Image Quality and Blur Checker

Measure a bounded blur estimate and dark/light pixel fractions for manual image review.

On-device No sign-up Free to use

Measure image texture controls

  1. 1 Add files
  2. 2 Arrange/Options
  3. 3 Download

Drop your files here

or choose them from your device

Up to 20 images; 24 megapixels and 128 MiB combined. Each file up to 64 MiB. Every image is measured on its own small raster and all are ranked together.

    ?

    An uncalibrated cutoff for sorting the list. Images below it are flagged for review. It compares neighboring-pixel change, so a plain background scores low and a sharp photo of something flat can too. It cannot certify quality.

    Result

    Showing an example.

    An 80 × 60 PNG with a solid red left half and solid green right half, measured at the default cutoff of 100.

    One row with Laplacian variance 421.17, marked At or above cutoff. Dark, light and transparent pixel fractions are all 0.0%. Downloads: image-quality-review.json and image-texture-thumbnails.png.

    Loading the tool. Your result will appear here.

    https://usefreetools.net/image-quality-blur-checker · 20 November 2026

    Was this helpful?

    How to use Image Quality and Blur Checker

    1. Choose up to 20 JPEG, PNG or WebP images within the combined limits. Use images with similar content and scale when comparing measurements.
    2. Set Texture review cutoff. This is the only adjustable threshold; dark/light pixel boundaries are fixed.
    3. Read the ranked table alongside the numbered thumbnails. Rank is the sorted position; Input number identifies the original file-list position.
    4. Inspect the original images at their intended size. Download the JSON record if you need the measurements and chosen cutoff for a later comparison.

    Example: Image Quality and Blur Checker

    An 80 × 60 PNG with a solid red left half and solid green right half, measured at the default cutoff of 100.

    You add
    One 80 × 60 PNG: left half #ff0000, right half #00ff00, all pixels opaque. Texture review cutoff: 100.
    You get
    One row with Laplacian variance 421.17, marked At or above cutoff. Dark, light and transparent pixel fractions are all 0.0%. Downloads: image-quality-review.json and image-texture-thumbnails.png.

    Options

    Threshold
    Texture review cutoff is your rule for measured Laplacian variance. It cannot certify sharpness or quality; a smooth scene can have a low value without blur.
    Sampling
    Images are reduced to the documented longest side before measurement. Scale, texture, noise and transparency can change the result.

    Supported inputs and limits

    • Up to 20 JPEG, PNG or WebP files, 64 MiB each, 128 MiB and 24 megapixels combined.
    • Each measurement raster has a longest side of at most 512 pixels and must provide a usable neighborhood of at least 3 × 3 pixels.
    • Alpha below 242/255 is excluded; a Laplacian sample is omitted if it or any of its four neighbors is excluded.
    • The metric is population variance of the four-neighbor Laplacian on gamma-encoded grayscale.
    • Dark is below 30 and light above 225 on the 0–255 gray scale; those boundaries are fixed.
    More details

    The entered cutoff is an uncalibrated review aid, not proof of blur or quality.

    Noise, texture, sampling and alpha change the numbers.

    Thumbnails label input positions, while table rows sort by decreasing metric.

    Where your input is processed

    This tool does not upload your inputs. See the limits for anything saved on your device.

    Compare like images under the same conditions

    A noisy photo can have a high edge metric without useful detail. A smooth subject can have a low value despite being in focus. Use the measurement to order a manual review, then inspect the original at its intended size.

    Questions about Image Quality and Blur Checker

    Does a low metric always mean an out-of-focus photo?

    No. A smooth sky or other textureless subject can also have little edge variation.

    Can I compare values from differently sized images?

    Sampling and scale affect the measure. Compare results under the same stated conditions and inspect the images.

    Do dark pixels prove underexposure?

    No. A dark subject can be intentional. The fraction is a review cue, not a camera diagnosis.

    Page updated · Project managed by Tony Hines · Report a problem with Image Quality and Blur Checker