易于硬件实现的集合树块图像压缩编码

Image compression algorithm of set-tree block coding for hardware implementation

提出一种易于硬件实现的集合树块图像压缩算法,将小波系数按照零树的方法排列为树集合,并将树集合排列为集合树块,利用分辨率之间的相似性和相邻系数之间的相关性进行编码,提高数据压缩效率;树块之间独立进行编码,以提高系统抗误码能力。对算法进行了系统仿真,结果表明本算法压缩效率与JPEG2000相当,优于SPIHT压缩算法,但编码算法更简单,有利于系统的硬件实现。

An novel image compression algorithm valid for hardware implementation is proposed,which is entitled set-tree block coding,first,arrange the coefficients of wavelet decomposing into set-trees by the way of zero-tree,and arrange the set-trees into set-tree block,then code them by the comparability of resolution and correlation of neighboring coefficients,so the compressibility is improved.The set-tree block is coded independently to improve the anti-BER（bit error rate）.The result of simulation indicates：this algorithm has a quite compression efficiency with JPEG2000 and better than SPITH,but this algorithm has low complexity,and is easy for hardware implementation.