| 针对纹理图像压缩的改进SPIHT算法 |
| |
| 作者姓名： | 潘志刚 高鑫 |
| |
| 作者单位： | 中国科学院电子学研究所,北京 100190 |
| |
| 基金项目： | 国家863计划项目(2008AA121805-1)资助 |
| |
| 摘 要： | 针对纹理图像压缩提出了一种改进的SPIHT算法. 算法对一阶小波变换产生的LL、HL、LH、HH 4个子带系数分别进行小波再分解,之后对上述4个子带利用SPIHT算法进行联合编码;子带之间的量化比特分配嵌入在编码过程之中,因此可实现4个子带编码比特的自动精确分配. 编码过程充分利用了小波系数的高频分量信息,可有效保持图像中的纹理细节. 最后通过对纹理图像的压缩实验,证实了改进算法的有效性.
|
| 关 键 词： | 图像压缩 SPIHT 高频分解 纹理图像 |
| 收稿时间： | 2009-06-08 |
| 修稿时间： | 2009-11-05 |
An improved SPIHT algorithm for texture image compression |
| |
| Authors: | PAN Zhi-Gang GAO Xin |
| |
| Institution: | Institute of Electronics, Chinese Academy of Sciences, Beijing 100190, China |
| |
| Abstract: | An improved SPIHT algorithm for texture image compression is proposed. A re-decomposition scheme is applied to the one order wavelet transformation coefficients independently in the LL, HL, LH, and HH sub-bands, and then the SPIHT algorithm is adopted to encode the mixed coefficients of the four sub-bands jointly. This method does not process the four sub-bands separately. The quantization bit allocation in the sub-bands is embedded in encoding process and precise bit allocation among the four sub-bands is automatically implemented. This method adequately utilizes the high frequency information of wavelet coefficients and efficiently preserves the image texture while the PSNR is not degraded. The experimental results validate the effectiveness of this improved algorithm for texture image compression. |
| |
| Keywords: | image compression SPIHT high-frequency decomposition texture image |
| 本文献已被 CNKI 等数据库收录！ |
| 点击此处可从《》浏览原始摘要信息 |
| 点击此处可从《》下载免费的PDF全文 |
|
