10-21-2013 06:12 AM
Hi guys,
I have managed to train single line of text in OCR (NI Assistant)..as attached..however I dont know how to train two (or more) line of texts..for example the car's plate number below:
WDP
3194 - as attached
I want my script.scr file to be abled to read any type of line/string..whether it's single, double, triple etc..
Can anyone show me the way how to do it.
Thank you in advance 🙂
Suliadi
10-25-2013 11:57 AM
Suliadi,
How were you able to train the NI Vision Assistant to recognize a single line of text in OCR? Could you attach your VI or show a block diagram? Also, what version of LabVIEW and Vision are you using?
I don't know how to do this myself but I'm hoping that my response will bump up your question so that others in the community will see it.
10-25-2013 12:47 PM - edited 10-25-2013 12:49 PM
Starting with Vision Development Module 2013, OCR supports multi-line reading. You can create a single region of interest that emcompasses all of the lines that you need to read along with the expected number of lines and it will output the text for the best lines. If you are using an earlier version, you will need to have a separate region of interest for each line of text. If you don't know how many lines you will have, one method is to do an edge detection to determine if text is present.
You can learn more about the multi-line support in this white paper and even more information in the Vision Concepts Manual.
07-21-2015 05:27 AM
Dear Carlton
I currently use a IMAQ OCR Read Text 4 VI in my OCR code. However, I have determined the number of lines 2 but IMAQ OCR Read Text 4 VI determine 1 line!!!!
please help me!