Full-Field Vietnam Vehicle Registration (Cavet) OCR

The Challenge: Moiré Patterns, Red Stamps, and Diacritics

Processing Vietnamese vehicle registration certificates (Giấy chứng nhận đăng ký xe / Cavet) presents unique optical hurdles. Images are frequently low-quality photos shared via messaging apps like Zalo, often taken of a computer screen (introducing severe moiré patterns). Furthermore, critical data fields like Owner Name (Tên chủ xe) and Address (Địa chỉ) are routinely obscured by large, red official stamps and signatures. Standard OCR engines struggle with this intense visual noise and the complex diacritics of the Vietnamese language. Learn more about our robust capabilities on our Vehicle Registration OCR product page.

Semantic Parsing & Localized Data Delivery

StructOCR is specifically engineered to conquer the complexities of Vietnamese registration documents. Our engine isolates exact text locations by bypassing red ink artifacts, background patterns, and screen glare. It uses advanced semantic Key-Value mapping to locate both universal fields (VIN, Make) and localized Vietnamese data points. Crucially, it dynamically returns a `localized.vietnam` object containing fields like `owner_name`, `address`, and `engine_number`, perfectly handling Vietnamese diacritics. Discover how this fits into broader compliance workflows in our Vehicle Registration Solutions guide.

Core Industry Applications

Integration & API Documentation

Easily integrate our full-field OCR API. For full technical details, visit our Developer API Documentation.

import requests
import base64

# 📸 Test for free with your own image upload: https://structocr.com/developers/vehicle-registration-ocr

with open("vietnam_registration.jpg", "rb") as image_file:
    base64_image = base64.b64encode(image_file.read()).decode('utf-8')

url = "https://api.structocr.com/v1/vehicle-registration"
headers = {
    "x-api-key": "YOUR_API_KEY",
    "Content-Type": "application/json"
}
payload = {
    "img": base64_image
}

response = requests.post(url, headers=headers, json=payload)
result = response.json()

if result.get('success'):
    data = result.get('data', {})
    std = data.get('standardized', {})
    print("✅ Standardized VIN:", std.get('vin'))
    print("✅ Standardized Plate:", std.get('plate', {}).get('formatted'))
    
    loc = data.get('localized', {})
    if 'vietnam' in loc:
        vn = loc['vietnam']
        print("✅ Vietnam Localized (Owner Name):", vn.get('owner_name'))
        print("✅ Vietnam Localized (Address):", vn.get('address'))

{
  "success": true,
  "data": {
    "document": {
      "country_code": "VN",
      "document_type": "VEHICLE_REGISTRATION",
      "confidence_score": 0.96
    },
    "standardized": {
      "vin": "LSVSH7C61RN020846",
      "plate": {
        "number": "29A 123.45",
        "formatted": "29A 123.45"
      },
      "make_normalized": "Honda",
      "model_raw": "Air Blade"
    },
    "localized": {
      "vietnam": {
        "owner_name": "Nguyễn Văn A",
        "address": "123 Đường Lê Lợi, Quận 1, TP Hồ Chí Minh",
        "engine_number": "JF51E-0123456"
      }
    }
  }
}