Comparing display using Vision Acquision

[Nghtcrwlr]

Hi all
I am not quite sure whether i can post this here in this forum.Sorry, if not allowed.

I am in the process of evalutaing different solutions for a problem. I have a display that have lights on its both sides. It changes its value ( lets say display ON and display OFF just two states) within 0.5sec. That means ON state  stays for apprx 0.5 sec and OFF state stays for approx.0.5sec. ( Exact  value of time needs to be found out later. Now i am doing only a rough calculation.)  Unless a STOP button is pressed, the display keeps on changing between ON & OFF states.

My goal is to compare these two states with respective template images. i.e I am planning to acquire 2 images ( ON state first and then OFF state) and then compare them with template images. planning to acquire images using a monochrome camera. (grayscale) .If the display is having some problem ( like if lights are ON only on one side ) then the grayscale image will be half dark and half white . and thereby comparing with template image i can say whether the display is good or bad.

Option A: Compare ON state with template and if It is not matching the template, then stop the process. If matching, then acquire OFF image and compare . If both results are OK, thenpress STOP and quit program.

Option B: acquire both images simultaneously one after the other and compare them each with template images.

With both options A& B, I am not quite sure whether VI could compare & process the image in ON state within 0.5 sec and acquire OFF state and compare. Is there any other better option to evaluate the display other than this idea , using Vision Acquisition?
Thanks

[Alexander_Sobolev]

0.5 s is plenty of time to check whether light is on or off.

Though I would not use template images, subtracting images and thresholding difference will be much faster. If you know position of the light, you can select this part of image and to resample it to 10 pixels.

More global: Do you really need a camera? Single photodiode and very slow daq board seems enough... though more expensive than 10$ webcam...

[Nghtcrwlr]

Hi Alexander

Thank you for reply. Subtracting images would be great idea. Another idea that came to my mind with your input is  :

Would it be better if i subtract OFF image and ON image ( i.e take OFF image---> take ON image ---> subtract both images. In standard situation, the result will be a fixed value. If any one side of the light is not working properly, then the subtraction may result in another value. )

I do have other tasks to be visually verified and that is why i thought of using cameras.