Deep Learning-Based Coral Reef Segmentation: A Satellite Imagery Study of the Trincomalee District
Bibliographic record
Abstract
Coral reefs, vital yet endangered ecosystems, face rising threats from climate change and humans. Accurate assessment of coral reef health is essential for early detection of ecosystem decline and effective conservation planning. We present a deep learning framework that utilizes satellite imagery and rugosity index analysis to automate coral reef segmentation in the Trincomalee District, Sri Lanka. High-resolution Google Earth Pro images were processed to compute rugosity index values, distinguishing coral reefs and enabling the creation of a new training dataset. A U-Net model, trained on 300 annotated images augmented to 1,200 samples, achieved robust segmentation of coral reefs (Dice coefficient$=0.86$, specificity$=0.98$). Our case study found differences in reef rugosity and extent across sites exposed to varying hydrodynamic conditions, emphasizing the interplay between hydraulic forces and reef health. Applying the model to satellite images allowed us to quantify declines in reef area and structural complexity in response to increased sediment loads and wave exposure. Furthermore, as no public coral reef training datasets exist for Sri Lanka to enable automated analysis, we prepared a new dataset. These insights aid in identifying vulnerable zones and support conservation, targeted hydraulic management, and future health assessments of Trincomalee's coral reefs.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.002 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.000 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".