一种改进的暗原色单幅图像去雾方法

暗原色先验方法可以较好地处理单幅图像去雾,但对理图像中的灰白区域处理效果不好。通过分析暗原色先验原理,得出了暗通道图像和雾的透射分布率以及雾的浓度系数的关系,提出了一种结合峰值信噪比和暗原色优先法则的去雾方法。由实验结果分析,该方法能够更清晰的表现去雾后图像的细节,并且一定程度上克服了原方法在处理图像中灰白区域效果不佳的弱点。

Method of Single Image Removal Based on Improved Dark Channel Priority

Though the method of dark channel priority achieves meaningful results in the process of single image removal,it doesn't work well in the treatment of the gray area in images.According to the analysis of the principle of the dark channel priority,the relation among dark channel image,the transmission distribution rate of the haze and the coefficient of the haze concentration is given.A haze removed method is proposed which combined the Peak Signal to Noise Ratio(PSNR) with the rule of the dark channel priority.This method is proved to present the haze removal image detail more distinct and overcome the weaknesses of the original method in the processing of the gray area in images to some extent by the analyse of the experimental result.