摘要
提出一个基于小波分析和非线性PCA(nonlinear principal component analysis,NLPCA)的图像压缩算法,该算法通过对小波分解系数的能量大小对各个小波子图进行分类,然后用具有高强压缩能力的NLPCA对各类子图进行不同程度的压缩。其主要特点是在高压缩比条件下能达到很高的图像重构信噪比。实验结果表明该算法性能良好,在压缩比为93.1时其图像重构信噪比都能达到30以上。
An image compression algorithm based on wavelet analysis and nonlinear principal component analysis （NLPCA） is put forward. The wavelet sub-image via the energy magnitude of coefficient decomposed by wavelet analysis is classified, and then each sub-image is compressed to different extents with NLPCA. High picture reconstruction signal noise ratio （PSNR） is achieved at a high compression rate. The experiment validates the algorithm＇s good performance. The PSNR is greater than 30 at the compression rate of 93.1.
出处
《系统工程与电子技术》
EI
CSCD
北大核心
2005年第9期1532-1535,共4页
Systems Engineering and Electronics
关键词
小波分析
非线性主成分分析
重构
图像压缩
wavelet analysis
nonlinear principal component analysis
reconstruct
image compression
