【作者】 张梁斌； 奚李峰； 范申；

【Author】 Zhang Liangbin,Xi Lifeng,Fan Shen(Faculty Of Computer Science And Information Technology,Zhejiang Wanli University,ningbo,315100

【机构】 浙江万里学院计算机与信息学院；

【摘要】 分形图像压缩中由Fisher提出的固定门限四叉树方法,子块与父块的实验门限值一般由经验来设定,直接决定了分形图像编码的效率,门限值的设定是分形图像压缩中的技术难点之一。本文考虑了输入图像的特点,给出了自适应门限的计算推导过程,提出了门限与子块的方差成正比的自适应门限的分形图像压缩方法。实验表明,对同类图像该方法压缩时间短,还原图像PSNR高,提高了分形压缩编码效率。更多还原

【Abstract】 Fisher put forward the method of fixed threshold value quadtree in fractal image compression. Experiment threshold between child bolck and father bolck is always decided by experience, which directly decides the efficiency of fractal image compression and came to be hard technology difficulty. According to inputting image characteristic, this paper put forward computering deductive procession and new quadtree fractal image compression on adaptive threshold value direct ratio child Variance. Experimental results show that this method can abtain shorter compression time and higher image psnr with the same image, which improved fractal image compression efficiency greatly.更多还原