ID Document Verification

Document verification is the process of digitally authenticating identity documents, such as passports or driver’s licenses, to ensure their validity and protect against fraud. Veridas’ AI-driven solution automates this process by analyzing key security features and comparing the data to global standards. With support for over 99% of identity documents worldwide, this technology checks for signs of forgery or manipulation without requiring manual review. The system guarantees fast, accurate verification, ensuring that documents are genuine and providing a secure, seamless user experience​

Yes, Veridas can quickly add new documents to its coverage. With a dedicated Technology and Legal team that continuously monitors official state publications for new identity documents, Veridas can include these documents in less than a week. Since the solution is fully proprietary, all that’s required is a template of the new document, such as one from the PRADO database. Once added, the document benefits from the full range of Veridas’ capabilities, including classification, OCR, and validation features. This ensures up-to-date and comprehensive global document coverage.

The validation of an identity document depends on two factors.

• The document itself: There are documents with a higher level of security than others. Therefore, the more security measures a document implements, the more reliable its digital verification can be.
• Capture technology: Capture technology with native components (iOS & Android) allows Veridas to offer a higher level of security than capture from the web environment. This is due to the restrictions imposed by web technology in terms of resolution, focus control, etc. In addition, a high-end device offers greater capture performance than a low-end one. The lowest step is the desktop camera, which generally offers very poor resolutions (VGA or HD).

Listed below are the checks the product performs to verify that the document is authentic.

• 1. Verification of the expiration date of the document and the age of majority. Veridas verifies that the document has not yet expired and that the holder is of legal age.
• 2. Verification of MRZ checksums. Veridas reads and verifies the MRZ check digits of the ID document.
• 3. Verification of consistency between printed data (VIZ) and MRZ. Veridas verifies the degree of similarity or consistency of the data written in the document (VIZ) against the data printed in the MRZ.
• 4. The Veridas solution detects black and white photocopies. 
• 5. The Veridas solution detects the presence of colored photocopies of documents on different types of paper (office paper, photographic paper, laminated documents, polycarbonate, etc.).
• 6. Detection of documents displayed on a screen. The system is capable of detecting the presence of different types of screens: smartphones, tablets and desktop computers.
• 7. Detection of impersonation and alteration of the printed photo in the document. For example, situations where a photo is superimposed on the original photo of the document.
• 8. Detection of documents generated with generative AI (deepfakes) and digital manipulations. The system is capable of detecting total or partial image manipulations using programs such as Photoshop, as well as artificially generated documents.
• 9. Decoding of bidimensional codes (OBR) of different types: Code 128, Code 39, Barcode, DataMatrix, QR and PDF417. The comparison is made between the data printed in the document (VIZ and MRZ) with the data stored in the previous bidimensional codes to verify their consistency.
• 10. NFC reading. The system has technology to read the NFC chip of any passport in the world, as well as national identity documents and residence permits. Comparison between the data printed on the document (VIZ and MRZ) and the data stored on the chip is carried out to verify consistency. Additionally, a biometric comparison is made between the photo printed on the document and the photo stored on the chip. In this way, it can determine if there is manipulation in the printed photo.

With over 10 years of experience and millions of documents analyzed, Veridas has detected several common types of fraud. These include:

Yes, Veridas uses NFC technology in both Android and iOS environments to read the NFC chips embedded in passports, national identity documents, and residence permits. This technology extracts all personal data stored on the chip, such as the identification number, name, date of birth, and more. Using NFC ensures that the data is error-free, as it bypasses the need for OCR. Veridas also compares the data on the chip with the information printed on the document (VIZ and MRZ) for consistency and conducts biometric verification between the printed and chip-stored photos to detect potential tampering.

Yes, Veridas can process previously captured images, even if the documents are in different orientations or on various types of backgrounds. The engine is capable of accurately processing documents regardless of the background, including light or dark homogeneous backgrounds, non-homogeneous backgrounds with or without patterns, and even out-of-focus backgrounds. This flexibility allows clients to use images already captured in a wide range of environments for document verification.

Detailed technical documentation is available to explain the different scenarios in which these images can be processed.

Yes, Veridas allows the processing of up to three documents during the same identity verification process. This is useful in scenarios where more than one document is required, such as verifying both an ID and a driver’s license for remote vehicle rentals, or when combining a document with fewer security measures with a more secure document like a passport. Veridas verifies the authenticity of each document individually and performs a biometric comparison between the photos on the documents to ensure they belong to the same person. If a document lacks a printed photo, the client can implement additional checks, such as comparing identification numbers or names.