基于空域梯度算子和小波域多分辨分析的图像增强算法研究  

Image enhancement algorithm of airspace gradient operator and wavelet-domain multiresolution analysis based

作　　者：刘鑫淼[1] 汪西原[1] 马润渊[1] 

机构地区：[1]宁夏大学物理电气信息学院,宁夏银川750021

出　　处：《宁夏工程技术》2009年第2期117-120,共4页Ningxia Engineering Technology

基　　金：宁夏自然科学基金资助项目(NZ0724)

摘　　要：为改善图像的视觉效果,提出一种基于Sobel算子和Laplacian算子相结合的混合算法,并采用基于Sobel算子、Laplacian算子、小波变换和提升小波变换的算法,对滚珠缺陷图像和医学CT图像进行图像增强处理的算法研究。实验结果表明,混合算法适于滚珠缺陷图像的锐化,其效果优于另外4种算法;提升小波变换算法适于细节信息较为丰富的医学图像增强,与其他5种方法相比其效果最优。This article presents a mixed algorithm based on combining Sobel operator and Laplacian, and aiming at the defected image of ball bearing and CT image, makes uses of Sobel operator, Laplacian, wavelet transform and lifting wavelet transform algorithm to study enhanced algorithm. Experimental results demonstrate that the effect of mixed algorithm is the best of four algorithm to ball defect image sharpening and lifting wavelet transform algorithm is applied to enhance medicinal image with particular information, and the effect is the best of all.

关 键 词：SOBEL算子 LAPLACIAN算子 多分辨分析 提升小波变换算法 图像增强 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]