Image formats

All endpoints accept JPEG and PNG images, sent as binary parts of a multipart/form-data request.

Limits

• Size: at most 25 MiB per image part.
• Dimensions and aspect ratio: validated at runtime, after the image is decoded.

Image parts

The part names depend on the endpoint.

• /api/v3/verify and /api/v3/extract:
  • imageFirstSide
  • imageSecondSide
  • imageBarcode.
• /api/v3/extract-barcode: image, which is the only image part the endpoint accepts.

Which sides you scan is determined by which parts you send, not by which endpoint you call.

Resolution

The resolution should be at least 1080p (1920×1080 px). Smaller resolutions are accepted, but won't produce optimal results. Larger resolutions are also accepted, but won't improve the scanning process significantly, and going above 4K isn't recommended in any case.

There is no need to downscale images before sending them.

JPEG is recommended; use a high quality compression setting. Don't use lossy compression when passing images captured by an SDK: encode them as lossless PNG instead.

Keep in mind that the extraction engine is optimized for images taken with mobile phone cameras. Images taken with web cameras, which usually have much lower resolutions, might not perform as well.

Image quality

Format and size are necessary but not sufficient. Framing, lighting, and focus determine whether extraction succeeds, and they're assessed at runtime and reported back in the imageAssessment object of the response.

For the full guidance on positioning, framing, lighting, and blur, see Meet image requirements.

Returned images

Images returned in the response are base64-encoded JPEG strings, regardless of the format you sent.

The images object can carry firstSideCropped, secondSideCropped, face, signature, and barcode, depending on what you enabled in the configuration.