Design and construction of a multimodal perception system based on LiDAR for underwater applications

Although the oceans and seas conform most of the earth surface, and present many different topics of interest for scientific research, most of them haven’t been properly catalogued and mapped. Historically, sonar and camera techniques have been used for mapping purposes. Underwater LiDAR has emerged during the last few years as a novel approach to solve the problems traditional techniques face. While cameras can create 3D maps, they face significant limitations underwater due to reduced visibility in turbid waters and lack of texture, resulting in limited effective range. Sonar systems, while capable of longer ranges, face an inherent trade-off: their resolution is limited by the wavelength of the signal, where shorter wavelengths provide better angular resolution but reduced range, and longer wavelengths allow greater range but at the cost of lower resolution. The current commercial systems are expensive and rather scarce, due to several challenges, both optical and non-optical. The work here presented describes a multimodal underwater LiDAR system, integrating a LiDAR system, with RGB and polarimetric cameras. The system is designed to operate at depths of up to 10 meters, with potential for deeper operations through future adaptations of the housing structure. Applications for such systems span across multiple fields, including detailed mapping of submerged archaeological sites, inspection of underwater infrastructure, and marine life studies. With this system, we aim to acquire comprehensive data across different Jerlov waters and scenarios, enabling new research in underwater visibility and detection algorithms.