SOUNDTILES: Mapping the invisible Underwater

Project reference: 2016 PROD 00130
: Universitat de Girona
Budget: 98.750,00 €
Duration: 25/07/17 – 25/01/19
Research team: Pere Ridao Rodriguez, Natàlia Hurtós Vilarnau


Activities in the underwater domain are essential to the economy and to society in general, holding high expectations for future growth. Eventually, most underwater operations require to have an overview of the submerged area. Reasons are diverse, e.g. the oil and gas industries, the hydroelectric dam companies and the aquaculture farms need to regularly inspect their submerged facilities, the security authorities are required to search underwater for evidence or victims, the environmental agencies want to assess the presence of a determined specie in the seabottom, etc. The traditional solution consists in gathering images with a video camera and often, apply photomosaicing techniques to have a full map of the area (in a similar way as the well-known Google Maps).


However, very frequently, these activities take place in environments with poor water visibility conditions. A new generation of imaging sonars, namely the two-dimensional Forward-Looking Sonars (FLS), have produced a recent breakthrough in the market. These devices, also termed acoustic cameras, are able to directly deliver 2D images of unprecedented clarity and resolution compared to other available sonar technologies. Then, analogoulsy to photomosaicing techniques, individual sonar frames can be registered to offer overviews of underwater areas regardless of the visibility conditions. Given the lack of solutions in the market to generate maps with these kind of data, a PhD thesis from the University of Girona developed a mosaicing pipeline tailored to the peculiarities of FLS images, providing a pioneering way to map underwater environments in low visibility conditions.


Many potential sectors could benefit of such technology, saving vast amounts of time and money in their periodic inspections, maintenance routines, studies or investigations, and in general contributing towards more efficient and safe production/operation in underwater environments.


Thus, the goal of this proposal is to turn the research algorithms into a user-oriented software product, which we dub SOUNDTILES, for automatic processing of high quality sonar mosaics. The budget will be targeted to two main actions: first, to cover the programming effort required to build a user software package with a graphical interface and an intuitive workflow, that can be easily mastered by a non-specialist. Second, to determine a suitable commercialization model through a spin-off company in order to bring the product to the market. We believe that by adapting the existing technology into a user-oriented product with the appropriate commercialization model we can bring the ability to generate underwater maps in zero visibility conditions to a vast number of underwater-related markets.