Machine Vision

cancel
Showing results for 
Search instead for 
Did you mean: 

Matching analysis in case of LCD

Hi anusree,

Some sugessions,

  1. Do not use jpeg for image processing. Save as bmp and do the processing.
  2. The error you have mentioned will come - when you keep the two images for the XOR input differently, either by type (8 bit, 16 bit or RGB) or the size itself.

Make your postings clear, so that people can help you easily.

thanks,

logic

 

 

0 Kudos
Message 11 of 28
(2,480 Views)

Hi logic,

The same error pops up when I use file as *.bmp, both  images used are similar type. How can we make the image snapped from camera is same type and same size as template to compare both.

The comparision is between the snapped image of the device from the camera with the template.

with regards,

anusree

 

 

Message Edited by anusree on 08-02-2005 04:53 AM

0 Kudos
Message 12 of 28
(2,480 Views)

Hi anusree,

I am not sure where is your source of problem is? i have used the image posted by you and tried the XOR operation. i guess the image you saved is in RGB format, but the camera looks like monochrome camera. Anyway see the attached vi & script, see whether it helps you. Read the XOR vi supporting image types in help, make sure that you have the same in code. See what is the image type you have used for image creation.

Still if you are not successful post your code.

Thanks,

logic

0 Kudos
Message 13 of 28
(2,472 Views)

Hello logic,

The two images which are used to Xor is HL_V.jpg  and untitled1.jpg  which are posted earlier, The image captured from the camera of a testing device is untitled1.jpg and the template created is HL_V.jpg  the region of analysis. The  image captured from camera may vary slightly depending upon placing the test device on the fixture.

I hope you get my problem.

 

With regards

anusree

0 Kudos
Message 14 of 28
(2,469 Views)

Anusree,

I have attached a vi that do the job :

1/ The screen is located and the corresponding part of the image is cut from the original. This is done on both the reference image and the test image. The reference width and height are used to cut the test screen image.

2/ The reference image is subtracted from the test image.

Enjoy.

Message Edited by chilly charly on 08-02-2005 05:34 PM

Chilly Charly    (aka CC)
Message 15 of 28
(2,446 Views)

Thanks CC,

I will check with this Vi

With regards,

anusree

0 Kudos
Message 16 of 28
(2,433 Views)
Hello CC ,
 
The example vi is tested using subtraction in my application. It detects the default LCD and result is as required. The problem is when the device under test is placed on the fixture, The placement of device on fixture sometimes moves device slightly (in mm , maximum 5mm )
up or down or left or right , Since the position of camera is fixed., postion of LCD  in the captured image changes and as well the range of operation changes which leads to fail result. Please help me in fixing this.
 
With regards,
 
Anusree
 
0 Kudos
Message 17 of 28
(2,431 Views)
You will then need to establish a coordinate system based off of a feature of the part (PCS). Offset all finders by the actual part position.
~~~~~~~~~~~~~~~~~~~~~~~~~~
"It’s the questions that drive us.”
~~~~~~~~~~~~~~~~~~~~~~~~~~
0 Kudos
Message 18 of 28
(2,420 Views)

Hello,

As I am very new in using Vision  software, It is helpfull if you send me an example to  use the  coordinate to extract the position of LCD in the image .

 

With regards,

anusree

0 Kudos
Message 19 of 28
(2,417 Views)

Anusree:

 

I do not have time to send an actual example, best I can offer is to search IMAQ Vi Reference for coordinate under LabView Help.

The three vi's available in LV IMAQ to set up coordinate systems are under NI Measurements/Vision/Machine Vision/Coordinate Systems palettes.

I have never set up coordinates systems in IMAQ, but I do use them regulary on programs I write for a View Engineering vision system (non LV).

 

Good Luck!

~~~~~~~~~~~~~~~~~~~~~~~~~~
"It’s the questions that drive us.”
~~~~~~~~~~~~~~~~~~~~~~~~~~
0 Kudos
Message 20 of 28
(2,451 Views)