图像去雾的小波域 Retinex 算法

针对传统Retinex算法在图像去雾处理中易造成光晕伪影、边缘模糊以及算法复杂度高等缺点，提出了一种图像去雾的小波域Retinex算法。该算法将Retinex算法拓展到小波域，利用雾天图像中雾霾与景物细节信息在小波域中能量的不同分布特点，采用Retinex算法抑制雾霾分量，用锐化处理增强景物细节信息分量，以降低雾霾对图像的影响，增强图像细节的清晰度；该算法用双边滤波替代传统Retinex算法中的高斯滤波，以克服高斯滤波对图像边缘的模糊效应和光晕伪影。实验结果表明，该算法能有效地改善雾天图像的退化现象，提高图像的清晰度。

Wavelet-based retinex algorithm for image defogging

A wavelet-based Retinex algorithm for image defogging is proposed to avoid Halo artifacts , edge fuzzi-ness in traditional Retinex and improve its efficiency .In light of different energy distribution of fog , haze and de-tails of subjects in the wavelet domain , this algorithm uses Retinex to restrain the influence of the fog and haze , and adopts sharp processing to enhance the definition of the subjects′details.The bilateral filter is utilized to re-place Gaussian filter in traditional Retinex to eliminate mistiness in image edges and Halo artifacts .Experiment results show that this algorithm can effectively improve the degradation of foggy and hazy image and enhance its definition.