Extract Passport Data in C# with IronOCR

How Do I Extract Passport Data in C#?

As an example, we will use a passport image as input to showcase the functionality of IronOCR. After loading the image using OcrInput, you can utilize the ReadPassport method to identify and extract information from the passport. This method returns an OcrPassportResult object, which contains properties such as GivenNames, Country, PassportNumber, Surname, DateOfBirth, and DateOfExpiry. All members of the PassportInfo object are strings.

The ReadPassport method is part of IronOCR's specialized document reading capabilities, which also includes methods for reading license plates, MICR cheques, and other structured documents. The method uses advanced computer vision techniques to locate and extract the MRZ (Machine Readable Zone) area automatically.

What Passport Image Format Should I Use?

IronOCR supports various image formats including JPG, PNG, TIFF, and BMP. For optimal results, ensure your passport image has adequate resolution (at least 300 DPI) and proper lighting. The DPI settings can be adjusted if working with lower quality scans.

:path=/static-assets/ocr/content-code-examples/how-to/read-passport-read-passport.cs

using IronOcr;
using System;

// Instantiate OCR engine
var ocr = new IronTesseract();

using var inputPassport = new OcrInput();

inputPassport.LoadImage("passport.jpg");

// Perform OCR
OcrPassportResult result = ocr.ReadPassport(inputPassport);

// Output passport information
Console.WriteLine(result.PassportInfo.GivenNames);
Console.WriteLine(result.PassportInfo.Country);
Console.WriteLine(result.PassportInfo.PassportNumber);
Console.WriteLine(result.PassportInfo.Surname);
Console.WriteLine(result.PassportInfo.DateOfBirth);
Console.WriteLine(result.PassportInfo.DateOfExpiry);

Imports IronOcr
Imports System

' Instantiate OCR engine
Private ocr = New IronTesseract()

Private inputPassport = New OcrInput()

inputPassport.LoadImage("passport.jpg")

' Perform OCR
Dim result As OcrPassportResult = ocr.ReadPassport(inputPassport)

' Output passport information
Console.WriteLine(result.PassportInfo.GivenNames)
Console.WriteLine(result.PassportInfo.Country)
Console.WriteLine(result.PassportInfo.PassportNumber)
Console.WriteLine(result.PassportInfo.Surname)
Console.WriteLine(result.PassportInfo.DateOfBirth)
Console.WriteLine(result.PassportInfo.DateOfExpiry)

What Output Can I Expect from ReadPassport?

The extracted data is returned in a structured format that makes it easy to integrate with existing systems. The OcrResult class provides comprehensive access to all extracted information.

We then access the PassportInfo data member obtained from the OcrPassportResult object. The extraction process automatically handles various passport layouts and formats, providing consistent results across different issuing countries.

• GivenNames: A property of PassportInfo returns the given names of the passport input as a string. This corresponds to the first MRZ data row, positions 4 to 44.
• Country: A property of PassportInfo returns the country of the passport input as a string. This corresponds to the first MRZ data row, positions 2 to 3. The returned string spells out the full name of the issuing country instead of the abbreviation. In our example, USA returns 'United States of America'.
• PassportNumber: A property of PassportInfo returns the passport number of the passport input as a string. This corresponds to the second MRZ data row, positions 1 to 9.
• Surname: A property of PassportInfo returns the passport input's surname as a string. This corresponds to the first MRZ data row, positions 4 to 44.
• DateOfBirth: A property of PassportInfo returns the passport input's date of birth as a string in the format YYYY-MM-DD. This corresponds to the second MRZ data row, positions 14 to 19.
• DateOfExpiry: A property member of PassportInfo returns the passport input's date of expiry as a string in the format YYYY-MM-DD. This corresponds to the second MRZ data row, positions 22 to 27.

---

What MRZ Information Can I Extract from Passports?

IronOCR reads the MRZ information contained at the bottom two rows of any passport that follows the standard of the (International Civil Aviation Organization) ICAO. The MRZ data consists of two data rows, with each set of positions containing unique information. Here's a brief table showing which information corresponds to the index of the row.

