Understanding Checks & Check Statuses

Document Verification V2 runs a set of checks on every uploaded document. This page explains each check that appears in the Verification Details section of the Trust Validation Results page and in the verificationStatuses, securityChecks and imageQuality arrays of the API response. For the checks supported by each document, see Supported documents.

How the response is structured

The checks perform during Document Verification V2 is shown in the table below:

Title	Response key	What it contains
Overall Verification Checks Statuses	`verificationStatuses[]`	List of checks performed and status (`verificationType` + `verificationStatus`).
Document Security Checks	`securityChecks[]`	The individual tamper, forgery and liveness checks behind `DocumentSecurityCheck`.
Document Image Quality	`imageQuality[]`	The individual capture-quality checks behind `DocumentQualityCheck`.
Document Type	`documentType`	Contains information about the document that was submitted
Document Extracted Data	`extractedData`	Contains the OCR values for each field on the document, plus the document images and extracted portraits.

Example of Document Verification V2 response

{
  "verificationStatuses": [
    {
      "verificationType": "DocumentTypeCheck",
      "verificationStatus": "Passed"
    },
    {
      "verificationType": "DocumentSecurityCheck",
      "verificationStatus": "Passed"
    },
    {
      "verificationType": "DocumentExpirationCheck",
      "verificationStatus": "Passed"
    },
    {
      "verificationType": "DocumentQualityCheck",
      "verificationStatus": "Failed"
    },
    {
      "verificationType": "DocumentTextExtraction",
      "verificationStatus": "Passed"
    },
    {
      "verificationType": "MinAgeRequirement",
      "verificationStatus": "Passed"
    }
  ],
  "extractedData": {
    "images": [],
    "textFields": []
  },
  "documentType": {
    "documentName": "Indonesia - Id Card (2011) #2",
    "country": "Indonesia",
    "documentYear": "2011",
    "documentType": "identity_card",
    "filteredCountry": null,
    "filteredDocumentTypes": null,
    "documentTypeMatchesFilter": null
  },
  "imageQuality": [
    {
      "imageQualityCheckType": "ImageFocus",
      "result": "Passed"
    },
    {
      "imageQualityCheckType": "Potraits",
      "result": "Passed"
    },
    {
      "imageQualityCheckType": "Bounds",
      "result": "Passed"
    },
    {
      "imageQualityCheckType": "Perspective",
      "result": "Failed"
    },
    {
      "imageQualityCheckType": "ImageResolution",
      "result": "Passed"
    },
    {
      "imageQualityCheckType": "ImageColorness",
      "result": "Passed"
    }
  ],
  "securityChecks": [
    {
      "securityCheckType": "DocumentImagePatterns",
      "result": "Passed"
    },
    {
      "securityCheckType": "PhotoForgery",
      "result": "Passed"
    },
    {
      "securityCheckType": "HumanFacePhoto",
      "result": "Passed"
    },
    {
      "securityCheckType": "LivenessCheck",
      "result": "Passed"
    },
    {
      "securityCheckType": "BlackAndWhiteCopyCheck",
      "result": "Passed"
    },
    {
      "securityCheckType": "Occlusion",
      "result": "Not Performed"
    }
  ]
}

Check statuses

Every check returns one of three values.

Status	Meaning	Affects the verification results
`Passed`	The check ran and found no issue.	Yes, positively
`Failed`	The check ran and found an issue.	Yes, see How the final verdict is decided
`Not Performed`	The check did not run.	No

Not Performed is not a failure. It means the check was not run, because:

the check is not supported for this document type or version
the information the check needs is missing from the document (for example, no expiry date or no portrait area)
image quality was too low for the check to run
only one side was uploaded and the check needs the other side
the check is not enabled for your template (for example, MinAgeRequirement)

1. Document Type Check

verificationType: "DocumentTypeCheck" Checks whether the document was recognized as a known, valid version of a document in the reference database (country, document type and design year). If Document Type Filtering is enabled, it also matches the document type filtering based on allowed country and document types.

