基于景深的单幅图像快速去雾算法

在雾、霾天气条件下,由于大气粒子的散射作用导致采集的图像质量严重下降,给计算机视觉系统成像带来极大不便。针对上述问题,提出一种基于景深的单幅图像快速去雾算法。通过暗通道对图像处理,得到图像传输图,从而简化大气散射模型,利用景深信息估计大气散射模型,得到边缘突变景深关系比,通过近水平方向上的景深关系比值优化传输图,利用双边滤波对景物边缘进行处理。从3个颜色通道出发,降低波长对景深估计的影响。实验结果证明,该算法能够恢复景物边缘的细节对比度,有效提高图像的清晰度与视见度。

Quick Defogging Algorithm for Single Image Based on Field Depth

Imaging in the atmosphere is often degraded by scattering due to atmospheric particles such as haze, fog, and mist, which brings great inconvenience to the imaging of computer vision system. This paper presents a method based on the depth of field of the single image defogging algorithm quickly to solve this problem. Through dark channel prior on image processing, it gets image transmission chart, thereby simplifies the atmospheric scattering model. It uses depth of field information estimation of atmospheric scattering model to get the depth relationships of scene edge. Through nearly horizontal death ratio improving transmission chart, it uses bilateral filtering for image edge processing. Starting from three color channels reduces the effect of wavelength on estimating depth of field. Experimental results show that this algorithm can effectively recover scene edge contrast details, and effectively improves image clarity, resulting in a great improvement in image visibility.