基于暗原色的单一图像去雾算法的研究  

Research on single image defogging algorithm based on dark primary colors

作　　者：和晓军[1] 乔寅[1] 

机构地区：[1]沈阳理工大学信息科学与工程学院,沈阳110159

出　　处：《计算机应用研究》2014年第1期304-306,共3页Application Research of Computers

摘　　要：暗通道优先算法在处理单幅户外图像去雾方面有较好的效果,但是该算法处理的时间过长,算法速度较慢。在暗原色优先算法的基础上,提出了改进的方案,即使用指导性滤波代替暗原色优先中的软抠图算法,并增加盒子滤波以提高指导性滤波的处理速度。同时为提高图像的输出质量,采用了分块处理方法,使用3×3模板求解图像的暗原色,得到了更好的透过率t值。实验表明,该算法在提高速度的同时优化了图像输出效果,图像更加自然、真实。Dark channel prior can achieve better results when dealing with single image, but this algorithm which is slow takes a long time. This paper brought out a better method that replaced the soft matting algorithm in dark channel prior with guiding filter. Furthermore, it also employed box filter to get high speed. At the same time, in order to improve the output quality of images, it used block processing and 3×3 template to solve the dark color of the image, and obtained the better transmittance t values. According to the experiments, the above algorithm achieves not only high speed processing but also better image quality. The processed images are much natural and realistic.

关 键 词：暗原色 分块处理 透过率t值 指导性滤波 盒子滤波 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]