CoralSCOP-LAT: Labeling and analyzing tool for coral reef images
with dense semantic mask

Overview

The CoralSCOP-LAT is a coral reef image analysis and labeling tool that automatically segments and analyzes coral regions. By leveraging advanced machine learning models tailored for coral reef segmentations, it enables users to generate dense segmentation maks with minimal manual effort, enhancing both the labeling efficiency and precision of coral reef analysis

Demo

Installation

Please follow the instruction on Github to install the tool.

Workflow

The CoralSCOP-LAT workflow begins with users selecting target coral reef images for analysis. During the Project Preparation stage, the selected images are processed by CoralSCOP-LAT to extract image features and automatically segment coral regions. The outcomes of this preparation are saved in a project file, which can be reloaded by the user for subsequent analysis. Additionaly, the system enables the automated generation of visualizations and statistical reports based on the analyzed data.

Performance

CoralSCOP-LAT automatically segments coral regions with high accuracy. We evaluated its segmentation performance on coral images from 10 distinct reef sites, demonstrating strong generalization capability across diverse reef environments.

Citation

                        
@article{WONG2025103402,
    title = {CoralSCOP-LAT: Labeling and analyzing tool for coral reef images with dense semantic mask},
    journal = {Ecological Informatics},
    volume = {91},
    pages = {103402},
    year = {2025},
    issn = {1574-9541},
    doi = {https://doi.org/10.1016/j.ecoinf.2025.103402},
    url = {https://www.sciencedirect.com/science/article/pii/S157495412500411X},
    author = {Yuk Kwan Wong and Ziqiang Zheng and Mingzhe Zhang and David J. Suggett and Sai-Kit Yeung},
    keywords = {Coral reefs, Coral segmentation, Semi-automatic annotation tool, Machine learning},
}