一种基于深度信息的图像雾效模拟算法

在已知的图像场景深度信息和大气散射模型的基础上，提出了一种基于深度的图像雾效模拟算法。算法引入了最大能见度概念，通过场景深度与雾透射率之间的比例关系将深度图直接转化为透射率图，使用引导滤波对透射率进行优化，最终通过大气散射模型生成雾效模拟图像。实验表明，算法所生成的雾效场景符合空间透射原则，在不同的能见度和大气光结合的条件下可获得变化丰富的有雾效果，具有较强的真实感。

An image haze simulation algorithm using depth information

A haze simulation algorithm is proposed,exploiting known scene depth information and the atmosphere scattering mod-el.The concept of maximum visibility is introduced.The proportionate between scene depth and haze transmittance is applied to transform the depth maps to tranmission maps.The guided filtering algorithm is used to refinethe transmission maps and simu-late haze effect on the image.The experimental results demonstrate that the simulated scene by the proposed algorithm follows the aerial perspective rules.And the generated haze effects present good reality with the combination of different atmospheric visi-bility and light.