Manual Validation for Digitize Documents

The example below explains how to manually extract data from an image and present the output in a separate file. It presents activities such as Digitize Document or Present Validation Station. You can find these activities in the UiPath.IntelligentOCR.Activities package.

Note: This workflow is using an older version of the UiPath.IntelligentOCR.Activities package.

This is how the automation process can be built:

1. Open Studio and create a new Process named by default Main.

   Note: Make sure to add all the needed files (.json files and all the images) inside the project folder.

2. Drag a Sequence container in the Workflow Designer and create the following variables:

   Variable Name	Variable Type	Default Value
   Text	String
   DOM	UiPath.DocumentProcessing.Contracts.Dom.Document
   Data	UiPath.DocumentProcessing.Contracts.Taxonomy.DocumentTaxonomy
   DocumentTaxonomy	UiPath.DocumentProcessing.Contracts.Taxonomy.DocumentTaxonomy
   TaxonomyJSON	String
   HumanValidated	UiPath.DocumentProcessing.Contracts.Results.ExtractionResult

3. Drag a Read Text File activity inside the sequence.

   • In the Properties panel, add the name of the file, in this case "taxonomy.json", in the FileName field.
   • Add the variable TaxonomyJSON in the Content field.

4. Add an Assign activity below the Read Text File activity.

   • Add the variable Data in the To field and the expression DocumentTaxonomy.Deserialize(TaxonomyJSON) in the Value field. This activity builds the taxonomy for extraction.

5. Drag a Digitize Document activity below the Assign activity.

   • In the Properties panel, add the value 1 in the DegreeOfParallelism field.
   • Add the expression "Input\Invoice01.tif" in the DocumentPath field.
   • Add the variable DOM in the DocumentObjectModel field.
   • Add the variable Text in the DocumentText field.

6. Drop a Google OCR engine inside the Digitize Document activity.

   • In the Properties panel, add the variable Image in the Image field.
   • Select the check box for the ExtractWords option. This option extracts the on-screen position of all detected words.
   • Add the expression "eng" in the Language field.
   • Select the option Legacy from the Profile drop-down list.
   • Add the value 2 in the Scale field.

7. Place a Present Validation Station activity below the Digitize Document activity.

   • In the Properties panel, add the variable DOM in the DocumentObjectModel field.
   • Add the expression "Input\Invoice01.tif" in the DocumentPath field.
   • Add the variable Text in the DocumentText field.
   • Add the variable Data in the Taxonomy field.
   • Add the variable HumanValidated in the ValidatedExtractionResults field.

8. Drag a For Each activity under the Present Validation Station activity.

   • In the Properties panel, select the option UiPath.DocumentProcessing.Contracts.Results.ResultsDataPoint from the TypeArgument drop-down list.
   • Add the expression HumanValidated.ResultsDocument.Fields in the Values field.

9. Drag a Log Message activity inside the Body of the For Each activity.

   • Select the option Info from the Level drop-down list.
   • Add the expression item.FieldName in the Message field.

10. Drag a Log Message activity below the first Log Message activity.

    • Select the option Info from the Level drop-down list.
    • Add the expression item.Values(0).Value.ToString in the Message field.

11. Drag a Write Line activity under the Log Message activities.

    • Add the value "" in the Text field.
12. Run the process. The robot uses the IntelligentOCR activities to manually process the data and to present the results.

Download example from here.