Machine Vision
Having to change the ROI for different systems
Hi,
Attached is a VI which is generated using Vision Assistant 2009. ( Please dont comment on the messy layout)
( I have also attached the Vision Assistant script and the character set)
The idea is it is surposed to detect the edges of a two line LCD display then place a ROI around the each line and do an OCR and return the character string of each line.
It works fine on two systems but on a third system, it doesn't recognise some characters. The problem lies around the ROI descriptor for the input to IMAQ TransformROI.vi. The same camera system is used on test systems.
I ran the vision assistant script on the problem system and generated a new VI for this system. This give me a set of figures that worked. Unfortunately these new figures didn't work on the other systems.
I need to make this system independent and I dont want to have to supply different data from outside the VI (from Teststand)
Any clues.
Hello
It's difficult to see what the problem is as you didn't post any tes images. Moreover, ther's a vi missing.
Regards
I'll post a couple of images. The missing VI is not important and can be delete. I'll re-post with a front end.
It works fine for me, but difficult to say what goes wrong with only two images.
What I suggest is to use the find straight edges function instead of simple edge detector to find your borders, it will be more robust.
Then check that you pass the good points to your set coordinate system function.
regards
