vCF-1: Linear Curvature-Change Denoising With Minimal Low-ℓ Bias (Medley 2025)

This record presents the vCF-1 (Variable Curvature Filter – Version 1) denoising method developed by Michael Medley. vCF-1 is a linear, prior-minimal curvature-change filter designed to remove high-frequency noise while preserving large-scale structure. It uses only first-order and second-order finite differences, with no physical scale, no statistical noise model, and no hand-tuned hyperparameters. This release includes: • Cleaned vs. raw example (single image) • The exact operator equation used • Full copy-and-paste explanation of how vCF-1 works • Ready-to-use text for scientific reproducibility This note is modest, honest, and verifiable — presenting vCF-1 exactly as it is: A stable, linear, low-bias denoiser designed for comet images, astronomical frames, and full-sky maps. All content © 2025 Michael Medley.