Abstract

An underwater optical imaging system is indispensable for many oceanic engineering tasks, yet still plagued by poor visibility conditions. The serious degradation of underwater image results from light scattering and absorption. Removal of the backscattered light is the focus issue of underwater imaging technology to improve the image visibility, particularly in turbid water. In this paper, we present an approach for underwater image recovery using structured light imaging and flood light imaging to compose a combined imaging model with which the backscatter component is completely offset. The convolutional image is obtained using the structured light scanning imaging mode where the backscatter intensity is proportional to that of the flood light image with an unknown scale parameter. An algorithm to refine the matching of the backscatter components of both the convolutional image and the flood light image is proposed. Thus, subtraction of both images gives rise the combined imaging model without the backscatter component. Consequently, image restoration is completed by employing the deconvolution process. Results of underwater experiments are given.

© 2021 Optical Society of America

Underwater image sharpening based on structure restoration and texture enhancement

Sen Lin, Kaichen Chi, Tong Wei, and Zhiyong Tao
Appl. Opt. 60(15) 4443-4454 (2021)

Underwater image restoration via depth map and illumination estimation based on a single image

Jingchun Zhou, Tongyu Yang, Wenqi Ren, Dan Zhang, and Weishi Zhang
Opt. Express 29(19) 29864-29886 (2021)

Adaptive color correction and detail restoration for underwater image enhancement

Quan Wang, Chengtao Cai, Weidong Zhang, Peitong Li, and Boyu Xin
Appl. Opt. 61(6) C46-C54 (2022)

Underwater image restoration via feature priors to estimate background light and optimized transmission map

Jingchun Zhou, Yanyun Wang, Weishi Zhang, and Chongyi Li
Opt. Express 29(18) 28228-28245 (2021)

Local linear model and restoration method of underwater images

Zijie Zhuang, Zhiguo Fan, Haihong Jin, Kaiqiang Gong, and Jin Peng
Opt. Express 30(17) 30949-30968 (2022)