Coral Detection, Ranging, and Assessment (CDRA) algorithm-based automatic estimation of coral reef coverage.

Area of Science:

• Marine biology
• Computer vision
• Ecological monitoring

Background:

• Climate change threatens coral reef health, necessitating accurate coral coverage assessment.
• Traditional manual methods for coral coverage estimation are inefficient and labor-intensive.
• Automated processing of coral video imagery is underdeveloped, hindering comprehensive monitoring.

Purpose of the Study:

• To develop an automated system for efficient analysis of large volumes of coral reef video data.
• To create a novel algorithm for reliable and comprehensive coral reef surveys.
• To advance coral reef conservation through improved monitoring techniques.

Main Methods:

• Development of the Coral Detection, Ranging, and Assessment (CDRA) algorithm.
• Integration of You Only Look Once version 5 (YOLOv5) object detection, binocular stereo vision, and tracking algorithms.
• Analysis of continuous video segments to identify, count, and estimate size/location of coral colonies.

Main Results:

• The CDRA algorithm achieved an average accuracy of 90.7% in estimating coral coverage in controlled environments.
• Field trials validated the algorithm's effectiveness and robustness under diverse water and lighting conditions.
• The study demonstrates a significant advancement in automated coral reef detection.

Conclusions:

• The CDRA algorithm offers a reliable and automated solution for coral reef monitoring.
• This technology can transform imaging devices into autonomous sensors for comprehensive surveys.
• The findings contribute to enhanced coral reef conservation efforts globally.