How the face cropper works
The tool finds each face in your photo with one face detection call, then cuts a crop centred on it in your browser. There is nothing to drag or line up: change the shape or the framing and every crop updates. A photo with one face gives one crop. A group photo gives one crop per person, up to 48 from one photo, numbered to match the boxes on the photo.
A download is cut from your original file, read at up to 4,096 pixels on the longer side, so it is as sharp as the photo allows. Crops are never scaled up.
Shapes, framing and sizes
- Shape: square, 4:5 portrait, 3:4 portrait, or a circle on a transparent background. The circle is saved as PNG and the others as JPG.
- Framing: face is tight on the face, headshot takes in the hair and chin with some room, and head and shoulders is the wider cut used for profile and staff pages. Wider framings sit a little lower, so the extra room goes to the shoulders and not to empty space above the head.
- Size: the crop's own pixels, or 1,024, 512 or 256 pixels on the longer side.
A crop never reaches outside the photo. When a face is near an edge the crop slides back in, and when the photo is too small for the framing the crop shrinks, so no blank bars are added.
What people use it for
- A profile picture or avatar from a photo that has other people in it.
- Matching headshots for a team page, cut from one group photo.
- Photos for ID badges, member lists and class pages.
- Contact photos for a phone or a CRM.
- Thumbnails of each person in an event photo, with their agreement.
To see what was found before cropping, use the face detector. To hide faces instead of cutting them out, use the face blur tool.
Crop faces with the API
To crop faces in bulk, send each image to POST /v1/detect, the SightRadar equivalent of AWS Rekognition DetectFaces. It returns a box for every face as ratios of the image. Multiply by the image size, grow the box to the framing you want, and cut it out with any image library. A call costs 93 credits per photo whatever the number of faces, and stores nothing. See the detect endpoint reference.
curl -X POST "https://api.sightradar.com/v1/detect" \
-H "Authorization: Bearer $SR_API_KEY" \
-F "file=@group.jpg"
# each face: "bbox": { "left": 0.29, "top": 0.19, "width": 0.38, "height": 0.53 }Frequently asked questions
- How do I crop a face out of a photo?
- Add the photo. The tool finds each face and shows a crop centred on it. Choose a shape (square, 4:5 or 3:4 portrait, or circle) and a framing (face, headshot, or head and shoulders), then download the crop you want.
- Can it crop every face from a group photo?
- Yes, up to 48 faces from one photo. Each face gets its own crop, numbered to match the boxes on the photo. Download them one at a time, or all together as a ZIP file.
- What size are the crops?
- By default a crop keeps the photo's own pixels. Photos over 4,096 pixels on the longer side are scaled down to that first. You can also save at 1,024, 512 or 256 pixels. Crops are never scaled up, so a small face in a wide shot gives a small crop.
- Can I use it for a profile picture?
- Yes. Square and circle crops with the headshot framing suit most profile pictures. The circle is saved as a PNG with a transparent background; the other shapes are saved as JPG.
- Does it make passport photos?
- It crops to a 3:4 portrait, but it does not check passport rules such as background colour, head size or expression. Use your government's own photo checker for that.
- Are my photos stored?
- No. The photo is uploaded once so the faces can be found, processed in memory and discarded. The crops are cut in your browser and saved on your device. No photo, face crop or faceprint is kept.
- Is the face crop tool free?
- Yes. There is no sign-up and no card. A fair-use limit per visitor keeps the tool fast for everyone; if you reach it, the tool shows how long to wait.
- Can I crop faces in bulk from my own code?
- Yes. POST /v1/detect in the SightRadar API returns a box for every face in an image, in ratios of the image. Your code cuts the crops from those boxes, for one photo or a million. Detect stores nothing.
More free tools
- Face compare Upload two photos and get a similarity score, a match verdict and a plain-English reading of it.
- Look-alike test Upload a photo of each person and see how much the two faces have in common.
- Mom or dad test Upload a photo of yourself and one of each parent to see whose face scores closer to yours.
- Face blur Upload a photo and the faces in it are blurred for you. Keep the ones you want sharp, then download.
- Face pixelator Upload a photo and the faces in it are pixelated. Pick how coarse the blocks are, then download.
- Face detector Upload a photo to count the faces in it, see a numbered box on each one, and check which are clear enough to match.
- API cost calculator Price your monthly face workload on AWS Rekognition, Azure Face, Google Cloud Vision and SightRadar, operation by operation.
- Rekognition calculator Price an AWS Rekognition image workload in any region, tier by tier: face search, detection, storage and liveness.