摘要
分形图像编码算法是一种很有前途的限失真编码方法,但其编码时间长、计算复杂度高。针对此问题,文中提出了一种利用相对误差排序的快速分形图像编码算法。这种算法首先将码本按照相对误差大小进行排序,然后使用折半查找搜索法寻找给定Range块在相对误差意义下的最佳匹配块,进而利用一个联系相对误差和均方根的不等式来在这个最佳匹配码块的邻域中搜索Range块在均方根意义下的最佳匹配码块,从而加快了搜索速度。仿真实验结果显示,与全搜索基本分形编码算法相比,在主观判断没有明显失真的前提下,该算法能够加快编码速度。
Fractal image encoding algorithm is a promising Iossy compression technique, but it has long encoding time and high computa- tional complexity. A fast fractal image encoding algorithm based on relative error-ordered is thus proposed in this paper. In the proposed algorithm,after the codebook blocks are sorted according to their relative error intensities,then use binary search method to find out the best matched codebook blocks regarding to relative error of a given range block. And then, the proposed algorithm uses an inequality link- ing the root-mean-square （ RMS ） and relative error to convert the range-domain block matching problem to the nearest neighbors search problem in the sense of RMS. By comparing the baseline fractal algorithm with the full search, the simulation result shows that the pro posed algorithm not only can guarantee the image＇s quality ,but also can achieve the speed-up.
出处
《计算机技术与发展》
2012年第12期105-107,111,共4页
Computer Technology and Development
基金
国家自然科学基金(61070234)
南京邮电大学校科研基金(NY210018)
关键词
分形
图像压缩
分形图像编码
相对误差
fractal
image compression
fractal image encoding
relative error
