Reproducible dataset for channel-specific ocular chromatic similarity in bird photographs

The study tested whether an image-defined iris region showed greater similarity to the ocular highlight in individual RGB channels than expected from nearby non-eye colour.</p><p dir="ltr">Six original wild-bird photographs were used as a discovery set, and six independently selected public bird photographs were used for validation.

The analysis used normalized RGB-channel values, local negative controls, permutation testing, and sensitivity analyses of region-of-interest definitions. The primary analysis settings were fixed before application to the independent validation set.</p><p dir="ltr">The archive contains derived numerical data, per-image measurements, quality-control information, discovery sensitivity-analysis results, summary tables, figures, and Python code for reproducing the principal results and figures from the deposited derived data.</p><p dir="ltr">Third-party validation photographs are not redistributed.