Status	Meaning
`Passed`	The document matched a known, accepted template.
`Failed`	The document could not be matched to a valid template. It may be an unsupported, deprecated or altered document.

The recognized document is returned in documentType:

"documentType": {
  "documentName": "Philippines - Driving License (2017)",
  "country": "Philippines",
  "countryCode": "PHL",
  "documentYear": "2017",
  "isDeprecatedDocument": false,
  "documentTypeMatchesFilter": true,
  "filteredCountry": "Philippines",
  "filteredDocumentTypes": ["driving_license"]
}

documentTypeMatchesFilter shows the result of the document type filter set up in Trust Flow Tool Settings:

Value	Meaning
`true`	The submitted document matches an allowed document type.
`false`	The submitted document is not an accepted document type.
`null`	No document type filter is set up.

If the uploaded document does not match the allowed document type in filteredDocumentTypes, the entire verification will be rejected. Example of rejected document type response due to not matching the document type filter:

"documentType": {
    "documentName": "Philippines - Criminal police permit #1",
    "country": "Philippines",
    "documentYear": "",
    "documentType": "nbi",
    "filteredCountry": "ID",
    "filteredDocumentTypes": [
      "indonesia_ktp"
    ],
    "documentTypeMatchesFilter": false
  }

2. Document Security Check

verificationType: "DocumentSecurityCheck" Checks whether the document is genuine and has not been forged. It looks for signs of counterfeiting, a replaced or pasted-on photo, a photocopy, a document shown on a screen instead of held in front of the camera, and parts of the document being covered. Types of security checks:

`securityCheckType`	What it checks	Common failure causes
`DocumentImagePatterns`	Compares the shape, size, colour and location of the document’s security patterns against the reference stored in the Documents Database. Uses white, UV and IR light.	Counterfeit or altered document, reprinted document, design that doesn’t match the detected version
`PhotoForgery`	Whether the holder’s photo was replaced, altered or attached incorrectly (printed or glued on, as this document type specifies). Looks for the edges of a pasted-on photo.	Photo substitution, a photo pasted or printed over the original
`HumanFacePhoto`	Whether a human face is present in the document’s photo area.	Photo area blank, covered or unreadable, or contains something other than a face
`LivenessCheck`	Whether a physical document was captured, rather than an image of it shown on a screen or replayed in another way.	Photo of a screen, screenshot, digital image re-captured from a device
`BlackAndWhiteCopyCheck`	Whether the document is a black-and-white photocopy, which lacks the original’s colour and security features.	Greyscale photocopy or printout
`Occlusion`	Whether any part of the document (text, photo, signature, barcode, security features or edges) is covered.	Fingers over the card, stickers, objects lying on the document

3. Document Quality Check

verificationType: "DocumentQualityCheck" Groups the capture-quality checks below. These check the photo, not the document, so a failure usually means the user should take the picture again. The parent status is derived the same way as for the security check. A quality failure on its own leads to Review Needed, not Rejected.

`imageQualityCheckType`	What it checks	Typical failure cause
`ImageFocus`	Whether the image is sharp enough and not blurred, evaluated by edge clarity and contrast.	Camera didn’t focus, hand moved
`Potraits`	Whether a face can be detected on the document. This checks that a portrait is present, not that it is genuine.	Portrait covered, blurred or hit by glare
`Bounds`	Whether the whole document, including all four corners, is in the image.	Cropped edge or corner
`Perspective`	Whether the document was photographed at too steep an angle.	Camera not parallel to the document
`ImageResolution`	Whether the document’s pixel density (DPI) is above the required minimum.	Document too far from the camera, low-resolution camera
`ImageColorness`	Whether the image contains colour or is effectively greyscale.	Greyscale scan, black-and-white filter

4. Document Text Check

verificationType: "DocumentTextExtraction" Checks the text read from the document (OCR). Each field is checked in two ways:

Source validity: each value is checked on its own: check digits, date and number formats, and consistency between pages.
Source comparison: when a field appears in more than one place on the document, the values must match. For example, the document number in the MRZ against the printed (VISUAL) number, or the barcode against the printed text.

