限制D块耦合极点插值的分形图像压缩算法  

Fractal image compression algorithm based on limited D block coupled pole interpolation

作　　者：邓宾[1] 

机构地区：[1]东营职业学院电子信息与传媒学院,山东东营257091

出　　处：《信息技术》2014年第7期143-148,共6页Information Technology

摘　　要：由于当前分形图像搜索编码都忽略了对比度因子约束,且对于负载性较高的图像,其压缩比较低,继而降低了解码图像质量以及计算效率。对此,构造了限制D块搜索范围耦合极点插值复原技术的分形图像压缩算法。引入对比度因子约束,设计了一种限制D块搜索范围的编码机制。并基于皮亚诺扫描,提出了极点插值复原技术。测试算法性能,结果显示:与当前的分形图像压缩算法相比,文中算法的压缩效果优异,其平均PSNR提高了1dB^3dB,且编码速度提高了约11.6倍,可满足实时性,在压缩比越高时,提高越明显。On the basis of researching fractal coding related theory, principle and realization of the basicfractal coding method, an improved method limiting the searching range of the D block is proposed in thispaper, the result shows that the average PSNR of the decoded images of the improved method drops 0.78dB, but the coding speed is up about 11. 6 times, but it is still different to meet the real-timerequirement. Considering the drawbacks of the fractal based coding method, an adaptive method isproposed, and it explores a new coding method using the geometric properties of the image data curve, acoding method based on extremum interpolation is proposed and the experiment is done with 6 differentimages. The result shows that this method is reasonable and effective. This method is close to fractalbased method in coding time, but is better in coding effect. The PSNR of decoded images of this methodcan rise 1 - 3dB in the premise of close compression and the rising is more obvious when the compressionrate larger.

关 键 词：图像压缩 分形编码 分形基 极点插值 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]