一种基于K-均值聚类优化的快速分形图像压缩算法

A fast fractal image compression algorithm based on K-mean clustering optimization

对搜索窗中的父块和子块,根据其方差的不同,利用K-均值聚类优化方法分别对子块和父块进行聚类,子块只对同一类中的父块进行匹配,从而大大缩短了编码时间.仿真实验结果表明,在不影响信噪比和压缩比的前提下,与经典分形压缩算法相比,该算法编码速度可提高大约5倍;同近期文献报道的基于方差的快速分形压缩算法相比,该算法的结果也有明显的改善.

Range and domain blocks are clustered by using K-mean clustering method, and range blocks search domain blocks in the same category, which can shorten encoding time significantly. The encoding speed of our method is about 5 times faster than that of the classical Jacquin＇ s algorithm, and the quality of the decoding images can be retained as well when the compression ratio is fixed. We also tested some other fast encoding schemes based on variance, and the experimental results show that our algorithms are superior to them.