Corals Classification Dataset

This enhanced dataset facilitates the binary classification of coral health, distinguishing between healthy and bleached corals—a crucial indicator in coral reef ecosystem management. Initially sourced to overcome the lack of image diversity and pre-sorted training categories, this dataset now features:

• Total Images: 10,292, meticulously augmented and categorized into training, testing, and validation sets.
• Image Distribution: Training (7,384 images), Testing (923 images), Validation (985 images).
• Data Splitting: Strategically split in an 80-10-10 ratio to optimize machine learning outcomes.

Purpose and Impact:

This dataset aids in monitoring coral health, assessing environmental impacts, and enhancing conservation strategies. By providing a clear differentiation between bleached and healthy corals, it supports scientific research, conservation planning, and raises awareness about coral ecosystems’ vulnerabilities and the necessity for sustainable interventions.