OCR in video

[ingnoura]

Hello every one,

Please can any one help me in programming my application. I'm trying to make number recognition from digital displayer. i found tutorials about OCR for still images. but my problem is that I have a 20 mn video as duration and i want to take the values from the displayer every 1s and plot them. But I couldn't do that because I know such task for just one frame. my webcom has a rate of 30 f/s.

i already setup labVIEW8.6 and Vision toolkit. but not IMAQ Vision OCR.

To metion that i'm just beginning to learn LabVIEW.

Thank you very much in advance for your help.

[crelf]

...i want to take the values from the displayer every 1s and plot them. But I couldn't do that because I know such task for just one frame. my webcom has a rate of 30 f/s.

You'll need to install the OCR toolkit component if you want to use OCR (last I checked it was included in IMAQ Vision?), then you'll need to extract frames (images) that you want (every 1s) out of the video and do the OCR on them individually.

[ingnoura]

You'll need to install the OCR toolkit component if you want to use OCR (last I checked it was included in IMAQ Vision?), then you'll need to extract frames (images) that you want (every 1s) out of the video and do the OCR on them individually.

Thank you very much crelf. But my problem is how to extract every 1s a frame from the video? Is their a function that let me doing that? To mention, I think have already setup IMAQ Vision, isn't it with LabVIEW toolkit?

[crelf]

Thank you very much crelf. But my problem is how to extract every 1s a frame from the video? Is their a function that let me doing that? To mention, I think have already setup IMAQ Vision, isn't it with LabVIEW toolkit?

No, you get a handful of functions with the normal LabVIEW install, but the video and OCR functions are in a toolkit called the NI Vision Development Module. Come to think of it, maybe you don't need LabVIEW at all - have you considered NI's Vision Builder AI? It's a more interactive environment that doesn't require LabVIEW programming...

[ingnoura]

No, you get a handful of functions with the normal LabVIEW install, but the video and OCR functions are in a toolkit called the NI Vision Development Module. Come to think of it, maybe you don't need LabVIEW at all - have you considered NI's Vision Builder AI? It's a more interactive environment that doesn't require LabVIEW programming...

I've installed NI Vision Developpement Module and I'm working on it. Thank you very much

[Minh Pham]

I've installed NI Vision Developpement Module and I'm working on it. Thank you very much

sequence to try:

1. load the video then play it.

2. capture a frame every 1s to use as an image input for OCR purpose.

3. OCR the captured image

4. Get the result from OCR and do calculation/analysis/display as required.

5. Loop back

[ingnoura]

sequence to try:

1. load the video then play it.

2. capture a frame every 1s to use as an image input for OCR purpose.

3. OCR the captured image

4. Get the result from OCR and do calculation/analysis/display as required.

5. Loop back

Thank you very much Minh, that's exactly what I did.