Denoising Methods

Choose between Gaussian smoothing for speed or Non-Local Means for better edge preservation.

Gaussian Filter

A classic smoothing filter that replaces each pixel with a weighted average of its neighbors. Weights decrease with distance following a Gaussian (bell curve) distribution.

When to use

• Random, uniform noise (like camera sensor noise)

• When processing speed is important

• When some edge blurring is acceptable

Non-Local Means (NLM)

An advanced algorithm that finds similar patches throughout the image and averages them. This preserves edges better because similar structures are averaged together.

When to use

• Structured images with important edges

• When edge preservation is critical

• When you can afford longer processing time