基于双通道及图像质量评价模型的去雾方法

Image dehazing method based on dual channel and image quality evaluation model

针对传统暗通道先验去雾算法易产生明亮区域失真、去雾图像整体偏暗等问题,提出一种基于双通道及图像质量评价模型的去雾方法。首先,划分出图像的明亮区域与非明亮区域;其次,利用双通道先验算法准确估计出大气光值;接下来,将暗通道先验的透射率作为非明亮区域的透射率,在明亮区域单独构建透射率,将二者融合细化,得到带有参数的透射率;最后,通过构造参数驱动图像质量评价模型,迭代选取最优的无雾图像。实验结果表明,算法去雾效果良好,可以有效地抑制Halo效应,避免区域的失真,改善复原图像的综合质量。

Aiming at the problems that traditional dark channel prior dehazing algorithm is prone to produce bright area distortion and the whole dehazing image is dark,an image dehazing algorithm based on dual channel and image quality evaluation model is proposed.Firstly,the bright region and the non-bright region of the image are divided.Secondly,the atmospheric light value is accurately estimated by using the two-channel prior algorithm.Next,the transmittance of the dark channel prior is taken as the transmittance of the non-bright region,and the transmittance is constructed separately in the bright region.The two are fused and refined to obtain the transmittance with parameters.Finally,a parameter-driven image quality evaluation model is constructed to iteratively select the optimal fog-free image.The experimental results show that the proposed algorithm can effectively suppress Halo effect,avoid regional distortion and improve the comprehensive quality of therestored image.