Publication Date

Fall 2025

Degree Type

Master's Project

Degree Name

Master of Science in Computer Science (MSCS)

Department

Computer Science

First Advisor

Philip Heller

Second Advisor

Nada Attar

Third Advisor

Amith Kamath Belman

Keywords

Arabian Gulf Coral Reefs, Deep Learning, Computer Vision, Segmentation, Coral Bleaching

Abstract

Coral reefs in the Arabian Gulf are a thermally tolerant reef system compared to others globally, making them critical models for studying climate resilience in marine ecosystems [1],[2],[3],[4]. As climate change introduces stressors such as rising sea surface temperatures and ocean acidification there is growing interest in developing tools to monitor coral health over time [5],[6],[7],[8]. Deep learning techniques such as Mask R-CNN offer an automated way to detect coral in photoquadrat images commonly used in ecological surveys [9],[10],[11],[12],[13],[14],[15],[28]. This project report covers the use of Coral Vision as a Deep Learning Computer Vision tool to support Arabian Gulf reef monitoring. Using Coral Visions’ RGB histogram equalization capacities has shown to be an effective method for preprocessing images resulting in more accurate segmentation. Moreover, RGB histogram equalization has shown to be a valid method for data augmentation to supplement small data sets and resulting in more accurate and robust models for detecting corals. Finally, this report will highlight the use of Open CV packages that can be used in a post processing capacity to feature tag masked corals and assess bleaching events in the Arabian Gulf’s coral reef system.

Download

Available for download on Friday, December 18, 2026

Share

COinS