Hi,
How does your inspection work? Do you drive the dashboard to display some specific text, then perform the visual inspection to verify that the text is correct? Basically is it an OCR (Optical Character Recognition) inspection where you don't know a priori what's going to be displayed or an OCV (Optical Character Verification) where you want to verify that the text displayed is correct?
You can tell the engine what type of characters each letter should be. For example, you can specify that the first character should be an uppercase letter, the second a lower case letter etc. To do that, go to the Options tab, check Use Text Pattern, and click on Specify Text Pattern. You can add a specific type for each character. This works well in an OCV type application, where you have some knowledge about the text you expect to read.
Also, in general, make sure to use the same threshold and size options that were used to train the character set file.
You might also want to retrain the characters that are not recognized correctly with the training interface, untill you get perfect recognition.
Christophe
