Coda Octopus has released the beta version of a new software package for automated seabed classification. Integrated within its Survey Engine® range, this new software automatically interprets users’ sidescan sonar data to detect and classify different seabed types.
The seabed classification module uses AI (artificial intelligence) based methods to detect and classify seabeds in both type and geopgraphical extent from sidescan sonar data. The extent boundaries are instantly visible to the user for validation and quality control and can then be exported for use in chart and map generation, direct import to the users chosen GIS platform, or for further processing. Within the survey engine, the seabed type boundaries are stored using the GeoKit feature set, allowing them to be manually edited, reported or supplemented. The boundary generation process automatically creates nodes in a way that avoids any gaps between adjacent seabed types, allowing for contiguous segmentation and reporting. The software also displays closed boundary areas as colored polygons, helping to identify and distinguish these seabed types.
Ideal geophysical survey applications include site and geohazard investigation in support of pipelay, jacket and riser installation or subsea cable laying operations. For example, ripples in the seabed that alert operators to strong and potentially dangerous currents, can now be more quickly and more repeatably identified.
Seabed Classification Beta is programmed to recognize six different types of seabed compositions. This Seabed Classification package is fully extensible and specific types can be added on request.