Scan Machine-Readable Documents with Identity Documents for Laravel

Published on May 5th, 2021 by Paul Redmond

Identity Documents for Laravel is a package that allows you to handle documents like passports and other machine-readable documents. We originally covered Laravel Identity Documents v1, but the package has improved with the release of V2 with a complete rewrite. Package author Hergen Dillema summarized to us the rewrite in V2 of this package:

Over the past few months I've completely rewritten the package to make the api very easy to work with, to allow other optical character recognition (OCR) services besides Google Vision (even tesseract for example) and to provide face detection that extracts the passport picture.

Here's the high-level list of new features available in V2 of this package:

A new MRZ searching algorithm
Improved MRZ parsing
Compatible with more document types
Improved VIZ parsing and confidence score
Face detection to find passport picture
Configurable ORC and Face Detection services
Improved interaction with the package through oop, instead of having everything static

The new version allows you to use other OCR services besides Google Vision, such as the Tesseract Open Source OCR Engine, providing a unified API interface:

$document = new IdentityDocument($request->front, $request->back);
 
// Get the machine-readable zone (MRZ)
$mrz = $document->getMrz();
// Parsed version of the MRZ
$parsed = $document->getParsedMrz();
 
// Get the passport picture from the document
// Returns an instance of \Intervention\Image\Image
$face = $document->getFace();

Here's an example of the MRZ that this package scans and returns a parsed version:

Credit: Machine-readable passport - Wikipedia

If you'd like to get everything provided by the scanned document at once, you can use the all() method. The returned array looks like the following:

$response = IdentityDocuments::all(
    $request->front,
    $request->back
);
 
/*
[
    'type' => 'string', // TD1, TD2, TD3, MRVA, MRVB
    'mrz' => 'string', // Full MRZ
    'parsed' => [], // Array containing parsed MRZ
    'viz' => [], // Array containing parsed VIZ
    'face' => 'string', // Base64 image string
]
*/

Learn More

You can learn more about this package, get full installation instructions, and view the source code on GitHub.

The author also published Processing Identity Documents in Laravel for an in-depth article on all things related to parsing MRZ-enabled documents. The article also goes into writing a custom OCR implementation to use with the package.