An In-Depth Analysis of Machine Learning and Deep Learning Methods for the Classification of Underwater Marine Species

Effective methods for classifying undersea creatures are necessary because marine environments are essential to the biodiversity of the world. This article presents an extensive analysis of contemporary deep learning (DL) and machine learning (ML) methods tailored for the underwater classification of marine creatures. Advanced deep learning models like Alex Net, Squeeze Net, and Darknet19 are investigated alongside traditional machine learning methods like Decision trees, k-NN, and Support Vector Machines (SVM). Using criteria including accuracy, precision, recall, and F1 score, the study thoroughly assesses each algorithm's effectiveness in categorization. Results unveil nuanced distinctions, with deep learning models demonstrating notable advancements in accuracy. Squeeze Net stands out with exceptional performance, while Alex Net and Darknet19 showcase competitive capabilities. The traditional ML algorithms, though effective, exhibit varied accuracies. This review synthesizes findings to offer insights into the strengths and limitations of different approaches, providing a roadmap for researchers, marine biologists, and conservationists. The implications extend to the development of automated systems, advancing marine species identification and contributing to the broader discourse on ecological preservation. The findings presented herein aim to guide future endeavours in the realm of underwater marine organism classification, fostering advancements at the intersection of technology and marine conservation.