针对暗原色先验模型对于图像明亮区域不适应, 暗原色估计偏大, 导致透射率估计偏小, 出现色彩失真现象, 本文介绍一种新的暗原色修正方法。提出一种逆暗原色概念, 将雾化图像的暗原色与逆暗原色进行融合处理得到一种新的修正暗原色, 从而获得比较真实的明亮区域透射率, 有效消除了明亮区域的色彩失真。以有效细节强度、色调还原程度、结构信息及综合测评作为图像质量评价指标, 与目前流行算法进行对比实验, 本文算法的色调还原程度指标平均值提高41.1%, 综合测评指标平均值提高48.7%。实验结果表明, 本文算法在改善明亮区域色彩失真及提高去雾图像总体质量方面优于目前流行算法。

Aiming at the problem that dark channel prior model does not suit the image bright region, since dark channel estimate is too small, the estimated transmittance is small, and color distortion occurs, a new method for correcting dark channel has been proposed. One kind of inverse dark channel concepts have been proposed, dark channel and inverse dark channel of the haze image are fused to get a new dark channel, thereby obtaining more real bright area transmittance, effectively eliminating color distortion in the bright region. Effective strength of the details, the degree of reduction of tone, structure information and comprehensive evaluation as the image quality evaluation index, compared with the current popular algorithms, the average value of tonal reproduction degree is improved by 411% and the average comprehensive evaluation is increased by 48.7%. Experimental results show that proposed method is superior to the current popular algorithms in improving the color distortion of the bright region and improving the overall quality of the image haze removal.