Glossary · Sharpness score
What is a sharpness score for a photo?
Short answer
A sharpness score is a number that estimates how much fine edge detail a photo contains. Software calculates it by measuring how quickly brightness changes between neighbouring pixels: crisp edges score high, smooth blur scores low. It is most useful for comparing near-identical shots of the same scene, not for judging unrelated photos against each other.
Last updated
How is a sharpness score calculated?
The most common method runs an edge filter over a greyscale version of the image. The Laplacian filter, a standard tool in image processing, responds strongly wherever brightness jumps between neighbouring pixels. The spread of those responses, called the variance, becomes the score.
- Convert the photo to greyscale, since colour adds little to edge detection.
- Shrink it to a fixed size so that a 50-megapixel and a 12-megapixel photo are judged on equal terms.
- Apply the edge filter and measure how varied the result is.
- Compare that number with other frames of the same scene.
Some tools score only the centre or the detected subject, so that a deliberately blurred background does not drag the score down.
Why is a sharpness score relative, not absolute?
Scenes differ in how much detail they naturally hold. A brick wall is full of edges and scores high even when slightly soft; a foggy beach has few edges and scores low even when perfectly focused. A single fixed threshold would flag every misty morning as blurry.
| Comparison | Is the score meaningful? |
|---|---|
| Five frames of the same burst | Yes, this is what it is built for |
| A city street vs a sunset | No, the content differs too much |
| The same photo at two resolutions | Only after resizing both to one size |
That is why cleaners compare scores inside a similar-photo group rather than across the whole gallery.
What can fool a sharpness score?
- Noise. Grainy low-light photos contain many tiny brightness jumps that look like detail to a simple filter.
- Over-sharpening. Phone processing adds halos around edges, which raises the number without adding real detail.
- Text and patterns. Screenshots and printed pages score very high, which is one reason cleaners classify content before ranking it.
- Shallow depth of field. A sharp subject on a soft background may score lower than a fully sharp but dull frame.
How do photo cleaners use it?
In a cleaner, the score is one input among several, and it is rarely shown to the user. Its job is to rank, not to judge a photo's worth on its own. It flags candidates for the blurry photo category and helps choose the keeper in a burst, with exposure and framing weighed alongside it. GalleryJury, an Android cleaner launching soon on Google Play, treats sharpness as one part of picking the best frame and always shows you the group before removing anything.
Frequently asked questions
Is a higher megapixel photo always sharper?
No. More pixels can record more detail, but shake or missed focus blurs every one of them. A steady 12-megapixel shot often beats a shaky 50-megapixel one. See megapixel.
Can I see a photo's sharpness score?
Most gallery apps do not show one. Desktop tools and some camera review sites do, but for everyday use, zooming in to 100% on the subject tells you the same thing. If text or fine hair is crisp at that zoom, the frame is sharp enough.
Sources
- Files by Google Help: Delete blurry or bad pictures (checked 31 October 2026)
- OpenCV documentation: Laplace operator (checked 31 October 2026)
Related
More in Photo cleanup terms
GalleryJury is coming soon to Android
It keeps the best shot from every burst and clears the clutter around it. 50 free photos to try, then $5 for 5,000 photos or $15 for 20,000. Pay once, nothing renews. No subscription.