How Mean and Median Filters Clean Salt-and-Pepper Noise, Explained Visually
cneuralnetwork · x · 2026-09-11
- A quoted post shows how salt-and-pepper noise (random black and white pixels) degrades a photo, then demonstrates cleaning it with two classic filters: Mean and Median.
- Mean filtering averages neighborhood pixels to smooth noise but blurs edges; median filtering takes the neighborhood's median value, removing salt-and-pepper noise more effectively while preserving detail.
More from Research
- NYU mathematician says OpenAI pushed him to drop co-author's name after private call — eyishazyer · 2026-09-11
- NYU mathematician says OpenAI pushed to scrub collaborator from Navier-Stokes proof announcement — eyishazyer · 2026-09-11
- Arch breakdown: dropping convs for Muon optimizer, 3x3 pixel unshuffle for vision — stochasticchasm · 2026-09-11
- The 'Trickle Test': a new eval measuring whether user models disclose information gradually like humans — gharik · 2026-09-11
- Humans& Releases Persimmon, First Large-Scale Model Simulating How People Talk — niloofar_mire · 2026-09-11
- Yacine: RL can run in real time — almost nobody is seriously trying — yacineMTB · 2026-09-11