HABs image recognition software under development to support monitoring network

SCCWRP and its partners have begun working to develop image recognition software that can autonomously identify and count harmful algal bloom (HAB) species in marine waters, part of an ongoing pilot study to develop a statewide early-warning system for coastal HABs.

Work on the software, which began this spring, will support development of a pilot network for HABs monitoring that will use special underwater microscopes to provide a continual stream of information on the types and abundance of HAB species.

Already, two Imaging FlowCytobot (IFCB) units have been successfully deployed – at Del Mar mooring in San Diego and Newport Beach Pier in Orange County – to begin capturing hourly images of phytoplankton a few meters beneath the water’s surface. A third unit is monitoring offshore waters, and five more units will be deployed during the pilot project.

The real-time stream of HABs data generated through this network has the potential to dramatically improve water-quality managers’ ability to predict when a bloom event is imminent.