一种基于小波零树编码和K-mean聚类的图像压缩的实现

摘要

针对在低比特率时图像存在严重的方块效应等不足，本文提出了一种基于小波零树编码和 K-mean 聚类的图像压缩算法。其中基于小波零树的图像编码克服了图像JPEG标准压缩算法在统计值较少时的方块效应现象，通过K-mean聚类算法可以改善小波零树编码算法中没有充分考虑到小波系数同一子带中相邻元素之间相关性的缺陷。实验结果表明，此改进的图像编码算法具有较好的信噪比和压缩比。%This paper presents an image coding algorithm based on wavelet transform and K-mean clustering.Image coding using wavelet coefficients are compressed by the JPEG,overcoming the blocking effect that the compression algorithm is serious in low bit rate conditions.Through the K-mean clustering,it overcomes the defects of algorithm without considering the image wavelet coefficient in the same sub band correlation between adjacent elements.The ex-perimental results show that it is good at PSNR and compression ratio.