Abstract: An image compression method based on a new class of wavelet-nonseparable wavelet is proposed. Some wavelet filter banks with different characteristics (such as orthogonality, symmetric, vanishing moment) are constructed. Using these filters, an image compression method in non-separable wavelet-domain is implemented; the comparative studies of the application of the non-separable wavelet transform and the tensor product wavelet transform in case of different characteristics are proposed. Experimental results show that non-separable wavelet is a class of multi-dimensional wavelet with good application potentials, which has more advantages compared with wavelet and multi-wavelet.

Key words: Non-separable wavelet, Tensor-product wavelet, Image compression

摘要： 提出了一种基于非分离小波的图像压缩算法，构造了不同性质(正交性，对称性，消失矩等)的小波滤波器，利用这些滤波器实现了非分离小波域的图像压缩，比较研究了在采用不同性质的小波函数的情况下非分离小波变换和张量积小波变换在图像压缩中的应用。从实验结果可以看出，非分离小波在性质接近时，在很多情况下要优于单小波与多小波，是一类具有良好应用前景的高维小波。

关键词: 非分离小波, 张量积小波, 图像压缩

CLC Number: 

• TP391.41