NSCT变换域PCNN彩色图像增强算法

为了改善由于光照不均对真彩色图像的影响，根据人类对颜色的感知特性，利用YCbCr彩色模型空间，提出了一种基于非抽样Contourlet变换（ NSCT）的PCNN模型的图像增强算法。首先，将图像从RGB彩色空间转换到YCbCr彩色空间；然后对亮度分量进行NSCT分解，得到低频子带系数和高频方向子带系数，对低频子带，利用PC－NN增强，对高频子带采用非线性变换进行增强；最后，使用NSCT逆变换重构图像的亮度分量，并将图像从YCbCr色彩空间模型还原到RGB空间得到增强后的图像。实验结果表明，算法增强效果明显优于同态滤波、空域PCNN及NSCT域的PCNN算法，不仅增加了彩色图像的明亮度，而且图像保真性好，纹理更清晰。

An Enhanced Method of Color Image Combined PCNN Based on NSCT

A new enhanced method of non-uniform illuminaton color image based on NSCT was pro-posed.First,a color image is transformed from RGB color space to YCbCr color space .Second,luminance channel is decomposed by NSCT ,and We can obtain the low-frequency sub-image and a series high-fre-quency sub-images.Third the PCNN was applied to the low-frequency subimage ,and nolined translation was applied to high-frequency sub-images.Forth,luminance component image is obtained by inverse NSCT.Last,the color image is transformed from YCbCr color space to RGB color space .Experiments il-lustrate that this algorithm can not only corrected non-uniform illumination in images ,but also maintained a good image color and the local details .