Corals, vital for marine ecosystems, are facing threats from climate change-induced thermal stress. To monitor their health, the Marine Ecology Laboratory (LECOM) at the Federal University of Rio Grande do Norte (UFRB) initiated the “#DeOlhoNosCorais” project. It encourages individuals to share coral reef photos on Instagram, contributing to monitoring efforts.
Current Results
At DSBD, we employed machine learning models for image classification and semantic segmentation. The combination of these models proves effective in identifying coral species and assessing health. The U-Net Pix2Pix model achieves an 86% pixel-level accuracy in semantic segmentation. The results suggest that this approach can streamline and improve the efficiency of coral reef image analysis, offering potential for broader applications across various regions by incorporating additional datasets.