top of page

CDI Blog

Using Handwriting Recognition in Laserfiche Cloud

Updated: Jan 26

By Kyle Knebel

January 2023 Consultants Corner

As a Laserfiche Cloud Administrator or Process Designer, you have a new tool in your arsenal. Handwriting recognition is now built into the Capture Profile suite of Process Automation tools. Capture profiles and extract data from documents to be used in a business process. These profiles can extract both typed and handwritten data. This month’s article will focus on building and using a capture profile to capture data from a U.S. Federal W-4 form.

Below is an example of a document with the template fields filled with the handwritten information from the scanned W-4.

So how do we build this so that a form, such as this W-4 Federal Employee Tax Withholding Certificate, can be scanned to the repository and the information captured from it?

We’ll need to do two things, build a Capture Profile and then use it in a Workflow. Both of these steps are done in the Process Automation area of Laserfiche Cloud.

Step 1: Build the Capture Profile

Navigate to “Process Automation” by selecting it from the Application Selector at the top-right of the Cloud interface.

Select the “Capture Profiles” tool on the Process Automation screen to build a capture profile.

Next, select New > Capture Profile and name it.

You’ll then want to upload a sample document. Click add a sample document to add one or more images (.png, .jpg, .gif, .bmp, and .tiff) or .pdf files that are samples of the pages you want to capture data from.

Having selected your sample document, preview your sample document, then press Save.

The sample file will upload, and then you can add Zones around the areas where the text will be captured.

In the Profile Editor, you will move your cursor to the top-left corner of the area you want to “box” and then drag while holding down on the mouse button. The cursor will look like a “cross-hairs” icon, as highlighted below.

Drag down to the bottom-right corner of the area to be “boxed” and release the mouse button.

Best Practice: The green zone box will grab nearby text, so it’s better to make the box a bit smaller and tighter to the area in which text will typically be found.

It will then pop up with a Zone dialog box. Choose “Text” unless you are reading a Barcode. Give the Zone a name, like “First Name” or “SSN,” etc.

Best Practice: Use the “Relative” Zone position to improve the accuracy of captured data by automatically adjusting the zone position to account for the shifting of the page's text.