数字图像的NLEMD分层压缩研究  

A New Multi-layered Digital Image Compressing Approach Based on NLEMD

作　　者：兰国辉[1] 胡程鹏[1] 王秀森[1] 

机构地区：[1]海军大连舰艇学院航海系,大连116018

出　　处：《科学技术与工程》2011年第26期6364-6369,共6页Science Technology and Engineering

摘　　要：二维限邻域经验模式分解(NLEMD)通过设定最大邻域(时宽)和采用邻域内局部自适应均值算法代替包络均值算法对图像进行分解,克服了以往EMD分解算法出现的灰度斑现象,改善了分解图像的质量。采用NLEMD算法进行图像分解,并基于当前普遍使用的JPEG图像压缩标准提出了数字图像分层压缩的算法。实验结果表明算法在确保重构图像较高质量的前提下,各层图像都实现了较高的压缩比率,对图像的分层压缩与传输的实现具有较高的实用价值。Neighborhood limited empirical mode decomposition （NLEMD） is one novel time-frequency analysistool which has the adaptive features and meanwhile adopts neighborhood limited （ max time-width） to overcome other EMD＇ s gray spots in images , and enhances images. NLEMD is adopted to decompose images , and , at the same time , proposes a new multi-layered image compressing approach based on JPEG which is the most popular image compression standard. The experiment result shows that this approach can keep the quality of images , and compress each layer to a good ratio . So it is useful in multi-layered image compressing and transmission.

关 键 词：NLEMD 图像分解 分层处理 JPEG 压缩 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]