Abstract :

Coral Reefs Are Among The Most Diverse Ecosystems On Earth, Supporting Nearly 25% Of All Marine Life Despite Covering Only About 1% Of The Ocean Floor. In Addition To Their Ecological Importance, They Play A Vital Role In Human Livelihoods, With Nearly 500 Million People Depending On Them For Food, Income, And Coastal Protection. However, Coral Reefs Are Increasingly Threatened By Pollution, Climate Change, Overfishing, And Other Human Activities, Leading To Significant Degradation Worldwide. Monitoring And Identifying Coral Species Is Therefore Essential For Assessing Reef Health And Implementing Effective Conservation Strategies. Traditional Coral Identification Methods Rely Heavily On Manual Observation By Experts, Which Is Time-consuming, Labor-intensive, And Often Prone To Errors. Moreover, Underwater Imaging Presents Additional Challenges Such As Poor Visibility, Low Contrast, Varying Lighting Conditions, And The High Visual Similarity Between Different Coral Species. These Factors Make Accurate And Efficient Classification A Difficult Task. To Address These Challenges, This Project Proposes An Automated Imagebased Coral Classification System Using Advanced Deep Learning Techniques. Specifically, Vision Transformers (ViT) Are Employed To Extract Rich And Contextual Features From Underwater Images. Unlike Conventional Convolutional Neural Networks, ViT Leverages Attention Mechanisms To Capture Global Relationships Within Images, Improving Classification Performance. The Extracted Features Are Then Integrated With Powerful Machine Learning Classifiers, Including Natural Gradient Boosting (NGBoost), Histogram Gradient Boosting (HGB), Extreme Gradient Boosting (XGBoost), And Skope Rules Classification (SRC). Among These Approaches, The ViT Combined With Skope Rules Classification (ViT + SRC) Demonstrates Superior Performance, Achieving Higher Accuracy And Effectively Handling Imbalanced Datasets. The Proposed System Provides A Reliable And Efficient Solution For Large-scale Coral Species Classification, Supporting Marine Research And Conservation Efforts.

Published:

24-4-2026

Issue:

Vol. 26 No. 4 (2026)

Page Nos:

3059-3071

Section:

Articles

License:

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

How to Cite