The MRZ parsing functionality in IronOCR is designed to handle variations in print quality and image orientation. For challenging documents, you can apply image correction filters to improve recognition accuracy.

What Does the MRZ Section Look Like?

The MRZ is typically located at the bottom of the passport page and consists of two lines of standardized text. Understanding the MRZ structure helps in troubleshooting extraction issues and validating results.

First Row

Position	Field	Description
1	Document Type	Typically 'P' for passport
2-3	Issuing Country	Three-letter country code (ISO 3166-1 alpha-3)
4-44	Surname and Given Names	Surname followed by '<<' and then given names separated by '<'

Second Row

Position	Field	Description
1-9	Passport Number	Unique passport number
10	Check Digit (Passport Number)	Check digit for the passport number
11-13	Nationality	Three-letter nationality code (ISO 3166-1 alpha-3)
14-19	Date of Birth	Date of birth in YYMMDD format
20	Check Digit (Date of Birth)	Check digit for the date of birth
21	Sex	Gender ('M' for male, 'F' for female, 'X' for unspecified)
22-27	Date of Expiry	Expiry date in YYMMDD format
28	Check Digit (Date of Expiry)	Check digit for the date of expiry
29-42	Personal Number	Optional personal number (usually national ID number)
43	Check Digit (Personal Number)	Check digit for the personal number
44	Check Digit (Composite)	Overall check digit

---

How Can I Debug and Verify Passport Extraction Results?

We can also verify the results from IronOCR by obtaining the raw extracted text from the passport image and the confidence level to confirm whether the extracted information is accurate. Using the example from above, we can access the Confidence and Text properties of the OcrPassportResult object.

For debugging purposes, you might want to highlight text regions to visually verify what areas were recognized. This feature is particularly useful when troubleshooting extraction issues or optimizing scan regions.

:path=/static-assets/ocr/content-code-examples/how-to/read-passport-debug.cs

using IronOcr;
using System;

// Instantiate OCR engine
var ocr = new IronTesseract();

using var inputPassport = new OcrInput();

inputPassport.LoadImage("passport.jpg");

// Perform OCR
OcrPassportResult result = ocr.ReadPassport(inputPassport);

// Output Confidence level and raw extracted text
Console.WriteLine(result.Confidence);
Console.WriteLine(result.Text);

Imports IronOcr
Imports System

' Instantiate OCR engine
Private ocr = New IronTesseract()

Private inputPassport = New OcrInput()

inputPassport.LoadImage("passport.jpg")

' Perform OCR
Dim result As OcrPassportResult = ocr.ReadPassport(inputPassport)

' Output Confidence level and raw extracted text
Console.WriteLine(result.Confidence)
Console.WriteLine(result.Text)

• Confidence: The Confidence property from OcrPassportResult is a float indicating the OCR statistical accuracy confidence as an average of every character. This float is lower if the passport image is blurry or contains other information. One is the highest and most confident, and zero is the lowest and least confident. For production applications, consider implementing result confidence thresholds to ensure data quality.
• Text: The Text property from OcrPassportResult contains the raw, unparsed text extracted from the passport image. Developers can use this in unit tests to verify the extracted text of the passport image. For advanced scenarios, you can export results in hOCR format for further analysis.

Best Practices for Passport Scanning Applications

When implementing passport scanning in production environments, consider these additional factors:

1. Image Quality: Ensure input images meet minimum quality standards. The Filter Wizard can help optimize images for better recognition.

2. Performance: For high-volume processing, implement async support and consider batch processing multiple passports.

3. Security: Since passport data is sensitive, ensure proper data handling and consider integrating with secure document management systems.

4. Validation: Implement check digit validation for the extracted MRZ data to ensure accuracy. The MRZ format includes multiple check digits that can be used to verify the integrity of the extracted information.

5. Error Handling: Implement robust error handling for cases where passport images may be damaged, poorly lit, or contain non-standard formats.