How do I set up OCR?

Setting up OCR is done by going through steps (see below), but the steps can be navigated in any order. Note that OCR only works in single trigger mode. Navigate by using the Next and Back buttons below, or by clicking on the panels at the top of the OCR pane to go directly to a panel. The Summary panel lists all the settings, and red indicates an error. You can use the repair icon to navigate straight to the pane related to the error.

Fonts: You can upload fonts that will be used for OCR reads. The already used fonts are shown in the dialog box. Clicking on any of these allows for additional actions below the dialog box. Note that Edit Font works similar to Upload Font, but you can select characters that will be included in the upload. This allows for reducing file sizes, which will result in faster load times of the device. Use Ctrl+click to add characters one by one in the pop-up window. You can use the arrow next to the Font pop-up window’s OK button to reveal upload as, with which you can change the file name of the font file.

Fixturing: You can set up fixturing on symbologies or patterns using ROI (region of interest) fields on the read image in the image box so that OCR reads can be positioned in relation to the fixtured item. When fixturing on pattern, the yellow region is used to select the pattern for training, and the orange region is used to provide the frame where the pattern is searched for during operation. When fixturing on symbology, you have to set the appropriate code the device looks for. In both cases, you must use Train Fixture to finalize fixturing.

OCR Strings: To finalize the setup of OCR, navigate to OCS String 1. A green “Trained” or a red “Untrained” indicates the status of the OCR string (this color code applies to the tab name as well). You can select the Font for the string under Selected Font from the already uploaded font types.

Fielding can be ticked to set which characters can be in certain positions during an OCR read. The tooltip (navigate the mouse pointer over the text field to the right of Fielding) indicates what characters signify in the fielding string.

Next, navigate the OCR ROI field into position over the text in the image, and enter the training text.

Accept Threshold is a character score that lets you manipulate the threshold at which characters are accepted in an OCR read. Reads are color coded on the image.

Confidence Threshold is set on a minimum character score difference between the first and second matches in OCR reads that is still accepted.

After everything is set, click on Train OCR to finalize training the OCR string. Note that if you are using symbology fixturing, or want to read codes other than OCR, then keep Allow partial results checked!

Additional OCR strings can also be set up under OCR String 2-4, each with its own visual cue (separately colored ROI frames). Note that the more OCR Strings are trained, the slower the device becomes.

NOTE: Training with Multicode is not supported.