Our /document/ocr Robot

Recognize text in documents

🤖/document/ocr recognizes text in documents and returns it in a machine-readable format.

With this Robot, you can detect and extract text from PDFs using optical character recognition (OCR).

For example, you can use the results to obtain the content of invoices, legal documents or restaurant menus. You can also pass the text down to other Robots to filter documents that contain (or do not contain) certain phrases.

Warning: Transloadit aims to be deterministic, but this Robot uses third-party AI services. The providers (AWS, GCP) will evolve their models over time, giving different responses for the same input PDFs. Avoid relying on exact responses in your tests and application.

Note: Currently, this Robot only supports character recognition for PDFs. To use this Robot with other document formats, use /document/convert first to convert the document into a PDF.

Usage example

Recognize text in an uploaded document and save it to a JSON file:

{
  "steps": {
    "recognized": {
      "robot": "/document/ocr",
      "use": ":original",
      "provider": "gcp"
    }
  }
}

Parameters

  • use

    String / Array of Strings / Object required

    Specifies which Step(s) to use as input.

    • You can pick any names for Steps except ":original" (reserved for user uploads handled by Transloadit)

    • You can provide several Steps as input with arrays:

      "use": [
        ":original",
        "encoded",
        "resized"
      ]
      

    💡 That’s likely all you need to know about use, but you can view Advanced use cases.

  • provider

    Stringrequired

    Which AI provider to leverage. Valid values are "aws" and "gcp".

    Transloadit outsources this task and abstracts the interface so you can expect the same data structures, but different latencies and information being returned. Different cloud vendors have different areas they shine in, and we recommend to try out and see what yields the best results for your use case.

    AWS supports detection for the following languages: English, Arabic, Russian, German, French, Italian, Portuguese and Spanish. GCP allows for a wider range of languages, with varying levels of support which can be found on the official documentation.

  • granularity

    String ⋅ default: "full"

    Whether to return a full response including coordinates for the text ("full"), or a flat list of the extracted phrases ("list"). This parameter has no effect if the format parameter is set to "text".

  • format

    String ⋅ default: "json"

    In what format to return the extracted text.

    • "json" returns a JSON file.
    • "meta" does not return a file, but stores the data inside Transloadit's file object (under ${file.meta.recognized_text}, which is an array of strings) that's passed around between encoding Steps, so that you can use the values to burn the data into videos, filter on them, etc.
    • "text" returns the recognized text as a plain UTF-8 encoded text file.

Related blog posts