摘要
受到复杂成像环境影响,光学视觉系统获取到的水下图像普遍存在对比度低、模糊和颜色失真等问题.为此,本文提出一种基于修正散射模型的水下图像复原算法.首先,深入分析光在水下的吸收衰减特性,在简化大气散射模型的基础上,将水体背景光融入到模型的直接衰减项;其次,考虑到水下红光迅速衰减,采用红通道的逆通道对其进行补偿;然后,使用基于四叉树的分级搜索算法估计水体背景光值;最后,在修正的成像模型基础上,结合水下暗通道先验信息估计介质透射率进而复原水下图像.实验结果表明,本文算法水下复原后的图像色彩自然,能有效恢复出远景区域的细节信息,图像对比度、色度和饱和度的综合评价指标整体优于对比算法,适用于不同类型的水下退化图像.
Due to the influence of complex imaging environment, there exist many problems in the underwater images acquired by optical vision system, such as low contrast, blur and color distortion. To solve this problem, an image restoration algorithm based on the modified scattering model is proposed. Firstly, the absorption attenuation characteristics of light in water are analyzed in depth, and based on the simplified atmospheric scattering model, the background light of water body is incorporated into the direct attenuation term of the model. Secondly, the inverse channel of red channel is used to compensate for the rapid attenuation of red light. Then, the background light value of water body is estimated by the hierarchical search algorithm based on quadtree. Finally, the medium transmittance is estimated on the basis of the modified imaging model by combining the underwater dark channel prior, and the underwater image is restored. The experimental results show that the restored underwater image displays natural color and can effectively restore the details of the far scene in image. The comprehensive evaluation index of image contrast, color and saturation is better than the contrast algorithm, and the proposed algorithm is suitable for different types of underwater degraded images.
作者
林森
白莹
李文涛
唐延东
LIN Sen;BAI Ying;LI Wentao;TANG Yandong(School of Electronic and Information Engineering,Liaoning Technical University,Huludao 125105,China;State Key Laboratory of Robotics,Shenyang Institute of Automation,Chinese Academy of Sciences,Shenyang 110016,China;Institutes for Robotics and Intelligent Manufacturing,Chinese Academy of Sciences,Shenyang 110169,China)
出处
《机器人》
EI
CSCD
北大核心
2020年第4期427-435,447,共10页
Robot
基金
国家自然科学基金(91648118,61473280)
辽宁省重点研发计划(2019JH2/10100014)
辽宁省教育厅科研项目(LJ2019JL022)
辽宁省自然科学基金指导计划(2019-ZD-0038)。
关键词
水下机器人
图像复原
成像模型
红通道补偿
暗通道先验
underwater robot
image restoration
imaging model
red channel compensation
dark channel prior
