MRZ OCR
Enhance the effectiveness of your identity document processing with Optical Character Recognition. Find and extract MRZ strings automatically and securely with Klippa.
Trusted by more than 1000 brands worldwide
Also looking for a solid partner? Give us a ring at +31 50 2111631
MRZ data extraction
An image speaks more than a thousand words. Below is an example of the three steps our OCR engine takes to automatically extract MRZ data from identity documents.
Uploading an image containing an MRZ
The first step is to provide a picture or a PDF file of an identity document that contains an MRZ. Usually this is done from a mobile app or a web application. The image will automatically be cropped, skewed and converted into a usable format.
Image to text using OCR
As soon as the picture or PDF has been received it is converted to a TXT file. In this step all the text from the document is extracted, including the MRZ that you can see on the bottom left.
Getting JSON output from the API
The Klippa Parser takes the TXT gained from the OCR in step 2 and converts it into structured JSON using machine learning and MRZ parsing. The JSON is then returned as output from the API. From here the MRZ data can easily be processed into your database. Optionally additional data like the passport picture or signature can be extracted and provided.
Reduce cost
Spend less on processing passports and other identity documents by using MRZ OCR.
Improve speed
Process MRZ strings automatically within seconds. Simplify your KYC and onboarding.
Reduce errors
Prevent manual data entry errors with high quality data extraction for MRZ.
Prevent fraud
Automatically recognise errors, duplicates and fraud.
What fields are extracted?
An MRZ consists of the name, document number, nationality, date of birth, gender and passport expiration date. Sometimes there is optional additional information. When reading MRZ these fields will be extracted and optionally the signatures and pictures can be extracted.
Frequently Asked Questions
What does it cost?Can Klippa process all types of MRZ with OCR?How does Klippa check the MRZ validity?Is Klippa MRZ processing GDPR compliant?Is there documentation for the MRZ OCR API?
What does MRZ OCR cost?
The pricing structure for MRZ recognition depends on the monthly document volumes. Both pay per use and monthly licenses are available. Get in contact with our solution architects to get an exact quote.
Can Klippa process all types of MRZ with OCR?
Yes, Klippa can extract and validate all three types or MRZ strings on passports, ID cards and other travel documents.
How does Klippa check the MRZ validity?
MRZ strings are available in three structures: Type 1, Type 2 and Type 3. Klippa only accepts MRZs that have the correct structure, font and length to ensure validity. Next to the structure we validate the content of the MRZ by checking if it consists of only A-Z and 0-9 and separators. MRZ strings contain checksums to validate the extracted data. Klippa uses these checksums to ensure that the MRZ are valid and accurate.
Is Klippa MRZ processing GDPR compliant?
All the services that Klippa offers are fully GDPR compliant. We only use ISO certified servers within the European Union for processing and a data processor agreement is in place. We do not store any of your or your customers data.
Is there documentation for the MRZ OCR API?
Klippa was founded by developers. Therefore we understand the value of a well documented API. Our documentation is created using SWAGGER and can be found via this link.
Schedule a free online demonstration
Get a clear view of how Klippa can help make your MRZ processing more efficient. A demo takes just 30 minutes.
Automated MRZ processing with OCR
An MRZ is a coded zone that is used on identity documents like passports, ID cards and sometimes even driving licenses. MRZ stands for machine readable zone. It is an international standard (ICAO 9303) developed by governments, customs offices and airlines. It is a way of securely compacting identity document information into a coded string of characters that a computer system or smartphone camera can read. Almost all passports are nowadays fitted with an MRZ.
An MRZ consists of two or three strings of characters, checksum numbers and separators on the front of passports or the backside of ID cards. For the string a special font called OCR-B is used. This font is specifically developed to make it easier for computers to read and prevents reading errors as much as possible. The checksum numbers are used to validate the extracted data that is in the MRZ and is a safety measure against fraude.
What types of MRZ are there?
There are three types of MRZ, called MRZ 1, MRZ 2 and MRZ 3. Passports are usually fitted with Type 3 and ID cards with Type 1. The Type 3 MRZ consists of two lines of each 44 characters long and consists of name, passport number, nationality, data of birth, sex and expiration data. Sometimes additional data can be added. Only A-Z and 0-9 characters are included in the MRZ and the separator. The Type 2 format is not commonly used, but spans two lines of 36 characters long. The Type 2 only consists of A-Z and 0-9 characters next to the separator. The Type 1 format spans three lines and each line is 30 characters long. Also in the Type 1 only A-Z and 0-9 are included, next to the separator.
What are the use cases of MRZ OCR?
Any use case where large amounts of identity documents have to be processed to extract information such as the name, date of birth, nationality, document number, valid through and more can benefit from MRZ OCR. Many times these are cases with regards to identity validation, customer onboarding or KYC.
Questions?
Please feel free to ask all your questions.
Get in touch by mail, phone or chat!
+31 50 2111631