AI-Driven Deep Learning: Classifying Coral Reef Health with VGG19 and ResNet50

Coral reefs are an integral component of the marine ecosystem. In recent days the coral reefs face threats because of climate change and human activity. Timely and accurate assessment of the health of the coral will ensure its conservation. This article proposes a deep learning-based method for automatically categorizing coral reef images as healthy or bleached. It utilizes advanced models like VGG19 and ResNet50 for classification, The proposed method achieves promising results in discriminating between healthy and bleached coral reefs formations. The dataset has been meticulously curated, comprising high resolution coral reef images. The performance of VGG19 and ResNet50 architecture are compared with their accuracy scores. Also, the model is evaluated with Precision, Recall, F1- Score and Support score for both the above architectures. The model achieves an accuracy score of 87.661% for VGG19 and 92.9635% for ResNet50. The proposed model highlights the potential impact of deep learning techniques in aiding conservation of the coral reefs.