基于颜色衰减先验的自适应Retinex去雾算法

当前的图像去雾算法中对自适应的要求越来越高,而传统的Retinex算法无法根据雾天图像的实际雾化情况进行去雾,导致处理后的图像仍然存在细节不突出以及色彩失真等问题。针对上述问题,提出了一种基于颜色衰减先验的自适应Retinex去雾算法。利用颜色衰减先验理论求得有雾图像的景深信息,通过建立的景深和高斯尺度参数的线性模型实现对亮度分量的自适应Retinex去雾处理;其次采用饱和度自适应线性拉伸算法优化饱和度分量,最终实现雾天图像的自适应处理。实验结果表明,上述算法在突出图像细节的同时,能够更好地修复图像本来的色彩,改善了雾气浓度不均对图像的影响,为图像去雾的自适应处理提供了有益参考。

An adaptive Retinex of Image Haze Removal Based on Color Attenuation Prior

The current image defogging algorithms have higher and higher requirements for self-adaptation.However,the traditional Retinex algorithm cannot defog adaptively according to the actual atomization of foggy images,resulting in some problems such as unpromising details and color distortion.Therefor,in view of the above problems,an adaptive Retinex defogging algorithm based on color attenuation prior is proposed.Firstly,use the color attenuation prior theory was used to obtain the depth of field information of fogging image.Secondly,a linear model of depth of field and Gaussian scale parameters was established to realize the adaptive Retinex defogging of brightness component.Then,the saturation adaptive linear stretching algorithm was adopted to optimize the saturation component and the adaptive processing of foggy images was finally achieved.And the experimental results show that the algorithm can not only highlight the details of the image,but also better repair the original color of the image,and improve the effect of uneven fog concentration on the image,which provides a useful reference for adaptive image defogging.