Image groups

An image group refers to a collection of one or more images. The Ocrolus API allows users to upload images of any supported file type. Ocrolus will then classify the images for document processing. The form types supported for image classification, such as tax forms, ID proofs, bank statements, and pay stubs, can be found here.

Upload images to Ocrolus

To upload images using Ocrolus API, perform the following steps:

  1. Generate your API credentials as described in the Authentication section.
  2. Create a new Book to upload all the images. Save the book_uuid or Book pk for later use.
  3. Upload the input file to the Book using the Upload the image files. Use the book_uuid or Book pk from the response of create Book API while invoking the upload call to ensure the files are uploaded in the intended Book.
  4. Retrieve the document's classification results using the Book classification summary or mixed-document classification summary endpoint.
  5. To know the uploaded mixed document's status, call the Retrieve mixed document status endpoint.
  6. Once uploaded, use Mark image group as complete API to finalize the image group.

    📘

    Image formats

    We currently support a wide range of image formats including JPEG, PNG, BMP, and TIFF (only single-image TIFF file).

Setting the form_type to BANK_ACCOUNT functions the same as uploading PDF bank statements. For any other form type, it is equivalent to uploading PDF forms. For a complete list of supported document types, see all supported documents.

Additional form types can be configured for your organization upon request. Your Account Manager can provide details on the form types available for your organization. For any inquiries regarding form types, feel free to email us at [email protected].

Check for completion

To receive the complete results for a case, poll for completion for the Book. Alternatively, we recommend registering for webhook notifications, which lets you know exactly when a document has completed processing.

📘

Note

While you can request results at any point, you will only receive data from documents with VERIFICATION_COMPLETE status.

Extract results from Book

To extract the results from a Book: