> ## Documentation Index
> Fetch the complete documentation index at: https://docs.tensorlake.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Barcode Detection

> Detecting and decoding barcodes from document pages, returning type, value, and bounding boxes as structured output.

## Overview

Barcode detection identifies and decodes barcodes found in document pages. Each detected barcode is returned as a structured page fragment with its decoded value, barcode type, and bounding box — allowing downstream processing to handle barcodes separately from text, tables, and other content.

Barcode detection is available in the `model03` OCR model and is enabled via a flag in `ParsingOptions`.

## Enabling Barcode Detection

Set `barcode_detection="true"` in your `ParsingOptions` along with `ocr_model="model03"`:

<Note>
  Barcode detection requires Tensorlake Python SDK version `0.2.91` or later. Run `pip install --upgrade tensorlake` before using this feature.
</Note>

```python theme={null}
from tensorlake import DocumentAI, ParsingOptions

doc_ai = DocumentAI(api_key="YOUR_TENSORLAKE_CLOUD_API_KEY")

file_id = doc_ai.upload(path="barcode_file.pdf")

parsing_options = ParsingOptions(
    ocr_model="model03",
    barcode_detection="true",
)

parse_id = doc_ai.read(
    file_id=file_id,
    parsing_options=parsing_options,
)

result = doc_ai.wait_for_completion(parse_id)
```

## How It Works

When barcode detection is enabled, the pipeline:

1. Parses each page into fragments (text, tables, barcodes, etc.)
2. Runs a barcode detector and decoder over the page image
3. Emits `fragment_type: "barcode"` entries alongside other page fragments
4. Includes bounding boxes and page dimensions so barcodes can be positioned or highlighted in a viewer

## Fragment Output

Each detected barcode is returned as a fragment with the following structure:

```json theme={null}
{
    "fragment_type": "barcode",
    "content": {
        "content": "PDF417: 4QGDkVjpF7nuGhQiOgLHwc",
        "html": null
    },
    "reading_order": 9,
    "bbox": {
        "x1": 2,
        "y1": 444,
        "x2": 207,
        "y2": 475
    }
}
```

| Field             | Description                                                                 |
| ----------------- | --------------------------------------------------------------------------- |
| `fragment_type`   | Always `"barcode"` for barcode fragments                                    |
| `content.content` | The barcode type and decoded value, formatted as `"<TYPE>: <value>"`        |
| `reading_order`   | Position of this fragment in reading order relative to other page fragments |
| `bbox`            | Bounding box coordinates `(x1, y1, x2, y2)` in page pixels                  |

## Common Use Cases

Barcodes appear across many operational document types:

* **Shipping labels and packing slips** — tracking numbers and carrier codes
* **Lab reports and sample labels** — specimen and sample IDs
* **Insurance documents** — claim IDs and policy references
* **Utility bills, tickets, and receipts** — account numbers and confirmation codes

With barcodes returned as structured fragments alongside text and tables, you can:

* Match barcode values to internal IDs (shipment, claim, order, patient)
* Validate that a barcode value matches a printed text label on the same page
* Flag documents where an expected barcode is missing or unreadable
* Render barcode overlays in a document viewer using the provided bounding boxes

## Related

* [Parsing Overview](/document-ingestion/parsing/read)
* [Parse Output](/document-ingestion/parsing/parse-output)
* [Sample Notebook: Barcode Detection](https://tlake.link/notebooks/barcode-detection)