Kongsberg Discovery, in collaboration with Cathx Ocean, has launched a new analysis tool.
The tool the company claims will radically reduce the time and effort required to detect and classify objects of interest in seabed surveys undertaken by Kongsberg’s HUGIN autonomous underwater vehicles (AUVs), which is expected to reduce costs for survey owners and freeing up valuable resources to focus on more important tasks.
Kongsberg HISAS scans of the seafloor contain extensive visual information that requires a highly trained eye to manually detect objects of interest.
Automating the object identification process using the novel SAS Target Assistant provides major workflow benefits regarding earlier decision-making. The application runs in the background providing consistent and reliable detection, leaving users to focus their mental capacity on the most valuable task of evaluating and classifying the potential objects of interest.
“In conventional survey analysis, the operator has to laboriously play through the scanned image data, tagging, measuring, and classifying objects of interest and, finally, building a target list that can be exported. SAS Target Assistant automatically detects and measures objects of interest and creates a target list for review/confirmation, further analysis, and export,” said Adrian Boyle, CEO of Cathx Ocean.
The tool allows users to predefine the desired characteristics (size and shape of objects) they require to rapidly identify targets matching such specifications. SAS Target Assistant requires no additional window or display/monitor, as it is fully integrated into the Reflection PMA system.
“It’s a great example of how machine learning and sensor intelligence can be used to simplify often demanding workflows, reducing the burden on human operators while also producing reliable results according to survey specifications and deadlines,” comments Sigurd Fjerdingen, VP Products in Kongsberg Discovery.
