8 replies to this topic

[hookes2]

Hello everyone,
I want to acquire images of License Plate Numbers, extract the numbers and log them into a file either as a text file or an image file. I have no prior knowledge of image processing and my LabVIEW skills are still poor. Can anyone please help me out...maybe with the VIs and some materials. Thank you.

[jdunham]

Hello everyone,
I want to acquire images of License Plate Numbers, extract the numbers and log them into a file either as a text file or an image file. I have no prior knowledge of image processing and my LabVIEW skills are still poor. Can anyone please help me out...maybe with the VIs and some materials. Thank you.

In your position, I think you should buy an off-the-shelf solution. Why do you expect us to do this work for you when you can just purchase such a system?
http://www.eaglevisi...recognition.htm

Of course if this is educational, you're going to have to do the majority of the work yourself and show the work you've done so far before asking specific questions about how to reach the next step.

Good luck!

[DaveC]

This sounds like a good chance to improve your LabView skills ...

To get started do a search for image acquisition examples, you need to get images in reliably before you can analyse them. Then you'll need to start working with the Vision and Motion Toolkits, specifically the Machine Vision --> OCR (Optical character recognition) features. There are examples on their use, so use them to help you learn. Once you have something roughly working (or at least planned and partially implemented) then members here will be more than happy to help.

[hookes2]

Hello Everyone,
I am working on a project which will allow me to capture images of license plate numbers of vehicles and log them. Right now I can do OCR and ROI on a sample image which I captured and saved on my computer using the OCR example express VI in LabVIEW. But that is not enough, I want to be able to acquire images automatically using my camera and the ROI selected automatically. I just ordered the following hardware (though I am yet to receive them, and they cost me a fortune) and I wondering what I will do with them when they finally arrive. Please someone come to my rescue!

1. NI CVS-1456 Compact Vision System
2. NI Vision, I/O Terminal Block and Prototyping
3. Camera, Basler SCA1600-14FM, IEEE-1394B, 1628X1236, 14 FPS, Monochrome
4. Cable, CVS to Basler A601 trigger cable
5. Lens,35MM,Compact Fixed Focal Length, Edmund Optics

[jdunham]

Hello Everyone,
I am working on a project which will allow me to capture images of license plate numbers of vehicles and log them. Right now I can do OCR and ROI on a sample image which I captured and saved on my computer using the OCR example express VI in LabVIEW. But that is not enough, I want to be able to acquire images automatically using my camera and the ROI selected automatically. I just ordered the following hardware (though I am yet to receive them, and they cost me a fortune) and I wondering what I will do with them when they finally arrive. Please someone come to my rescue!

1. NI CVS-1456 Compact Vision System
2. NI Vision, I/O Terminal Block and Prototyping
3. Camera, Basler SCA1600-14FM, IEEE-1394B, 1628X1236, 14 FPS, Monochrome
4. Cable, CVS to Basler A601 trigger cable
5. Lens,35MM,Compact Fixed Focal Length, Edmund Optics

Sounds like a good setup. First you will want to hook up the camera and acquire the images automatically using the NI Vision Examples. If you are having trouble, NI tech support should be able to get you to that point. How do you plan to select the ROI automatically. That seems like the difficult part, but I'm sure it's possible.

[Hera_Walhalla]

Hello! Currently I am working on the same project (License Plate Reading) and I need some guidance regarding the best way for the plate localization. I read that this could be done with connected component analysis, edge detection or some algorithm to determine the plate location by finding the color space of the plate. I can't find enough resources about connected component analysis in LV, but there are some examples for edge detection... I really don't know which is the best method and the simplest to implement.

Any ideas will be appreciated.

Edited by Hera_Walhalla, 24 August 2011 - 11:06 AM.

[Phillip Brooks]

Since you are waiting for your hardware, this would be a good time to think about your environment and how to trigger.

Your environment sounds different than an automated assembly line with controlled lighting and a constant rate. The environment will define how you place your camera, your trigger(s) and possibly require lighting control (additional lighting or screening to prevent glare).

[hookes2]

Thanks Everyone...
I finally got an algorithm written using vision assistant and I automatically generated the LabVIEW code. I still have issues with auto-generating the ROI. Right now, I can only select it manually.
Thanks once again.

[Hera_Walhalla]

I got a solution for that(maybe it fits for you too): If you want to locate the plate, try geometric matching for the plate, make a template for the plate ignoring the content (chars and digits sequence) and keep only the borders. If you play a bit with the threshold and geometric matching, the program you build should be able to locate the plate with a high accuracy. For this you should make different templates and see which is the best(keep only borders, keep only the corners for the borders of the plate). Then you should be able to extract the matching and then do the OCR on it.

Good luck!