You don't need a pre-cropped, well-centred photograph. Give it a close-up headshot or even a full-length shot: it finds the face, straightens it if tilted, and produces a correctly framed output, ready for normalisation and ICAO verification.
Normalisation algorithms work best on an image that is already framed on the face. But the photographs that actually turn up — from a smartphone, from an archive, from a badly scanned passport photo — almost never are.
Face Finder bridges that gap. It locates the face, measures the inter-pupillary distance, corrects head tilt and produces a crop compliant with the required format, ready to hand over to PhotoNorm Base or PhotoNorm Enroll for colour normalisation, and to Digit-Ale MorpheusICAO for full ICAO 9303 / ISO-IEC 19794-5 verification.
Not a convenient simulation: the five samples include a full-length e-commerce photograph, a folded-arms pose and a portrait with the head tilted by 12.7°. They are real screens from the desktop application, with the actual processing times.
An already framed portrait, first with 5 landmarks and then with 68: you can see the point map used for centring, IED measurement and geometric checks.
Full-length shot and a non-standard pose at close range: the final crop stays centred and compliant with no manual intervention.
Head tilted by 12.7°: auto-rotation straightens it automatically before the final crop.
This page is built on the existing technical copy and on real screens from the application. Detailed system requirements, measured performance and final licensing terms are under review.
Pure normalisation for already framed images, batch capable.
Application for enrolment desks and counter staff.
SDK to embed normalisation into an existing application.
ICAO compliance verification, requirement by requirement. Real online demo.
For integration documentation or a demonstration licence.
