Driver License OCR vs PDF417 API

What the Failure Means

Front OCR and back-barcode parsing solve different problems. Treating them as interchangeable can hide capture errors or omit fields that exist only on one side of the credential.

A Reliable Recovery Path

Use /v1/driver-license for visible front-side fields and international licenses. Use /v1/driver-license-pdf417 for deterministic AAMVA data from US license backs; compare the two only as an application-level review signal. Consult the PDF417 error reference for status codes and use the scanner with fictional or authorized samples while tuning capture guidance.

Where This Workflow Fits

• Front image uses OCR endpoint: Validate this condition before the image leaves the capture screen.
• US back uses PDF417 endpoint: Include this check in support guidance and quality monitoring.
• No AI guessing for barcode fields: Include this check in support guidance and quality monitoring.

Handle the Result Explicitly

Treat capture errors as application states. Never substitute guessed identity values when the barcode cannot be validated.

Prerequisite: A server-side API integration and a capture UI that can request a new license-back image

# Front-side or international license
front = requests.post("https://api.structocr.com/v1/driver-license", headers=headers, json={"img": front_b64})

# Back-side US AAMVA barcode
back = requests.post("https://api.structocr.com/v1/driver-license-pdf417", headers=headers, json={"img": back_b64})

# Your application decides how to review differences.
print(front.json().get("data"), back.json().get("data"))

Example Recoverable Error

This illustrative error contains no personal data. Use the error code, rather than free-form text, to select the next application action.

{
  "success": false,
  "error": "PDF417_NOT_DECODED",
  "message": "Upload a clear image of the complete back of the US driver license or State ID."
}