pointcoral: Point-Count Processing for Coral Photoquadrats

Imports Coral Point Count with Excel extensions (CPCe) point annotations, matches them to local photoquadrat images, validates labels and coordinates, and computes point-count percent-cover summaries. Optional crosswalks map project-specific CPCe codes to full labels, ecological categories, and machine-learning classes. Quality-control overlays, point-centered patches, sparse weak-label masks, and leakage-aware dataset splits support review and downstream local model development.

Details

The package preserves original CPCe labels and coordinates. Users remain responsible for verifying CPCe file variants, image matching, coordinate conversion, crosswalk semantics, sampling design, and the interpretation of point-count estimates. Sparse masks label only neighborhoods around sampled points and are not dense benthic annotations.

See also

Useful links:

• https://el-cordero.github.io/pointcoral/

• https://github.com/el-cordero/pointcoral

• Report bugs at https://github.com/el-cordero/pointcoral/issues

Author

Maintainer: Elvin Cordero elvin.cordero@seamountgeo.com (ORCID)