Number Plate Recognition API Integration Guide

Integrating a number plate recognition API into your application doesn't have to be complicated. This guide walks you through the entire process — from obtaining API credentials to handling responses in production.

1. Get Your API Key

Start by creating a free account on NPR API. Once registered, navigate to your dashboard to generate an API key. The free Developer plan includes 1,000 requests per month — enough to build and test your integration.

2. Understand the Endpoint

NPR API exposes a single primary endpoint for plate recognition:

You can submit images as a base64-encoded string in the JSON body, as a multipart file upload, or by providing a publicly accessible URL.

3. Submit an Image

Here's a minimal example using cURL:

curl -X POST https://api.npr-api.com/v1/read \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"image": "BASE64_ENCODED_IMAGE"}'

4. Handle the Response

A successful response returns structured JSON with plate data:

{
  "results": [{
    "plate": "AB12 CDE",
    "confidence": 0.997,
    "region": "gb",
    "vehicle": {
      "make": "BMW",
      "model": "3 Series",
      "colour": "black"
    }
  }],
  "processing_time_ms": 42
}

Key fields to work with:

• plate — The recognised number plate text.
• confidence — A score between 0 and 1 indicating recognition certainty.
• region — The detected country/region of the plate format.
• vehicle — Optional vehicle metadata (available on Startup plan and above).

5. Error Handling

Your integration should handle these common scenarios:

• No plate found — The results array will be empty. This typically means the image doesn't contain a visible plate.
• Low confidence — Set a threshold (e.g. 0.85) and treat results below it as uncertain.
• Rate limiting — If you receive a 429 status, implement exponential backoff.
• Invalid image — Ensure images are JPEG, PNG, or WebP and under 10MB.

6. Best Practices

• Image quality matters — Higher resolution images yield better accuracy. Aim for at least 100 pixels across the plate width.
• Crop when possible — If you know the approximate plate location, cropping the image reduces processing time.
• Cache results — Store plate reads locally to avoid redundant API calls for the same vehicle.
• Use webhooks for batch — For high-volume processing, use the batch endpoint with webhook callbacks rather than polling.