First per-image PCA decomposition of Kodak suite reveals deliberate curation

Kodak PCD0992 Statistical Profile Series

Per-Image PCA and Inter-Channel Redundancy Analysis of the Kodak Lossless True Color Image Suite

Baetzel, J. (2026)

Overview

This repository contains the first published per-image statistical characterization of all 24 images in the Kodak Lossless True Color Image Suite (PCD0992). Each image is documented as a two-page reference data sheet reporting the complete inter-channel redundancy structure: covariance matrix, eigendecomposition, Pearson correlations, spatial autocorrelation, and derived classification metrics.

All statistics were computed directly from the 8-bit RGB pixel arrays of the standard 768x512 base-resolution PNG distribution. No subjective descriptions appear in any profile. All redundancy classifications are generated programmatically from the computed metrics using fixed thresholds documented in the methodology.

Related Research

Parent Paper: Baetzel, J. (2026). Statistical Characterization of Inter-Channel Redundancy Structure in the Kodak Lossless True Color Image Suite. Per-Image Principal Component Decomposition of PCD0992.

Focus: Theoretical framework establishing the first complete per-image PCA decomposition of the Kodak suite. Documents the dimensionality spectrum, blue channel independence range, eigenvector loading patterns, and evidence for deliberate curation across the 24-image collection.

Availability: Included in this repository ( baetzel_2026_kodak_pca_characterization.pdf ).

... continue reading