YOLO-Driven Automated Detection of Coral Reef Health Indicators in Underwater Imagery
- Version
- Download 1
- File Size 423.53 KB
- Download
YOLO-Driven Automated Detection of Coral Reef Health Indicators in Underwater Imagery
Elavarasi Kesavan, Full Stack QA Architect, Cognizant,
elavarasikmk@gmail.com , ORCID: https://orcid.org/0009-0008-3844-0286
Abstract
Coral reefs, critical marine ecosystems, play a vital role in supporting biodiversity, protecting coastlines, and stimulating global economic activity. However, these ecosystems are increasingly threatened by climate change, unsustainable human activities, and pollution. Consequently, the success of conservation initiatives hinges on the capacity for timely and accurate diagnoses of coral reef health issues—a necessity for effective management. This study showcases an innovative approach, employing the YOLO (You Only Look Once) algorithm, a prominent deep learning object detection framework, to develop an automated, real-time coral reef health and disease detection system with significant potential in marine biology [28]. Because it processes images both rapidly and with precision, YOLO presents an ideal solution for spotting underwater coral diseases, like coral bleaching, tissue loss, and color variations, from photo and video data. Our proposed system is grounded in a meticulously compiled and annotated dataset of coral reef images, encompassing both thriving coral formations and indicators of illness; this thoroughness ensures the algorithm's strength [29]. Furthermore, the YOLO algorithm is specifically tuned to cope with the challenges presented by underwater environments, such as variable lighting, low contrast, and common obstructions. Through in-depth testing, the system has achieved notable precision and recall rates in differentiating between healthy and diseased corals. This performance allows for the continuous, real-time monitoring of coral reefs, providing a robust tool for comprehensive evaluation. As a result, marine biologists and conservationists can respond quickly to safeguard and rejuvenate these delicate ecosystems. This work underscores the potentially transformative power of artificial intelligence—YOLO, specifically—for enhancing coral reef monitoring and conservation efforts. Moving forward, research will concentrate on expanding the dataset, further enhancing detection accuracy, and integrating the system with autonomous underwater vehicles (AUVs) to facilitate extensive coral reef health evaluations.
DOI: 10.55041/ISJEM00071