摘要
富含细节、纹理和边缘的图像中，重要信息大量集中在中、高频部分，基于小波变换的压缩方法只对低频信息进行多。次分解．针对这个问题，以SPIHT算法为基础，研究了基于小波包分解的图像压缩算法．仿真结果表明，采用小波包分解的SPIHT算法与传统SPIHT算法相比，在同压缩比情况下，峰值信噪比提高了0．35～1dB，适用于纹理丰富的图像的压缩.
Image compression methods based on wavelet transform regarding the information in the middle and high frequency is unimportant, so wavelet analysis decomposes the lower frequency bands. But in texture - rich images most significant information of texture and region boundaries often appear in the middle and high frequency. In our paper, Set Partitioning in Hierarchical Trees （SPIHT） algorithm is used to code the wavelet packet coefficients corresponding to the selected best tree. Simulation shows the SPIHT coder decomposed by wavelet packet improves the peak signal - to - noise ratio by 0. 35 to 1 dB for texture - rich images compared with the classical SPIHT coder at the same compression ratio.
出处
《哈尔滨理工大学学报》
CAS
2006年第5期72-75,共4页
Journal of Harbin University of Science and Technology
关键词
图像压缩
小波包分解
SPIHT
image compression
wavelet packet decomposition
SPIHT
