一种基于图像深度的单幅图像雾效模拟方法  

A Method of Fog Simulation of Single Image Based on Depth Map

作　　者：孙翠敏 李转运[1] Sun Cuimin;Li Zhuanyun(Anhui Post and Telecommunication College,Hefei 230001,China)

机构地区：[1]安徽邮电职业技术学院,安徽合肥230001

出　　处：《北华大学学报（自然科学版）》2018年第5期697-700,共4页Journal of Beihua University（Natural Science）

基　　金：安徽省高校自然科学研究重点项目(KJ2017A876)

摘　　要：为了通过在现实生活中拍摄的单幅图片生成不同能见度下的雾效模拟场景,提出一种基于图片深度的雾效模拟方法:通过单幅图片的散焦信息恢复深度图,把深度图转化成透射率,将最大能见度与大气散射系数相关联,通过大气散射模型模拟不同能见度下的雾效虚拟场景.利用此方法分别测试近景图和远景图,进行不同能见度下的雾效场景模拟.仿真结果表明,此方法简单有效,可行性高,仿真效果逼真.In order to generate fog effect virtual scene in a single real image under different visibility,a method of fog simulation based on depth information is proposed. The depth map of the image is restored by using the defocus information of a single image,to convert the depth map to the transmittance using the formula,and then associate atmospheric scattering coefficient with maximum visibility correlation,finally the simulation of fog effect virtual scene under different visibility is realized on the basis of atmospheric scattering model. This method is used to realize the fog scene simulation under different visibility for the vision map and close-range map severally.Experiments show that the algorithm generates a realistic scene fog,which can simulate fog visibility under different scenarios.

关 键 词：深度图 大气散射模型 最大能见度 雾效模拟 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]