The classification of underwater marine species plays an important role in marine biodiversity monitoring, ecological research, and conservation planning. However, identifying marine species from underwater images is a challenging task due to poor lighting conditions, water turbidity, background noise, and variations in species appearance. Traditional manual identification methods are time-consuming and require expert knowledge, making automated classification systems highly valuable. This project presents a comparative analysis of Machine Learning (ML) and Deep Learning (DL) techniques for the classification of underwater marine species. A real-time underwater image dataset containing 189 images across 20 different marine species is used for experimentation. Traditional machine learning models such as Support Vector Machine (SVM), K-Nearest Neighbors (KNN), and Decision Tree are evaluated alongside deep learning architectures including AlexNet, DarkNet-19, and SqueezeNet. Experimental results demonstrate that deep learning models significantly outperform traditional machine learning methods in terms of classification accuracy. Among all evaluated models, SqueezeNet achieves the highest accuracy, demonstrating its effectiveness in handling complex underwater visual patterns while maintaining computational efficiency. The study highlights the advantages of convolutional neural networks in extracting meaningful features from underwater images and emphasizes their suitability for marine species classification tasks. \\
Professor et al. (Thu,) studied this question.