How to use Camera Lens Distortion Corrector
- Choose one JPEG, PNG or WebP photo. Look for a straight edge, such as a door frame, to use as a visual reference.
- Choose Output size. Use The source size, when it fits this tool to keep dimensions, or enter a custom width and height. A changed aspect ratio can stretch the photo.
- Start with k1 and k2 at 0, then adjust a coefficient in small steps while comparing the left source and right corrected preview. Settings that fold or nearly flatten the map are refused.
- Choose Transparent or White for pixels sampled beyond the source. Create the result and inspect the full downloaded PNG; the small preview cannot show every resampling artifact.
Example: Camera Lens Distortion Corrector
Use zero coefficients and matching dimensions as a baseline before trying a correction.
Options
- Radial coefficients
- Both terms adjust a centered radial map. Zero coefficients leave that map unchanged; choosing new output dimensions can still resize or stretch the image. The tool refuses maps that fold or nearly flatten instead of guessing a camera profile.
- Output size
- Custom width/height controls the output raster. Source-size mode keeps both original dimensions only when the entire image fits the output bounds. Choose the mode before judging a correction.
- Pixels beyond the source
- Choose transparent or white for positions the map cannot sample. A different fill does not recover image content that was outside the original frame.
Supported inputs and limits
- Input is one JPEG, PNG or WebP up to 64 MiB and 24 megapixels.
- Output must stay within 2 megapixels and 2,048 pixels per side, with at least 2 pixels on each side.
- Source-size mode refuses a larger source; it does not silently reduce it.
- Custom dimensions can change aspect ratio. k1 and k2 are manual coefficients from -1 to 1, further restricted to a supported monotonic map.
- There is no camera/lens profile, automatic calibration, perspective correction or reconstruction of missing detail.
More details
Resampling can soften pixels and leave empty edges.
Compare the original and downloaded result before deciding whether the adjustment helps.
Where your input is processed
This tool does not upload your inputs. See the limits for anything saved on your device.
Manual correction needs a visual reference
Use a photo with a reference you expect to be straight, and compare the zero-coefficient result with small adjustments. This tool maps output positions back into the source using two radial terms centered on the image. It does not identify your lens, estimate its optical center or remove camera perspective. A result that looks better is a manual visual adjustment, not evidence of calibrated lens correction.
Questions about Camera Lens Distortion Corrector
Will the tool detect my lens model?
No. You enter the radial coefficients and review the result.
Why can the edges become uncovered?
A distortion map can sample outside the source image. Review the selected border treatment.
Can correction restore missing detail?
No. Resampling uses the pixels already in the uploaded image.