一种基于小波系数重要图编码的图像压缩算法

提出了一种新的基于小波系数重要图编码的图像压缩算法。该算法根据量化后的波系数的特点进行了一种期望排序，然后舍掉序列后面大量的零值小波系数，从而得到一个波系数子集，能以少的小波系数来很好地逼近原始图像，省去了零树编码中零树结构带来的大量比特开销。实验表明，该算法与MPEG-4的静止图像压缩算法相比较，重构图像的峰值信噪比（PSNR）值在相同码率下有较大的提高。

An image compression algorithm based on significant image of wavelet coefficients

In this paper,a new image compression algorithm based on significant-image coding of wavelet coefficients is put forward.Make the quantized wavelet coefficients into expectant sequence.Delete many latter ze-ro coefficients of the sequence and gain a subclass that can use the fewer coefficients to approach the original image.In this way,compared to zero-tree coding,the overhead bits of the zero-tree structure is saved.The peak signal-to-noise ratio(PSNR)of the reconstructed image is improved compared to still image compression algorithm of MPEG-4at the same bit ratio.