• Greeshma P M
• Pallavi K G
• BanuShree T
• Pranitha T Manur
• S Vinodh Kumar

This research proposes a robust and automated approach to categories coral reefs according to their state of health, which significantly advances the cause of coral reef conservation. Researchers, environmentalists, and marine biologists can monitor and assess coral reef ecosystem health more effectively and comprehensively with the help of this suggested approach. Our study focuses on using deep learning methods to automatically identify healthy and bleached corals by extracting important features from images of coral reefs. We specifically employ the DenseNet-169 and VGG19 architectures, which have demonstrated effectiveness in image classification tasks, making them well-suited for our coral reef classification objective. Using a sizable dataset of annotated images of coral reefs, including both healthy and bleached species, we thoroughly train and fine-tune these models in order to achieve this. The trained models are then assessed using an independent test set, and performance measures including accuracy, precision, recall, and F1-score are computed to determine how effective the models are. In order to evaluate the relative effectiveness of the Dense Net 169 and VGG19 architectures in coral reef classification, we also run a comparative analysis.

Visual Geometry Group (VGG-19), DenseNet-169

"EXPLORING MACHINE LEARNING MODELS FOR CORAL REEF CLASSIFICATION FROM IMAGES", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.11, Issue 5, page no.391-397, May-2024, Available :http://www.jetir.org/papers/JETIRGG06062.pdf

"EXPLORING MACHINE LEARNING MODELS FOR CORAL REEF CLASSIFICATION FROM IMAGES", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.11, Issue 5, page no. pp391-397, May-2024, Available at : http://www.jetir.org/papers/JETIRGG06062.pdf