摘要
通过分析小波零树编码和分形编码各自的优点与不足,结合自身的特点,提出一种基于分类的混合图像编码算法。该方法把图像分割后的块分为3类,对不同类型的块分别采用分形算法或零树编码的改进算法SPIHT算法编码;结果表明,与单独使用这两种算法相比,混合编码算法的重构图像质量有一定程度的提高。
The merits and disadvantages of wavelet zerotree coding and fractal coding are analyzed, and in view of the features of an image, a hybrid coding algorithm based on classification is proposed. In the algorithm, an image is partitioned into blocks, and then the blocks are classified into three types and coded separately with the fractal algorithm or the SPIHT algorithm, an improved zerotree coding algorithm. Experiments are conducted, and a comparison is made among the hybrid algorithm, the fractal algorithm and the SPIHT algorithm, which indicates that the quality of the reconstructed image by the hybrid coding algorithm is improved to a certain extent.
出处
《合肥工业大学学报（自然科学版）》
CAS
CSCD
北大核心
2007年第8期929-933,共5页
Journal of Hefei University of Technology：Natural Science
关键词
混合图像编码
小波零树编码
分形编码
分类
hybrid image coding
wavelet zerotree coding
fractal coding
classification
