What data fields can be extracted from an ID, passport, or driver's license?

The specific fields available depend on the document type and issuing country, but there's a fairly consistent core set that most identity documents share, plus some fields that are specific to certain document types.

Common core fields across most ID types include: full name, date of birth, sex, nationality, document number, issuing authority or country, and dates of issue and expiry. These appear, in some form, on the large majority of government-issued identity documents, whether it's a passport, a national ID card, or a driver's license.

Passports specifically also include the MRZ-derived fields discussed elsewhere in this FAQ, plus, for e-passports, chip-stored data that mirrors the printed information along with a digital signature for authenticity verification. National ID cards often include a residential address, which passports generally don't, plus sometimes a separate personal identification or tax number distinct from the document number itself.

Driver's licenses commonly include vehicle class information and any endorsements or restrictions (like a requirement to wear corrective lenses while driving), fields that don't have an equivalent on a passport or general-purpose ID card. In North America, they also typically carry a PDF417 barcode encoding much of this same data in AAMVA format, which a scanning system can decode directly rather than relying solely on OCR of the printed text.

Beyond text fields, most identity documents also yield a photograph, extractable as a cropped image for biometric comparison purposes, and often a signature field, though this is used less commonly in fully automated verification flows.

The exact combination of available fields really is document-specific, which is why a comprehensive extraction solution needs a large library of document templates covering different countries and document types, rather than assuming a single generic layout applies to every document a business might encounter.

ScanDoc's data extraction is built around this kind of broad document template coverage, extracting the full range of fields available on a given document type, core identity fields, document-specific details like vehicle endorsements, and the photo, across a wide range of countries and document formats.