用S+P变换替换标准小波变换,完成图像的低损耗压缩。实验结果表明:相对于经典的EZW编码算法,基于S+P变换改进的EZW编码算法在压缩比、编码效率以及峰值信噪比等性能方面都具有较大的改善。

The embedded zero tree wavelet （EZW） coding algorithm applies space multi-resolution characteristic of standard wavelet transform is made full use, and compresses important nodes and zero tree roots by eneode. However, the expression of characteristic information such as edge, contour and texture of image hy standard wavelet transform is not an optimal sparse expression. Transforming from integers to integers and lossless image compression can be realized by S ＋ P transform. In view of the defects of standard wavelet transform, an improved EZW coding algorithm is proposed based on the S ＋ P transform. That is, the standard wavelet transform is replaced by S ＋ P transform. The experimental results show that compared with the classic EZW coding algorithm, EZW eoding algorithm based on the S ＋ P transform is more superior in compression ratio, coding efficiency and peak signal to noise ratio.