针对传统的Retinex算法对雾天和彩色图像增强时，会出现色彩恢复不协调，光照分布不均匀这类缺点，提出了一种新的基于Zernike矩模型的Retinex图像增强方法。通过在HSV空间中求解图像V分量和S分量的Zernike矩来提取图像的背景灰度和阶跃高度，然后调节邻接像素内灰度变化差异和区域饱和度的相关性，进而增强图像的亮度，恢复图像的色彩。实验结果表明，该方法有效地解决了图像色彩恢复和光照恢复不足这一问题，而且对不同特点的图像都有良好的适应能力。

In view of such problems as the limited coloration effectiveness and the illumination distribute inharmonious for the traditional enhancement algorithm, a novel Retinex algorithm of image enhancement based on Zernike moment is proposed in this paper.By the Zernike moment of the V and S components of the image, the background gray level and step height in HSV space will be calculated. These results can be used to adjust the difference of gray level in intersection pixel and improve the correlation of regional saturation, and further to enhance the illumination and restore the image coloration. Experimental results show that the algorithm can both restore the image illumination, coloration effectively and adapt to various images of different characteristics.