基于小波的图像基本处理技术研究  

Image preprocessing based on wavelet

作　　者：汤仁民 李国芳[1] 王代强[2] 

机构地区：[1]贵州大学大数据与信息工程学院,贵州贵阳550025 [2]贵州大学人武学院,贵州贵阳550025

出　　处：《微型机与应用》2015年第2期44-46,共3页Microcomputer & Its Applications

摘　　要：鉴于小波多尺度分解与重构在图像的光照补偿、去噪、压缩方面的重要作用,介绍了小波变换的分解与重建,提出了包括IVIW光照补偿方法、基于Haar小波的去噪与压缩方法 ,并通过MATLAB R2009a平台对Yale人脸库中的图像进行成功验证。结果表明,基于小波变换的图像预处理方法简单易行且有效,解决了原始图像包含巨大数据量的问题,减少了计算的复杂度,尽可能消除了可变因素的影响,得到清晰的人脸图像。Given the important role of the wavelet muhi-scale decomposition and reconstruction in image illumination compensation, de-nosing and compression, this article introduces the decomposition and reconstruction based on the wavelet transform. It proposes illumination compensation method called IVIW, the Haar wavelet de-noising and compression algorithms. And it is validated successfully by carrying on the experiment in Yale face storehouse through the MATLAB R2009a platform. The results show that the image preprocessing based on wavelet transform is a simple, reliable and effective method. And it has solved the problem of the original image with huge amount of data to reduce the computation complexity and eliminated the influence of variable factors as much as possible to get a clear face image.

关 键 词：人脸识别 IVIW HAAR小波 多尺度分解 图像重构 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]