利用大气散射模型的图像去雾研究
首发时间:2019-04-18
摘要:为了获取清晰的去雾图像,提出一种基于暗原色先验和边界约束的单幅图像去雾算法。首先采用暗原色理论和边界约束理论分别获得天空区域和非天空区域的透射率图像。之后,采用引导滤波对透射率图进行优化处理;然后,采用改进的容差机制对天空区域的透射率做出放大处理。随后结合亮通道理论和暗原色理论得到更精确地大气光值。最后,对于天空区域和非天空区域,分别使用图像融合和直方图均衡的方法增强图像的明亮度。本文从峰值信噪比(PSNR)、均方误差(MSE)和结构相似性(SSIM)三个指标分析实验结果。指标数据表明我们方法的去雾图像失真更小,清晰度更高。
关键词: 暗原色先验 边界约束 引导滤波 亮通道先验 图像融合 容差机制
For information in English, please click here
Image Dehazing Research Using Atmospheric Scattering Model
Abstract:To obtain clear defogging images, a single image dehazing algorithm based on dark channel priori and boundary constraints is proposed. Firstly, for sky region and non-sky region, dark channel prior theory and boundary constraint theory are used to obtain transmission map respectively. Then, guided filtering is used to optimize the transmission map. Later, the improved tolerance mechanism is used to amplify the transmission in the sky region. Next, the precise atmospheric light value is obtained by combining the light channel prior theory and the dark channel prior theory. Finally, for the sky region and non-sky region, image fusion and histogram equalization are used to enhance the image brightness respectively. In this paper, the experimental results are analyzed from three indexes: peak signal to noise ratio(PSNR), mean squared error(MSE) and structural similarity(SSIM). Indicator data show that the dehazing image by our method has less distortion and higher clarity.
Keywords: Dark channel prior Boundary constraint Guided filtering Light channel prior Image fusion Tolerance mechanism
引用
No.****
同行评议
勘误表
利用大气散射模型的图像去雾研究
评论
全部评论0/1000