无链表SPIHT图像提升小波编码的硬件算法

Listless SPIHT Image Coding Hardware Algorithm Based on Lifting Wavelet

为了解决小波零树编码算法中,SPIHT(分级树的集合分裂)算法占用大量存储空间的问题,给出了一种无链表SPIHT图像压缩算法.在这种算法中,小波系数按照SPIHT的顺序输入到编码器中,编码器按照NLS算法编码,小波变换采用9/7提升分解算法,量化算法为均匀标量量化,图像边界延拓采用了补零延拓方法.实验结果表明,该算法图像压缩比与JPEG2000的EBCOT(基于优化截断的嵌入式块编码)算法几乎相同,编码速度大约是后者的2倍,一幅512×512的图像仅需要17.875KB额外内存.

To overcome the shortcoming of wavelet zero-tree coding algorithms, i. e. , the hardware implementation of SPIHT （set partitioning in hierarchical trees） algorithm needs too many memories, a listless SPIHT algorithm was proposed. In this algorithm, wavelet coefficients are input into an encoder in the SPIHT order, and then they are coded by using the NLS （no list SPIHT） algorithm. In addition, the 9/7 lifting schemes and uniform quantization are employed, and the adding zero method is applied for border extension of images. The experimental results show that the compression ratio of the proposed algorithm is almost the same as that of the EBCOT（ embedded block coding with optimized truncation） algorithm of JPEG2000, but its running speed is two times that of the EBCOT algorithm. Furthermore, the proposed algorithm needs fewer memories, such as only an additional 17. 875 KB memory for a 512×512 image.