Extract content from forms (Rinkt)

Extracts key value pairs or tables from documents(Pdf, jpeg, tiff) into structured data. To be able to extract key values, a project must be created and trained with at least 5 distinct documents using https://document.rinkt.com using the same credentials used for connecting to Studio or Portal To be able to recognize multiple type of documents, a master project can be created that it is composed of other projects representing the models(documents) that we want to recognize.

Form Recognizer resource keys

Designer Properties#

  1. File Path The pdf or image file to be analyzed.
  2. Project The project to be used for extraction. It can be a simple project, a composed project, or a shared project. It may take few seconds, depending of network band with to load all projects for your account.
  3. Result The prediction result.

Properties#

Table words extraction properties

Workflow Properties#

  1. Project The project to be used for extraction. It can be a simple project, a composed project, or a shared project. It may take few seconds, depending of network band with to load all projects for your account.

General Properties#

See General Properties.

Input#

  1. File Path The pdf or image file to be analyzed.
  2. Model id The id of trained model to be used for extraction. It can be used we want to use a specific model instead of the latest(when using the project). It overwrites the Project property.
  3. Process each page Process each page individually and then merge the results.

Misc#

See Misc.

Out Error#

See Out Error.

Output#

  1. Prediction Result The prediction result. //TODO ...