Fish Detection in Seagrass Ecosystem using Masked-Otsu in HSV Color Space

Seagrass ecosystems are coastal with high species diversity, especially fish. Fish diversity determines the abundance of communities based on number species. Detection fish directly (in-situ) and conventionally by catching them requires more energy, costs, relatively needs time. Therefore a computer vision method is needed that can detect well using underwater images. The detection model used Masked-Otsu Thresholding, HSV color space closing techniques in morphological operations. dataset form 130 images, divided into 80% training data 20% testing data. test results showed accuracy value 0.92, Precision 0.84, Sensitivity 0.93, F1 Score 0.88. With these results, could seagrass ecosystem.