Paper No. 51-3
Presentation Time: 8:30 AM-5:00 PM
INVASIVE ALIEN ALGAE DETECTION IN MAUNALUA BAY, ALONG PAIKO BEACH, OAHU, USING UNMANNED AERIAL VEHICLE (UAV) IMAGERY
Unmanned aerial vehicles (UAVs) have the potential to revolutionize the way that we map shallow water (<20m) benthic habitats. The fine-resolution (<0.5m) details provided by low altitude UAV imagery makes it possible to extract very specific information including substrate type and species specific benthic cover. These are essential data for coastal monitoring efforts, ecosystem modeling, change detection and planning decisions. This project demonstrates the usefulness of UAVs to detect and map invasive alien algae in Maunalua Bay, along Paiko Beach, Oahu. This region was once home to a 523-acre fishpond, but land-use and watershed changes have resulted in radical community shifts from a coral dominated to an invasive macroalgae dominated system. Imagery was collected using a DJI Mavic Pro. An orthomosaic map was constructed using Agisoft PhotoScan and then classified using eCognition and Matlab softwares. This presentation will show the results of two classification methods: (1) object based image analysis and (2) the maximum likelihood classification method and their respective ability to accurately classify Gracilaria salicornia (gorilla ogo), Avrainvillea amadelpha (leather mudweed), Acanthophora spicifera (prickly seaweed), Halophila hawaiiana (hawaiian seagrass), sand, silt, and uncolonized hardbottom.