[1]张爱华,唐婷婷,汪玮玮,等.基于主成分特征的快速分形图像压缩算法[J].计算机技术与发展,2018,28(05):77-80.[doi:10.3969/j.issn.1673－629X.2018.05.018]
　ZHANG Ai-hua,TANG Ting-ting,WANG Wei-wei,et al.A Fast Fractal Image Compression Algorithm Based on Principal Component[J].,2018,28(05):77-80.[doi:10.3969/j.issn.1673－629X.2018.05.018]
点击复制

基于主成分特征的快速分形图像压缩算法()
分享到：

《计算机技术与发展》[ISSN:1006-6977/CN:61-1281/TN]

卷:
28
期数:
2018年05期
页码:
77-80
栏目:
智能、算法、系统工程
出版日期:
2018-05-10

文章信息/Info

Title:
A Fast Fractal Image Compression Algorithm Based on Principal Component
文章编号:
1673-629X(2018)05-0077-04
作者:
张爱华; 唐婷婷; 汪玮玮; 张璟
南京邮电大学 理学院，江苏 南京 210023
Author(s):
ZHANG Ai-hua; TANG Ting-ting; WANG Wei-wei; ZHANG Jing
School of Science，Nanjing University of Posts and Telecommunications，Nanjing 210023，China
关键词:
分形; 降维; 主成分分析; 图像重构
Keywords:
fractal; dimension reduction; principal component analysis; image reconstruction
分类号:
TP301.6
DOI:
10.3969/j.issn.1673－629X.2018.05.018
文献标志码:
A
摘要:
作为一种有损压缩算法，高压缩比的分形压缩算法具有较好的解压质量，但其编码耗时过长。为了减少搜索匹配块的时间，在基本分形压缩算法的基础上，先对图像块进行处理，基于主成分分析法(PCA)对图像块进行降维，然后选取图像块中最有效的向量信息作为主成分，忽略掉信息量较小的一些向量，仅用保留两个信息量最大的主成分向量描述图像特征，实现降维分析 － 特征显示的目的。搜索匹配块时，对待编码的 Ｒange 块，仅在该 Ｒange 块主成分特征最为相近的Domain 块搜索邻域范围内找到其最佳匹配块，并且根据阈值来调节搜索邻域的大小。实验结果表明，改进算法能有效降低数据的复杂度，在保证图像重构几乎不受影响的情况下，缩短了编码时间。
Abstract:
As a lossy compression algorithm，the fractal compression algorithm with high compression ratio has better decompression quality，but its coding takes too long．In order to reduce the time of searching for matching blocks,based on the basic fractal compression algorithm,we firstly process the image blocks,reduce the dimension of the image blocks based on principal component analysis (PCA)，and then select the most effective vector information in the image blocks as the main component,ignoring some of the vector with a small a-
mount of information.The most effective information in image blocks is selected as the principal component,and the image components are described by it to realize dimensionality reduction analysis and characteristics display．For the encoded Range block,only the best match block is found within the domain of the Domain block search area where the main component of the Range block is closest．Experiments show that the improved algorithm can effectively reduce the complexity of the data and shorten the coding time in the case that the image reconstruction is almost unaffected.

相似文献/References:

[1]陈华光 王京文 张晓清.虚拟场景中基于LOD的树木真实感建模[J].计算机技术与发展,2008,(12):228.
　CHEN Hua-guang,WANG Jing-wen,ZHANG Xiao-qing.Realistic Modeling of 3D Tree Based on LOD in Virtual Scene[J].,2008,(05):228.
[2]张明书 张燕 王维民 梁立凯.三维植物的计算机模拟算法研究[J].计算机技术与发展,2006,(10):108.
　ZHANG Ming-shu,ZHANG Yan,WANG Wei-min,et al.Computer Simulation Algorithm Research of Three - Dimensional Plant[J].,2006,(05):108.
[3]刘萍 刘艳艳 宗岩.基于分形的树木建模算法的研究[J].计算机技术与发展,2011,(12):26.
　LIU Ping,LIU Yan-yan,ZONG Yan.Research of Trees Simulation Algorithm Based on Fractal[J].,2011,(05):26.
[4]赵春艳 马学强.基于DLA的虚拟植物根系模拟方法研究[J].计算机技术与发展,2012,(02):119.
　ZHAO Chun-yan,MA Xue-qiang.Research on Technology of Simulating Virtual Plant Roots Based on DLA[J].,2012,(05):119.
[5]张爱华 盛飞 杨培 常康康.基于相似比的快速分形编码算法[J].计算机技术与发展,2012,(11):176.
　ZHANG Ai-hua,SHENG Fei,YANG Pei,et al.A Fast Fractal Encoding Algorithm Based on Similar Ratio[J].,2012,(05):176.
[6]张爱华 杨培 盛飞 常康康.利用相对误差排序的快速分形编码算法[J].计算机技术与发展,2012,(12):105.
　ZHANG Ai-hua,YANG Pei,SHENG Fei,et al.A Fast Fractal Encoding Algorithm with Relative Error-ordered[J].,2012,(05):105.
[7]顾兵.WSN 中规则区域的最优覆盖研究[J].计算机技术与发展,2013,(01):107.
　GU Bing.Research on Optimal Coverage Problem of Regular Region in WSN[J].,2013,(05):107.
[8]郭斌,齐金鹏.基于改进LLE的Elman神经网络电路故障诊断[J].计算机技术与发展,2014,24(03):1.
　.Research on OIF-Elman Neural Network Circuit Fault Diagnosis Based on Improved LLE[J].,2014,24(05):1.
[9]李娜[],方恩浩[],滕飞[],等. 基于L-系统的分形植物模拟研究[J].计算机技术与发展,2014,24(08):199.
　LI Na[],FANG En-hao[],TENG Fei[],et al. Research on Fractal Plant Simulation Based on L-system[J].,2014,24(05):199.
[10]李远博,曹菡. 基于PCA降维的协同过滤推荐算法[J].计算机技术与发展,2016,26(02):26.
　LI Yuan-bo,CAO Han. Collaborative Filtering Recommendation Algorithm Based on PCA Dimension Reduction[J].,2016,26(05):26.

更新日期/Last Update: 2018-06-28