Date of this Version
U.S. Government Work
In this study, we examined the use of an unmanned aerial system (UAS) to monitor fish-eating birds on catfish (Ictalurus spp.) aquaculture facilities in Mississippi, USA. We tested 2 automated computer algorithms to identify bird species using mosaicked imagery taken from a UAS platform. One algorithm identified birds based on color alone (color segmentation), and the other algorithm used shape recognition (template matching), and the results of each algorithm were compared directly to manual counts of the same imagery. We captured digital imagery of great egrets (Ardea alba), great blue herons (A. herodias), and doublecrested cormorants (Phalacrocorax auritus) on aquaculture facilities in Mississippi. When all species were combined, template matching algorithm produced an average accuracy of 0.80 (SD = 0.58), and color segmentation algorithm produced an average accuracy of 0.67 (SD = 0.67), but each was highly dependent on weather, image quality, habitat characteristics, and characteristics of the birds themselves. Egrets were successfully counted using both color segmentation and template matching. Template matching performed best for great blue herons compared to color segmentation, and neither algorithm performed well for cormorants. Although the computer-guided identification in this study was highly variable, UAS show promise as an alternative monitoring tool for birds at aquaculture facilities.
Natural Resources and Conservation Commons, Natural Resources Management and Policy Commons, Other Environmental Sciences Commons, Other Veterinary Medicine Commons, Population Biology Commons, Terrestrial and Aquatic Ecology Commons, Veterinary Infectious Diseases Commons, Veterinary Microbiology and Immunobiology Commons, Veterinary Preventive Medicine, Epidemiology, and Public Health Commons, Zoology Commons