Documents and MRZ

Read an identity document from a photo: what it is, every printed and MRZ field, the portrait, and whether it passes its checks.

Read a document#

documents.read is the full read. Send the front, and the back when the document has one.

curl -X POST https://verisec.dev/v1/idv/read \
  -H "Authorization: Bearer $VERISEC_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "image_front": "iVBORw0KGgoAAAANSUhEUgAAAAEAAAABCAYAAAAfFcSJAAAADUlEQVR42mNkYPhfDwAChwGA60e6kgAAAABJRU5ErkJggg==",
    "scenario": "FullProcess"
  }'

The answer holds everything the reader found: the document type, the printed and MRZ fields, the portrait, and the security and image-quality checks. When you only need one part, ask for that part:

Parse an MRZ#

The MRZ is the machine-readable block at the bottom of a passport or ID. mrz.parse takes the lines as text, with no image, and recomputes every check digit:

curl -X POST https://verisec.dev/v1/mrz/parse \
  -H "Authorization: Bearer $VERISEC_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "lines": [
      "P<UTOERIKSSON<<ANNA<MARIA<<<<<<<<<<<<<<<<<<<",
      "L898902C36UTO7408122F1204159ZE184226B<<<<<10"
    ]
  }'

mrz.read finds the MRZ in a photo, reads it and parses it in one call, and mrz.recognize only reads the lines.

Proof of address#

proofOfAddress.analyze reads a utility bill, bank statement or letter and checks it against the name, address and country you expect, with a check that it's recent and hasn't been edited.

Every document and MRZ endpoint#