基于残差UNet的水下Mueller矩阵图像去散射算法  

De-Scattering Algorithm for Underwater Mueller Matrix Images Based on Residual UNet

作　　者：李晓欢 王霞 王丛赫 张鑫 Li Xiaohuan;Wang Xia;Wang Conghe;Zhang Xin(Key Laboratory of Photoelectronic Imaging Technology and System,Ministry of Education,School of Optics and Photonics,Beijing Institute of Technology,Beijing 100081,China)

机构地区：[1]北京理工大学光电学院光电成像技术与系统教育部重点实验室,北京100081

出　　处：《光学学报》2022年第24期91-101,共11页Acta Optica Sinica

基　　金：国家自然科学基金(62031018)。

摘　　要：针对高浑浊水体环境散射严重、目标成像不清晰和对比度低的问题,在传统UNet结构的基础上,结合偏振成像理论,提出了基于残差UNet(Mu-UNet)网络的水下Mueller矩阵图像去散射算法。该算法依据Mueller矩阵图像提供的目标强度信息和偏振信息,建立不同浑浊度水下多个目标物的图像数据集。在UNet基础上引入残差模块,利用构建的Mu-UNet网络提取目标底层信息,学习标签图像特征,重建出对比度高、细节信息更丰富的水下目标复原图像。对比实验结果表明,所提算法在峰值信噪比方面较原图提升了89.40%,结构相似度提升了82.37%。相比传统算法和UNet网络,所提算法得到的复原图像具有更明显的去散射效果,细节更精细,为水下偏振清晰成像探测提供了一种新思路。Considering the problems of severe scattering, unclear target imaging, and low contrast in high-turbidity water environments, a residual Unet(Mu-UNet)-based de-scattering algorithm for underwater Mueller matrix images is proposed on the basis of the traditional UNet structure and polarization imaging theory. According to the intensity and polarization information of targets provided by Mueller matrix images, this algorithm establishes the image data sets of multiple targets under different turbidities. The residual module is introduced on the basis of UNet, and the Mu-UNet is used to extract the underlying information of the targets, which learns the characteristics of the labeled images and finally reconstructs the underwater target images with high contrast and detailed information. The comparative experimental results reveal that compared with the original image, the image restored by the proposed algorithm is improved by 89. 40% in the peak signalto-noise ratio, and the structural similarity is improved by 82. 37%. Compared with traditional algorithms and UNet, the proposed algorithm can obtain restored images with a more significant de-scattering effect and finer details, which provides a new idea for the detection and high-quality imaging of underwater polarization.

关 键 词：图像处理 水下散射 偏振成像 MUELLER矩阵 残差UNet 目标探测 

分 类 号：O438[机械工程—光学工程]