Vietnam Passport OCR with the StructOCR Node.js SDK

Why Vietnamese Passport OCR is Difficult in Node.js

Building a reliable passport OCR pipeline from scratch is non-trivial, especially for Vietnamese documents. Open-source Node.js wrappers for tools like Tesseract frequently fail on the complex tonal marks and specialized characters (like ă, đ, ê, ô, ơ, ư) unique to the Vietnamese language. Furthermore, modern Vietnamese e-passports (blue cover) feature intricate background artwork, such as traditional lotus motifs, which confuse standard text recognition algorithms. Maintaining custom RegEx patterns in JavaScript to parse the Machine Readable Zone (MRZ) while correcting these optical errors is brittle, leading to high failure rates and manual review queues.

Enterprise-Grade Extraction with the StructOCR SDK

StructOCR simplifies document processing in your JavaScript ecosystem by replacing complex pipelines with a single async API call. Our service leverages pre-trained Deep Learning models optimized specifically for Vietnamese identity documents and their unique typography. Our passport mrz ocr api automatically handles perspective correction, denoising, and glare removal. Instead of returning raw, unstructured text strings, the StructOCR Node.js SDK provides standardized JSON output with validated fields. This capability is crucial for applications managing international borders, as it eliminates the need for manual parsing, delivering production-ready data directly to your Express, NestJS, or Next.js applications. For the fastest Node.js integration, install the official StructOCR SDK and see the Node.js SDK documentation.

Production Use Cases

• Digital Onboarding (e-KYC): Reduce drop-off rates by pre-filling user data from Vietnamese Passports into your web or mobile apps in under 2 seconds.
• Travel & Aviation Apps: Seamlessly integrate with Node.js backends for e-Visa processing or automated check-in systems at hubs like Noi Bai (HAN) and Tan Son Nhat (SGN).
• Financial Compliance: Ensure strict compliance with State Bank of Vietnam (SBV) and AML regulations by automatically verifying identity documents.

Vietnam Passport OCR with the Node.js SDK

The official Node.js SDK handles file buffers, API communication, and structured passport results. Use the SDK directly or open the TypeScript + Express tab for a server endpoint. Keep your API key in STRUCTOCR_API_KEY.

Prerequisite: `npm install structocr`; set the server-side `STRUCTOCR_API_KEY` environment variable.

Prefer another stack? Open the Python SDK + FastAPI integration.

const StructOCR = require('structocr');

// Initialize the client with your API key
const apiKey = process.env.STRUCTOCR_API_KEY;
if (!apiKey) throw new Error('STRUCTOCR_API_KEY is not configured');
const client = new StructOCR(apiKey);

async function scanVietnamPassport() {
  // Note: Supports JPG, PNG, WebP, PDF (Max 4.5MB)
  const imagePath = './vietnam_passport_sample.jpg';

  try {
    console.log(`Processing Vietnamese passport: ${imagePath}...`);

    // The SDK handles file reading and the API call seamlessly
    const result = await client.scanPassport(imagePath);

    if (result.success && result.data) {
      const data = result.data;
      console.log('✅ Extraction Successful!');
      
      // Basic Identity (MRZ + VIZ)
      console.log(`Passport #: ${data.passport_number}`);
      console.log(`Name:       ${data.surname} ${data.given_names}`);
      console.log(`Nation:     ${data.nationality} (${data.country_code})`);
      
      // Visual Zone Specifics
      console.log(`Birth Place:${data.place_of_birth}`);
      console.log(`Authority:  ${data.issuing_authority}`);
      
      // Dates
      console.log(`DOB:        ${data.date_of_birth} (${data.sex})`);
      console.log(`Expiry:     ${data.date_of_expiry}`);

    } else {
      console.error('❌ Extraction Failed:', result.error || 'Unknown Error');
    }
  } catch (error) {
    console.error('An unexpected error occurred:', error.message);
  }
}

scanVietnamPassport();

Sample JSON Output

The Node.js SDK returns an object matching this normalized JSON structure.

{
  "success": true,
  "data": {
    "type": "passport",
    "country_code": "VNM",
    "nationality": "VNM",
    "passport_number": "C1234567",
    "surname": "NGUYEN",
    "given_names": "VAN A",
    "sex": "M",
    "date_of_birth": "1990-09-02",
    "place_of_birth": "HA NOI",
    "date_of_issue": "2023-01-15",
    "date_of_expiry": "2033-01-15",
    "issuing_authority": "IMMIGRATION DEPARTMENT"
  }
}