基于非抽样的小波变换的彩色图像增强方法  

Color Images Enhance Algorithm Based on Undecimated Wavelet Transform

作　　者：吴粉侠[1] 段群[1] 李洪星[1] 

机构地区：[1]咸阳师范学院图形图像研究所,陕西咸阳712000

出　　处：《咸阳师范学院学报》2015年第4期52-55,共4页Journal of Xianyang Normal University

基　　金：陕西省教育厅科研计划项目(14JK1802);咸阳师范学院科研基金项目(13XSYK058);国家级创新训练项目(20131072002)

摘　　要：直接用灰度图像增强算法对真彩色图像增强,容易产生色彩偏差。提出一种新的增强算法:先将真彩色图像由RGB颜色空间转换到HSV颜色空间;再对饱和度分量V作非抽样小波变换,对变换域的低频系数采用对数变换以压缩动态范围进行增强,对高频系数进行分段线性变换以增强边缘及纹理;最后再用非抽样小波逆变换重构V分量,将图像由HSV空间还原到RGB颜色空间。实验表明,增强后的图像色彩基本无偏差,动态范围压缩良好,亮度对比度都得到明显提高,并且取得了较高的信息熵,空间频率,平均梯度,均方差。It is easy to produce color deviation that the enhancement algorithm of gray image used directly to true color image enhancement. In this paper, a new approach for enhancing contrast of color image based on undecimated discrete wavelet transform（UDWT） and HSV is proposed. The color im- age is converted to HSV（hue, saturation, value） color space. The V, which represents the luminance of color image, decomposed to its coefficients by NDWT, then applying the logarithmic transformation on low frequency coefficient and applying the piecewise linear transformation on high frequency coeffi- cient. Then, inverse undecimated transform is performed to reconstruct the enhanced V compoment. The S component is enhanced by the logarithmic transformation while the H component does not change to avoid degradation color balance between the HSV components. Finally the enhanced S and I together with H are converted back to its original color system.Experiments show that, the enhanced low contrast color images have basic no deviation, dynamic range compression is good, brightness and contrast are improved obviously, and has obtained the information entropy, the higher spatial frequency, the higher average gradient, the higher mean variance.

关 键 词：真彩图像 HSV空间 非抽样小波变换 对数变换 

分 类 号：TP317.4[自动化与计算机技术—计算机软件与理论]