自然场景的高真实感雾天退化图像仿真

Simulation of High Realism Foggy Image for Natural Scene

自然场景的天气图像仿真是图形图像和计算机仿真领域中的一个重要的研究课题。首先,以自然场景的雾天图像仿真作为研究对象,深入探讨了天气导致图像退化的物理机制,提出了一种基于成像设备特性的大气点扩散函数。其次,由于场景深度是大气点扩散函数中的一个重要参数,将基于单幅图像的场景深度重建算法应用于此项研究中最后,提出了雾天图像合成的后处理算法,实现了高真实感的自然场景雾天图像仿真。仿真实验的结果证明了本算法的有效性。

The simulation of the weathered image for natural scene is very important in computer graphics and computer simulation.A method was introduced to simulate the foggy image degradation of natural scene.Firstly,a new atmosphere point spread function（APSF） was proposed including the features of imaging equipment,based on introduce of the physical process of weathered image degradation.Secondly,a depth reconstruction algorithm based on single image was applied into this research since the depth was an important parameter in the APSF.Then,the post-processing methods were introduced.Finally,the simulation results show the validity of the proposed method.