PlatePrep

PlatePrep turns field photographs of settlement plates into annotation-ready imagery. It is a single HTML file that runs in the browser — no installation, no server, and the photographs never leave the computer. It was built for the Bioē coating experiment and is general enough for any campaign of flat, approximately square plates.

What it does

In five guided steps PlatePrep collects the campaign metadata — much of it read from the photographs' EXIF and GPS, with an optional weather lookup — groups the photographs by plate, and rectifies each plate through a four-corner homography at a fixed physical scale. The output folder receives the named crops, a metadata table ready to import into CoralNet, the triage map, a crop manifest recording the four source corners of every crop, the physical scale embedded in each JPEG and an ImageJ calibration macro. Because the manifest records the corners, any crop can be re-derived and between-operator variability measured without re-inspecting the images.

How to cite

Galembeck, E. & Schlosser, C. F. (2026). PlatePrep: browser-based preparation of settlement-plate imagery for annotation platforms (v1.1.2) [Software]. Zenodo. https://doi.org/10.5281/zenodo.21960482

The concept DOI, https://doi.org/10.5281/zenodo.21959574, always resolves to the latest version; machine-readable metadata is in the repository's CITATION.cff file, behind GitHub's "Cite this repository" button.

The validation data, the analysis code and the regression fixture are archived separately: Galembeck, E. & Schlosser, C. F. (2026). PlatePrep — validation data, analysis code and regression fixture. Zenodo. https://doi.org/10.5281/zenodo.21960679

A methods paper describing and validating the pipeline is under submission; this section will be updated with the reference.