摘要
水下环境会对可见光的传播过程造成吸收和散射影响,从而引起水下图像画质模糊、颜色失真、对比度低等问题。为此,提出一种基于雾线暗通道先验的水下图像复原方法。该方法使用雾线模型进行去模糊操作,在背景光预估时,结合结构森林边缘检测方法红通道信息获得准确的背景光值,并且通过暗通道先验模型和导向滤波算法优化透射图。经图像质量评价标准的客观分析,文中算法能够有效提升画质和对比度,还原色彩信息。
The underwater environment will cause absorption and scattering effects on the propagation of visible light,resulting in problems,such as blurred image quality,color distortion and low contrast on underwater images.An underwater image restoration method based on a priori of fog-line and dark channel is proposed.The method uses the haze-line model for deblurring operation.In the background light estimation,the structure forest edge detection method and the red channel information are combined to obtain the accurate background light value,and the transmission pattern is optimized by the dark channel prior model and the guided filter algorithm.By objectively analyzing image quality evaluation standards,the algorithm can effectively improve image quality and contrast,and restore color information.
作者
周丽丽
朱佳琦
王桥桥
蒋玉红
ZHOU Lili;ZHU Jiaqi;WANG Qiaoqiao;JIANG Yuhong(School of Electronic Information and Artificial Intelligence,Shaanxi University of Science&Technology,Xi’an 710021,China;School of Electrical and Control Engineering,Shaanxi University of Science&Technology,Xi’an 710021,China)
出处
《南京邮电大学学报(自然科学版)》
北大核心
2020年第4期64-69,共6页
Journal of Nanjing University of Posts and Telecommunications:Natural Science Edition
关键词
水下图像复原
雾线
暗通道先验
结构森林
图像质量评价
underwater image restoration
haze-line
dark channel prior
structure forest
image quality evaluation