As is well known, the acquisition and processing of underwater images is extensively affected by the effect of the aquatic medium on light propagation. The combined effect of absorption and scattering degrade the imagery in a much more significant way than the in-air imaging counterpart, leading to loss of color and contrast, and image blurriness. Capturing images underwater is essentially a short-range process, thus requiring special attention regarding the need to position the camera close to the scene or phenomena being observed. Further challenges arise when artificial illumination is needed, which amplifies back-scattering effects and creates strongly varying illumination conditions. Nonetheless, when compared to other forms of underwater sensing, optical imagery can provide unmatched levels of information, not only in terms of resolution, but also in terms of texture and color.

This Special Issue is aimed at bringing together new solutions for sensing and extraction of useful information from underwater imagery. We focus our interest on manuscripts thoroughly describing new vision-based systems to be used in highly unstructured and dynamic environments as well as innovative and efficient methods to process the data gathered in these scenarios. Both original research articles and reviews are welcome.

Original research papers should preferably not rely on processing information from public datasets but describe complete solutions to specific underwater applications, including sensor systems, fundamental methods, and experimental results. Manuscripts can alternatively focus on presenting new (annotated) datasets gathered by novel vision-based sensors and used in marine science and related applications, thus contributing to future benchmarking.
Reviews, presenting an analytical up-to-date overview of the state-of-the-art, would also be appropriate, provided they incorporate some quantitative and qualitative scoring of the exposed solutions through publicly available data.

We invite submissions from all areas of computer vision and image analysis relevant for, or applied to, underwater image sensing and analysis. Topics of interest include but are not limited to:

Marine imaging from drones;
New underwater optical sensor designs;
Underwater image enhancement;
Multispectral and hyperspectral sensing and calibration;
Underwater mapping in 3D and over time;
Physical models of reflectance and light transport;
Autonomous underwater navigation;
Optical sensing for autonomous underwater manipulation;
Underwater laser scanning and lidar;
Detection and monitoring of marine life;
Object tracking;
Automatic video annotation and summarization;
Classification, detection, segmentation using CNNs and deep learning;
Other forms of context-aware machine learning and image understanding.

If you have suggestions that you would like to discuss beforehand, please feel free to contact us. We look forward to your participation in this Special Issue.