基于色彩补偿与暗通道的水下图像复原
首发时间:2019-12-16
摘要:水下与雾天图像成像机理相似,但由于水下光学衰减特性,传统去雾算法针对水下颜色退化问题基本无效,因此本文提出基于色彩补偿与暗通道的水下图像复原方法。首先针对色偏问题,通过Retinex与白平衡算法对水下各颜色通道进行色彩补偿。其次,提出先对图像低通滤波再结合水下暗通道先验估计全局背景光。最后,利用水体背景颜色与散射系数的关系修正透射率。结果表明,本文算法能有效恢复图像颜色信息并去除后向散射光,且水下彩色图像质量评价、信息熵等指标整体优于对比算法。
关键词: 图像复原 色彩补偿 低通滤波 暗通道先验 散射系数
For information in English, please click here
Underwater Image Restoration Based on Color Compensation and Dark Channel
Abstract:The imaging mechanism of underwater and foggy images is similar, but due to the underwater optical attenuation characteristics, the traditional defogging algorithm for underwater color degradation is basically invalid, so this paper proposes an underwater image restoration method based on color compensation and dark channel. Firstly, for the problem of color deviation, Retinex and white balance algorithm are used to compensate the color of each underwater color channel. Secondly, the global background light is estimated by the low-pass filter and the dark channel. Finally, the relationship between background color and scattering coefficient is used to modify the transmittance. The results show that the algorithm can effectively recover the color information of the image and remove the backscattered light, and the underwater color image quality evaluation, information entropy and other indicators are better than the comparison algorithm as a whole.
Keywords: Image restoration Color compensation Low-pass filtering Dark channel prior Scattering coefficient
引用
No.****
动态公开评议
共计0人参与
勘误表
基于色彩补偿与暗通道的水下图像复原
评论
全部评论0/1000