Underwater imaging plays a key role in marine observation and exploration, supporting various applications such as object surveillance, geographic surveys, resource exploration, and marine ecosystem protection. However, because of the complex nature of the underwater environment, raw captured images are degraded by physical processes such as light absorption and scattering, which results in color deviation, low contrast, and uneven illumination. Restoration and enhancement of underwater images are essential for improving the accuracy and reliability of autonomous underwater vehicles (AUVs). Due to the limited communication bandwidth and high latency of the underwater environment, processing tasks must be executed on the devices to ensure timely responses to important events. On-device processing reduces reliance on high-latency acoustic communication links; however, it also imposes more constraints on computational power, memory, and energy consumption. Although numerous image processing algorithms have been developed, research on their implementation on embedded platforms remains relatively limited in the literature.
Joint supervision : the students will be supervised by two teachers/researchers from the Unversity of Brest and the University of Split.
In this internship, the objective is to perform a comprehensive survey and analysis of state-of-the-art techniques for deep-learning-based underwater image enhancement. The algorithms are evaluated and compared using three primary criteria: (1) output image quality, (2) real-time performance, and (3) object detection accuracy.
Underwater image enhancement and evaluation - Deep learning based underwater image enhancement - Embedded Development - selected techniques are implemented and tested on the Raspberry Pi 5, which is used in the open-source underwater camera KOSMOS.
Funded by the European Union. Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or the European Education and Culture Executive Agency (EACEA). Neither the European Union nor EACEA can be held responsible for them.
