基于物理模型的自适应快速单幅图像去雾方法

为提高雾天条件下降质图像的清晰度和色彩保真度,提出一种基于物理模型的自适应快速去雾方法。该方法求取暗原色的窗口选择为最小,并根据暗原色粗估计透过率,采用改进的增加补偿函数的双边滤波做优化,且对双边滤波中敏感的参数作自适应估计,最后通过物理模型恢复图像。实验结果表明,对于单幅图像在确保去雾效果的前提下,降低了算法的复杂度,且没有光晕效应。

An adaptive and fast defogging method based on physics model

In order to improve the clarity and the color fidelity of images taken in foggy weather,a fast and adaptive defogging method based on the atmospheric scattering model is proposed.This method sets the patch size of the dark channel as the minimum value and roughly estimates the transmission map based on the dark channal.To reduce the computational complexity of the original algorithm,an improved bilateral filtering with added compensation function is used to refine the transmission map,and the sensitive parameter of bilateral filtering is estimated adaptively.A haze-free image is then restored by the physical model.The experimental results indicate that this method can not only reduce the complexity of the algorithm greatly with the promised defogging effect,but also eliminate the halo effect.