Status	Meaning
`Passed`	Text was extracted and no field failed validation.
`Failed`	At least one field failed validation. See `textExtractionFailures`.
`Not Performed`	No text could be extracted.

Many national ID cards have no MRZ or barcode, so there is no second source to compare against. In that case the check returns Passed as long as text was extracted and no field was flagged. It does not return Not Performed just because there was nothing to compare.

When a field has values from several sources, textFields contains one entry per source, tagged with source:

{
  "fieldName": "Document Number",
  "value": "A1234567",
  "source": "MRZ",
  "status": "Failed"
},
{
  "fieldName": "Document Number",
  "value": "A1234561",
  "source": "VISUAL",
  "status": "Failed"
}

Failure reasons

The reason of failure is displayed in extractedData.textExtractionFailures.reasons[].type

`reasons[].type`	Meaning
`SourceValidity`	The value from the named source failed its own validation (check digit, format, and so on).
`SourceComparison`	Values from two sources don’t match each other.
`Unspecified`	The field was flagged as failed, but no specific reason was given.

A text failure on its own leads to Review Needed, because it is usually caused by a poor capture rather than by fraud.

5. Minimum Age Requirement

verificationType: "MinAgeRequirement"

This entry is only included when Restrict Age is turned on in your template and the minimum age is above 0.

Calculates the holder’s current age from the Date of Birth field. If the date can’t be read, it uses the document’s own Age field.

Status	Meaning
`Passed`	The holder is at or above the minimum age.
`Failed`	The holder is below the minimum age. The whole verification is Rejected.

6. Document Expiration Check

verificationType: "DocumentExpirationCheck" Checks the document’s Date of Expiry against the current date.

Status	Meaning
`Passed`	The document is not expired.
`Failed`	The document is expired.
`Not Performed`	The document has no expiry date, or it could not be read.

How the final verdict is decided

The overall result (verificationStatusCode, also sent in the webhook) is decided in this order. The first rule that matches wins.

Condition	Result
Document doesn’t match your Document Type Filter (any document in the request)	Rejected
Holder is below the minimum age (any document)	Rejected
Any security check is `Failed`	Rejected
Holder matched your custom watchlist	Review Needed
Checks failed, and all the failures are in `DocumentQualityCheck` and/or `DocumentTextExtraction`	Review Needed
Any other check failed (for example, `DocumentTypeCheck` or `DocumentExpirationCheck`)	Rejected
All checks passed	Verified
The result is inconclusive	Review Needed

Document-specific adjustments

Some genuine documents always trip a particular check because of how they are made (for example, a photo glued onto thin card stock, or a card that is always printed in black and white). For these documents, IDmeta overrides the check before the result is stored. The API response, the dashboard and the webhook therefore always show the same result.

Key takeaways

Not Performed never counts against a document.
A failed security check always rejects the verification.
Quality and text failures on their own lead to Review Needed. Ask the user to retake the photo.
On two-sided documents, a failure on either side counts.

Trust Flows

Custom Forms

Document Verification

Document Verification V2

Biometric Verification

Internal Blacklists

Philippines Philsys Check

Data Types

Data handling and privacy

Stateless API Sandbox

Understanding Checks & Check Statuses

How the response is structured

Check statuses

1. Document Type Check

2. Document Security Check

3. Document Quality Check

4. Document Text Check

Failure reasons

5. Minimum Age Requirement

6. Document Expiration Check

How the final verdict is decided

Document-specific adjustments

Key takeaways

​How the response is structured

​Check statuses

​1. Document Type Check

​2. Document Security Check

​3. Document Quality Check

​4. Document Text Check

​Failure reasons

​5. Minimum Age Requirement

​6. Document Expiration Check

​How the final verdict is decided

​Document-specific adjustments

​Key takeaways

How the response is structured

Check statuses

1. Document Type Check

2. Document Security Check

3. Document Quality Check

4. Document Text Check

Failure reasons

5. Minimum Age Requirement

6. Document Expiration Check

How the final verdict is decided

Document-specific adjustments

Key takeaways