​
Overview

Cardinal automatically detects and extracts barcodes during processing. For each detected barcode, you’ll receive:
  • image_url – a cropped image of the detected barcode.
  • bounding_box – exact coordinates (x_min, y_min, x_max, y_max, plus width and height).
  • original_polygon – polygon coordinates for the barcode region.
  • barcode_type – e.g., Code128, QR_CODE, EAN_13.
  • barcode_value – decoded string from the barcode.
  • page_number – the page where the barcode was found.
  • confidence – detection confidence score.
  • barcode_index – index of the barcode on the page.
Example of barcodes returned by the API Example: Detected barcode and accompanying metadata

​
How to Enable

To enable barcode extraction, set barcodeDetect: true in your API request.
  • Default: barcodeDetect = false
  • To enable barcode extraction, set barcodeDetect = true.
  • ⚠️ Enabling barcode extraction will add latency to your requests, since additional detection passes are run.

​
Example Response

{
  "barcodes": [
    {
      "image_url": "",
      "bounding_box": {
        "x_min": 1506.6869,
        "y_min": 107.3429,
        "x_max": 1835.3657,
        "y_max": 154.0509,
        "width": 328.6787,
        "height": 46.7080
      },
      "original_polygon": [
        7.5815,
        0.5861,
        9.1245,
        0.5861,
        9.1245,
        0.7195,
        7.5815,
        0.7195
      ],
      "barcode_type": "Code128",
      "barcode_value": "CFP2023112000005",
      "page_number": 1,
      "confidence": 1.0,
      "barcode_index": 1
    }
  ]
}
Bounding box coordinates are provided in the same unit system as other page geometry. Use page.width and page.height from the response to normalize values if you need to scale to your viewer.