How to use Image Noise Reduction
- Choose an image and set its bounded output size.
- Choose a 3-by-3 or 5-by-5 median and a blend amount.
- Compare the exported pixels with the source and check small details.
Example: Image Noise Reduction
A local worked example using the inputs and settings below for image noise reduction.
You add
Image: ["noise.png (366 bytes)"]
Median neighborhood: "3"
Filter blend: "1"
Longest output side (px): "1000"
You get
Filtered a 80 by 60 image with a 3 by 3 median and 100% blend.
median-noise-reduced.png (80 × 60 pixels)
Options
- Neighborhood
- A 3- or 5-pixel median neighborhood replaces local values under the stated channel policy. Larger neighborhoods can remove thin detail.
- Blend
- Blend controls how much of the filtered result replaces the original. Use a lower amount when full filtering smooths meaningful texture.
Supported inputs and limits
JPEG, PNG and WebP within 24 megapixels of input; processing/output is capped at 1 megapixel. Median kernels are bounded to 3 or 5 pixels with an explicit blend. No AI restoration. Fine texture, sharp edges and small text can be smoothed or removed.
Where your input is processed
This tool processes your input in this browser. Your text and files are not uploaded to UseFreeTools. Check this tool's limits for anything it may save on your device.
A removed bright pixel may have been detail
The filter does not distinguish sensor noise from every real feature. Compare small text, hair and sharp edges at the intended output size. The bounded processing raster may also be smaller than the input.
Questions about Image Noise Reduction
Will filtering restore lost image detail?
No. It replaces or blends existing pixels.
Can filtering blur small text?
Yes. Compare the original and result at a useful zoom before keeping the export.
What does the blend control do?
It controls how much filtered output is mixed with the original image.