How to do OCR?

Ok, sorry for the real simple question, but I’m stuck. Plus, I have very little experience with ECM. (And when you are used to the Epicor resources and documentation and expect the same from DocStar, you become :disappointed_relieved:)

I want to set up a batch import into ECM. I get creating the Content Type, where to drop the files after they are scanned, and that stuff. What I am stuck on is the OCR to name the files. The files will be all the same layout so I can do OCR to get the data I want, just not sure where that is done. I am thinking it is a workflow in ECM? But it could be something on the client that does it? If someone could give me a real high level, simple explanation on the steps I should be good. Any help is appreciated.

Yes you are right it is done in the DocStar Workflows that would be attached/assigned to a specific content-type. The “Tasks” you would use are either “Extract Text” (extracts all text in document) or “OCR” (extracts text from specific portion of document specified by user). I am not sure of the license required to be able to use these in a workflow.

Excellent! Thanks for the push in the correct direction.

How do you trigger the workflow though? And will the workflow always trigger regardless of how the document was added? I’m thinking of a batch import vs. a document added in Epicor.

You need to add workflow to the content type and assign it the workflow name that you would like the content type to be assigned to.

I think that’s how to do it off the top of my head.

1 Like

1 Like

You assign a workflow to the content-type via the “Content Type Builder” and it should run through the workflow anytime that content-type is imported. I’m pretty sure that includes documents imported via Epicor (since you assign a content type to these as well anyway). Then you can OCR the data you want, and use some other “Tasks” to form the document name. Then you can use a “Advanced Format” task to set the title to whatever you need. We’re on an older version of DocStar, but here’s an example of setting a title. One of the “$Field”'s could be something you OCR.


1 Like

Thank you! And you too @utaylor

I am starting to make workflows to read barcodes to file away receipt packing slips. I will do the same for all the rest of the paperwork we generate and file away from our processes. If you have any questions I can try and help. I haven’t used the OCR capabilities much beyond barcode reading.

@jkane John, you didn’t mention it so I’m asking - Do you also have IDC along with ECM?

@utaylor & @Asz0ka are correct and I’d like to add to your comment about “Batch Import”. There are a few ways of getting documents into the workflow besides attaching from inside ERP. The ECM Workstation client running on a server is key to doing batch imports automatically. It can watch File Folders and Email Inboxes for new files and as it creates the entry in ECM it will trigger the workflow to start.

Plus, you need to have at least one Client service running in order to do the OCR in the ECM workflows as the server uses the running Clients as OCR engines (among other things). Better to have more than 1 running just in case.


@MikeGross , no, we do not have IDC