Step 1
1. The user photographs the passport
The SDK shows a frame and prompts: «remove the glare», «move closer». A phone photo, a scan or a webcam frame all work.
Recognition of Russian internal and international passports
Automatic passport recognition online: in a second NeuroVision reads the photo page, the registration pages and the MRZ and returns ready fields to your form, CRM or core banking system through the API.
Recognition result Russian passport · pp. 2–3
- Last name high IVANOVA
- First name and patronymic high SVETLANA VIKTOROVNA
- Series and number high 45 13 739149
- Issuing authority code high 770-121
- MRZ and visual zone high match
- Every version of the Russian passport since 1997
- International passports and passports of 195+ countries and territories
- Fields and MRZ in JSON in a second
Data is stored in Russia Russian Software Registry No. 19972 Compliance with 152-FZ
Which passports NeuroVision recognises
The model works out which passport and which page it is looking at by itself, so the user has nothing to choose. We recognise every version of the form that customers present — from the 1997 passport to the latest biometric international passport.
-
Russian citizen's passport: the photo page
Pages 2–3: full name, sex, date and place of birth, series and number, issuing authority and date, authority code. On forms with a machine-readable zone we read that too.
-
Pages with registration and stamps
Pages 5–12: address and date of registration, de-registration. Stamps and handwritten entries are recognised separately from printed text.
-
The older international passport
The five-year international passport without a chip: the photo page and the MRZ. Surname and first name in both Cyrillic and Latin.
-
The biometric international passport
The ten-year passport of the new format: the photo page and the MRZ.
-
Service, diplomatic and seafarer passports
The same fields, with their own form layouts. Also refugee passports and stateless person passports.
-
Passports of foreign nationals from 195+ countries and territories
Passports and ID cards of the CIS, Europe, Asia and the Middle East: Latin, Cyrillic, Arabic script, hieroglyphs.
Which passport data we extract
For every field the service returns a value and the model's confidence score.
Dates and numbers are brought to a single format, so the data can be written into the form straight away.
- Add this check
Russian passport
the internal passport of a Russian citizen: the photo page and pages 5–12. Not every form has an MRZ.
Personal data
- Surname, first name, patronymic
- Sex and date of birth
- Place of birth
- Holder's photo (crop)
Document details
- Series and number
- Date of issue
- Issuing authority
- Issuing authority code
- Address and date of registration
- Machine-readable zone (MRZ)
- Add this check
International passport
Russian international passports of the old and new format, service and diplomatic passports.
Personal data
- Surname, first name, patronymic
- Surname and first name in Latin
- Sex and date of birth
- Place of birth
- Citizenship
- Holder's photo (crop)
Document details
- Series and number
- Date of issue
- Issuing authority
- Valid until
- Machine-readable zone (MRZ)
- Add this check
Foreign passport
the passport of a citizen of another country;
the set of fields depends on the issuing country.Personal data
- Surname, first name, patronymic
- Surname and first name in Latin
- Sex and date of birth
- Citizenship
- Holder's photo (crop)
Document details
- Series and number
- Date of issue
- Valid until
- Machine-readable zone (MRZ)
How automatic passport recognition works
The whole cycle takes about a second per page. An operator steps in only if the model is unsure about a field.
Step 2
2. The neural network identifies the passport type and the page
The model tells a Russian internal passport from an international one and from passports of other countries, finds the photo page and the registration pages, corrects the perspective and removes glare.
Step 3
3. We extract the fields, including the registration address
Every field is read taking the typeface and language into account. The surname in an international passport is checked against its Latin transliteration.
Step 4
4. We validate the data
We cross-check the MRZ against the visual zone, calculate check digits, verify the expiry date and look for traces of editing.
Step 5
5. We return the result in JSON
Fields with a confidence score, a crop of the holder's photo and the status of the checks. The data goes straight into the form, CRM or core banking system.
Which errors recognition finds
Recognition does not merely copy out the passport — it validates the data. The result of every check comes back in the API response alongside the fields.
The MRZ does not match the visual zone
If the surname, number or date of birth was changed in one place in the document while the machine-readable zone kept the original, the service will see it.
Incorrect check digits
The MRZ contains check digits for the number, the date of birth and the expiry date. An error in any of them is a sign of a typo or a forgery.
The authority code does not match the issuing authority
We cross-check the authority code against the «Issuing authority» field. A mismatch is flagged in the response, and the application can be sent for manual review.
The passport is overdue for the holder's age
A Russian passport is replaced at 20 and 45. If the date of birth means the passport should already have been replaced, the service flags it. Validity against the Ministry of Internal Affairs database is a separate database check.
Traces of editing in the photo and fields
We look for signs of image editors, a pasted-in photo and a screen capture. In the API response this is the isEdited check in the fraud group.
The MRZ of international and Russian passports
The machine-readable zone (MRZ) is two or three lines at the bottom of the photo page: they encode the full name, the number, the dates and check digits. An international passport always has an MRZ; a Russian internal passport does not have one on every form. The service reads the MRZ and cross-checks it against the visual zone, and recognises passports without an MRZ from the visual zone. More about the MRZ
Instead of manual entry: filling in data automatically from a passport photo
An operator copies out the photo page and the registration in 3–5 minutes. A customer on the website types in 15–20 fields and often abandons the form. Recognition does the same in a second and validates the data on the spot.
| Criterion | Manual entry | NeuroVision recognition |
|---|---|---|
| Time per passport | 3–5 minutes | about a second |
| Registration address | copied out by hand | recognised from pages 5–12 |
| Typos | inevitable | doubtful fields are flagged |
| Data validation | by eye | MRZ, check digits, expiry dates, traces of editing |
| Where it works | in the office only | on the website, in the app, in the office |
| Integration | manual transfer into the system | REST API, SDK, webhooks |
Passport recognition API
One request — and the response contains the passport fields, the document type and the results of the checks. Integration through an SDK, the REST API or deployment inside your perimeter.
