VU cloud services

Extract the document's data and validate front and back

A document assembled from one person's front and another's back passes any check that only looks at whether the reading worked. ID Analysis is a service you can consume via API to read the front and the back from a photo, identify the document type and validate that the back belongs to the document on the front.

To generate your API Key, you need a paid VU plan.

Identity documentFront
Photo of the holder on the front of the document
Same documentBack
"code": 200 · "message": "Operation end successfully.""frontBackTemplateValidation": trueMatches the front · you continue with the data

Security backed by certifications and independent testing

iBeta PAD Level 3 compliant

ISO/IEC 30107-3 · Liveness

ISO/IEC 27001:2022

Information security management

ISO/IEC 27017:2015

Security for cloud services

The problem

Reading both sides doesn't guarantee they belong to the same document

If someone combines the front of one document with the back of another, the OCR extracts data from each side and the response arrives without errors. The same happens with a photo that contains no readable document: the API responds with code 200, a different message and no data. A flow that only checks the status code will accept both cases.

  • The API responded✓ 200
  • The message is a success✓ "Operation end successfully."
  • Back data was read✓ read
  • Does the back match the front?false
Our solution

The document's data, with front and back validated.

From the front, ID Analysis extracts the printed fields, reads the MRZ and the barcode (if the document includes them), identifies the document type and crops the face. From the back, it extracts the available data and validates that it matches the front. The decision to accept the customer isn't automatic: you receive the data and the validations, and your policy defines what to accept.

Case 01

You already have another provider

There's no need to migrate your whole operation to add document reading. You can consume only this API, in parallel with your current provider.

Case 02

You receive documents from different countries

It identifies the document type from a catalog of more than 90 templates from the region. Before integrating it, you can check which fields each one includes.

Case 03

You're developing your own onboarding

You need to read the document and crop the face to compare it with the selfie. You combine this API with Face Analysis and keep control over the rest of the process.

Key features

Three data sources, cross-checked against each other

It reads the visual zone, the MRZ and the barcode when the document has them, and tells you how much they match each other, field by field.

Reads from
  • Visual zone (VIZ)The document's printed fields
  • BarcodeWhen the document has one
  • MRZThe machine-readable zone, with its check digits
Returns
  • POST /frontDocument type"id": "VU-ARG-ID-04"Against more than 90 templates from the region
  • POST /frontFace cropReady to compare with Face Analysis
  • POST /backFront and back validation"frontBackTemplateValidation": trueIs the back from the same document?
  1. Visual zone (VIZ)The document's printed fields
  2. POST /frontDocument type"id": "VU-ARG-ID-04"Against more than 90 templates from the region
  3. POST /frontFace cropReady to compare with Face Analysis
  4. BarcodeWhen the document has one
  5. MRZThe machine-readable zone, with its check digits
  6. POST /backFront and back validation"frontBackTemplateValidation": trueIs the back from the same document?

The data it returns, and how much to trust each one

Consistency across sources

It compares the MRZ, the barcode and the visual zone with each other, with a match percentage from 0 to 100 for each field. The service doesn't impose a threshold: you decide for each field, depending on the data that feeds your process.

Template catalog

You check which fields, MRZ and barcode each document includes, and run tests with anonymized sample images before using real documents.

Integration

It does not replace Identity Link when you need a full verification. You can use it on its own or combine it with other VU cloud services.

Document antispoofing (optional)

It analyzes whether the image corresponds to a real document or a reproduction, and whether the holder's photo was tampered with. The result is informative: the response arrives with code 200 and your system decides what to do.

The integration

Two calls and one decision rule.

HTTP calls from your backend: the key is never exposed in a mobile app or the browser.

ElementHow it works
POST /frontAnalyzes the front and returns the uuid that links it to the back
POST /backAnalyzes the back and cross-checks it with the front, using that uuid as relatedUuid
GET /listTemplatesThe supported documents. With /{templateId}/details, which fields to expect from each one, and with /{templateId}/images, anonymized sample images
GET /versionChecks connectivity. The status code is evaluated, not the body
AuthenticationApi-Key header. Each service has its own key, per environment
Base URLVU provides it along with the key, depending on the environment
Inputinfo as a business reference and image in base64. On the back, also relatedUuid
Outputuuid, id (the document type), front and back with the data, crossMatchConfidence, code and message
RequestPOST /back
// Header: Api-Key: <your service key>
{
  "info": "onboarding-12345",
  "image": "<photo of the back in base64, no prefix>",
  "relatedUuid": "<the uuid returned by POST /front>"
}
  • "frontBackTemplateValidation": trueThe back matches the front. You review the data's consistency and decide.
  • "frontBackTemplateValidation": falseIt's the back of another document. You don't accept.
  • "message": "Cant get document fields"On the front: the document couldn't be read. You request another capture.
The decision ruleYour backend
// Front: accepting requires all three conditions.
const frontOk =
    status === 200 &&
    body.message === "Operation end successfully." &&
    body.id !== "unknown";

// Back: on top of the above.
const backOk = body.back.frontBackTemplateValidation === true;
// And you check body.crossMatchConfidence before trusting the data.

A 200 doesn't mean extracted data

A photo with no readable document also responds 200, with the message "Cant get document fields" and the document type as "unknown". And a back that doesn't match the front responds 200 with "Operation end successfully.". If you decide on the status code alone, you accept unreadable documents and documents assembled from the sides of two different documents.

Image requirements

The photo decides how many fields can be read.

AspectValue
ResolutionMinimum 480 × 480. Up to 1920 px per side improves the reading*
FramingThe whole document, laid flat and with no glare
FormatJPEG or PNG. If JPEG, quality 90 or higher
Encodingbase64 with no data: prefix and no line breaks

*Below 480 × 480 the API returns an error. Above 1920 px the image is downscaled automatically, so sending more doesn't improve the reading.

Glare empties fieldsGlare on the laminate and fingers covering data are the leading cause of empty fields and unreadable documents. Guide the capture before sending it: whole document, laid flat and with no glare.
How you start

Self-service: connect this API with no middlemen

With a paid plan, you generate the key from VU's panel and call the service from your backend.

  1. 02

    You confirm accesswith the method the panel offers you

  2. 03

    You open Cloud Servicesin the side menu

  3. 04

    You choose ID Analysisamong the available services

  4. 05

    You generate the API keywith a paid plan, for the environment you're going to use

  5. 06

    You copy the key and call GET /versionto confirm the network. The key is tested with the first /front

Generate your API Key

To generate your API Key, you need a paid VU plan.

Frequently asked questions

No. You can consume it independently, with its own key, over whichever flow you choose. It coexists with the provider you already have installed.
ID Analysis has a catalog of more than 90 document templates from the region. With GET /listTemplates you can check the supported documents, the fields, the MRZ and the barcode each one includes, and get anonymized sample images to run tests before using real documents.
You can analyze only the front with POST /front and get the data from that side. In that case, the front-and-back validation doesn't apply.
Yes, if you request it. Document antispoofing analyzes whether the image corresponds to a real document or a reproduction, and whether the holder's photo was tampered with. The result is informative: the response arrives with code 200 and your system decides what to do.
In the Api-Key header of each request. Each service has its own key for each environment, and VU provides it along with the corresponding base URL.
If there's no readable document in the photo, the API responds with code 200, with a different message and no data. That's why your flow shouldn't rely on the status code alone: also check code, message and that the expected data is present.

You already have the documents. You still need to read and cross-check them.

Generate your key and test with a document you already know, or with the catalog's sample images.

Generate your API Key

To generate your API Key, you need a paid VU plan.