一种BEMD第一余量图像压缩的方法  

Method for First Residue Image Compression of Bidimensional Empirical Mode Decomposition

作　　者：葛光涛 桑恩方[1] 刘卓夫 高云超[1] 

机构地区：[1]哈尔滨工程大学水声工程学院,哈尔滨150001

出　　处：《计算机工程》2008年第9期25-27,共3页Computer Engineering

基　　金：国家自然科学基金资助项目(10504004)

摘　　要：提出一种针对二维经验模态分解第一余量图像的压缩方法。解释了第一模态、第一余量和第一均值等概念,将可以利用网格特征点表示的各个均值曲面分别进行压缩。根据第一余量曲面和其各个均值曲面的加和关系,将各个均值曲面分别重建后再相加得到第一余量曲面(图像)高质量的还原结果。实验结果表明,该方法在图像低频信息压缩方面与传统的压缩方法相比具有明显优势。A method for the image compression of the first residue in the Bidimensional Empirical Mode Decomposition（BEMD） is introduced. It explains the several concepts such as the first IMF, the first residue and the first local mean, respectively compresses the several mean surfaces by representing them with the grid characteristic points, and gets the high quality reconstruction of the first residue surface（image） by adding the several reconstructed mean surfaces together. Experiments prove that the novel method has obvious advantages on the issue of compressing the low frequency information of one digital image compared with the current methods.

关 键 词：图像压缩 第一余量 二维经验模态分解 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]