- Fields in JSON
- Confidence for every field
- Registration and MRZ
- Sandbox and documentation
Terms of connection
A free one-month test, a dedicated manager, 24/7 support, payment only for recognised documents.
Who needs passport recognition
Everywhere employees or customers currently copy passport data by hand: onboarding, contract paperwork, registering guests and staff.
-
Banks and MFIs
- Remote issuing of cards and loans: the form is filled in from a passport photo
- Data for customer identification under 115-FZ
-
Insurance
- An online policy from photos of the passport and licence
- Fewer errors in contracts and payouts
-
Marketplaces and online services
- Seller verification under 289-FZ
- Age verification for 18+ goods
-
Car sharing and rental
- Sign-up in a minute: passport and driving licence from photos
- With no office visit and no manual review
-
Hotels and tourism
- Registering guests by internal and international passport
- Tourist data for the tour operator without manual entry
-
HR and personnel departments
- Onboarding employees, including foreign nationals
- Passport, residence permit and work patent — into the HR system without retyping
Why a business needs passport recognition
- 01
Higher form conversion
The customer photographs the passport instead of typing in 15–20 fields.
- 02
No errors in the data
Fields are cross-checked against the MRZ, and doubtful ones are flagged.
- 03
A second instead of minutes
The operator no longer copies out the photo page and the registration.
- 04
Less fraud
Edited and expired passports are filtered out before the decision.
Results of deploying document recognition
A case study from a large Russian bank
- +15% to onboarding conversion
- down to 0.01% fraud rate
- ×10 faster document processing
How much passport recognition costs
You pay only for recognised documents, with no subscription fee. The first month is a free test.
Calculate the cost- 01
Payment for recognised pages, no subscription fee
- 02
Discounts as the volume grows
- 03
Cloud, a dedicated server or deployment inside your perimeter
Why NeuroVision
Passport recognition is part of the customer verification platform: the same models, infrastructure and security requirements.
NeuroVision — an expert in computer vision and AI
AI-OCR works as part of customer verification (KYC), trusted by the largest companies in fintech, e-commerce and transport services.
We develop identification algorithms that automate checks, detect fraud and help meet regulatory requirements.
A Russian developer
A Skolkovo resident, software in the Russian Software Registry (No. 19972), Rospatent patent No. 2023661508.
Licences and certificates
The project «Development of a multimodal system using artificial intelligence "Neuromed" for managing the diagnostic processes of a clinic on the basis of the "Neurovision" computer vision platform» is supported by the Innovation Promotion Foundation within the federal project «Domestic Solutions»
The Neurovision computer vision platform is registered with Rospatent (certificate No. 2023661508) and included in the Russian Software Registry (entry No. 19972 of 13.11.2023)
The right to use the Platform is granted under a licence agreement concluded by Neurovision with its customers
Try passport recognition for free
Leave a request — we will give you a month of test access to the API and SDK and help you embed passport recognition into your form.
Can't put your request into words? Book a call and our team will contact you at a convenient time
Book a call
The company is included in the register of IT companies and the register of personal data operators of the Russian Ministry of Digital Development.
Address: Moscow, Presnenskaya Embankment 12 — Moscow City business centre, Federation Tower. Nearest metro stations: Delovoy Tsentr and Mezhdunarodnaya.
Read also
Materials on passport checks, the MRZ and recognition accuracy — in our blog.
385 13
KYC Document Verification
How KYC Systems Detect Forged Documents
354 17
KYC Document Verification
Passport Verification Online: How Automatic Verification Works in KYC
363 21
KYC Document Verification
Why users retake documents: how to reduce the number of repeat attempts during a KYC check
510 13
KYC Document Verification
Passport machine-readable zone verification: how the MRZ helps detect forgeries in KYC
Frequent questions about passport recognition
Answers about passport types, the registration page, accuracy, data storage and integration
The Russian citizen's passport in every version of the form, including the registration pages, international passports of the old and new format, service, diplomatic and seafarer passports, and passports of citizens of 195+ countries and territories.
Yes. The service reads the address and the date of registration from pages 5–12 of a Russian passport.
No. Recognition works in the browser and in a mobile app through an SDK, or on the server through the REST API: nothing needs to be installed. For work without an internet connection there is deployment inside your perimeter.
Field recognition accuracy is 99.85%. The service returns a confidence score for every field, and doubtful fields can be sent to an operator.
Older Russian passport forms have no machine-readable zone. Such passports are recognised from the visual zone, and checks that depend on the MRZ are marked as not applicable.
Ordinary OCR returns continuous text. NeuroVision identifies the passport type and the page, sorts the data into fields, cross-checks it against the MRZ and returns JSON that is written straight into the form.
Data is processed on servers in Russia. The service complies with 152-FZ, and with deployment inside your perimeter passport data never leaves your infrastructure.
You pay for recognised documents, and the first month is a free test. We will calculate the exact price for your volume.
