Machine Vision

cancel
Showing results for 
Search instead for 
Did you mean: 

lighting system

I am working on drying clay, I need to measure area decrease with a camera install on top and height decrease with a camera installed in front, my problem is whats the best lighting system to have good result in image processing? with present lighting from top system i only have good result to measure area but a bad result in height measurement.

Download All
0 Kudos
Message 1 of 8
(3,501 Views)

Hello hsadegh,

 

I have a couple thoughts on this. First of all, could you elaborate on the bad result you are receiving? This might help us better understand what is going wrong. Are you unable to find the block of clay from the camera on the side? Or are you able to correctly locate the clay, but getting variation in the height measurement?

 

The best lighting system would be to have TWO lighting sources, one per camera. Something like this image below:

Side Lighting.png

NI recommends working with partners for lighting - we work with Edmund Optics a lot on issues like this.

https://www.edmundoptics.com/illumination/led-illumination/led-ring-lights/ 

 

If you're able to find your clay block, but getting inaccurate height measurements, this might be a problem with the test setup (not lighting). 

 

Let's take three examples - one where your image is the same as what you sent, one where your part is placed slightly to the left, and one where your part is placed slightly to the right. If your clay part is closer to the camera, it will appear larger (and have a larger height measurement). If your your clay part is further away, it will appear smaller (and have a smaller height measurement). Because the part is always the same distance away from the top-down CAM 1, CAM 1 will be consistent with its area measurements. See below for a demonstration: 

 

Side Lighting.png

In order to correct this, you could consider having some kind of guideline on the surface your clay part is placed on! A "bottom right" corner would help add consistency to your measurements.

 

I hope this helps!

 

Derek B. (esoDerek)
Product Marketing Engineer
National Instruments
0 Kudos
Message 2 of 8
(3,452 Views)

for measurement best lighting system is back light system you can use two vertical and horizontal backlight system and switch off and on any time that you want  get image vertically 
or horizontally  but if you want  measure in same time use two backlight with different wavelength and use optical pass filter for camera to preventing  light problem 
this is best method

0 Kudos
Message 3 of 8
(3,430 Views)

 

Hello Derek B
First of all thanks for your useful suggestions.
you ask elaborate bad results,it means irregular results in height measurement with cam 2,as I said before and I illustrate it,I have just lighting from the top and the block is nearer to cam 1 therefore I have better and consistent results in area(length &width)
when I have lighting from top, it causes a bad effect on the cam 2(right cam.)to measure height, because it does not find edge consistently(high light density from the top), the second problem as you said the block is far from cam 2 and the smaller image cause the worse results.
I need to:
1-move the block nearer to cam 2 and as you said
2-A  lighting system for cam 2 to regulate the light density.
finally I will send results(image & files) showing new changes to my work.
 
0 Kudos
Message 4 of 8
(3,428 Views)

Hi hsadegh,

 

Thank you for the update! Good luck with the changes.

 

One important note though - please do not simply move the block closer to Camera 2. My recommendation was to make sure your block is always the same distance away from camera 2!

 

If your block is 1 inch tall, and is expected to be 10 inches away from the camera, if you move your block CLOSER to the camera, it will appear 1.1 inches tall, because it takes up more of your image. If you move your block FARTHER from the camera, it will appear 0.9 inches tall, because it takes up less of your image. What is important is that you keep all your objects at the exact same distance away from Camera 2. 

 

Finally, I'd suggest angling your camera 2 away from the bright light source if the bright light is affecting your image - or choosing a different region of interest (that excludes the light source). This should help reduce the glare. If you include information about how you're finding your edges, we can better help you diagnose the issue. 

Derek B. (esoDerek)
Product Marketing Engineer
National Instruments
0 Kudos
Message 5 of 8
(3,414 Views)

Hi

I do whatever you said, change the block nearer to front cam, put a lighting system for the front cam,

I don’t have problem for find the upper edge but problem in finding lower edge in LabVIEW, when the block is drying it moves from its position and it makes the measurement harder, saying that there is top cam and lighting working Simultaneous.

Download All
0 Kudos
Message 6 of 8
(3,315 Views)

Hi hsadegh,

 

Just to confirm: you are able to find the top edge in your image now, but not the lower edge? What method are you using to find the edges? Are you performing any image processing before doing so to make the image clearer?

One thing I notice in your images is that the edges are not clear, and it even looks like the image is out of focus quite a bit. Can you try getting a sharper image?

Charlie J.
National Instruments
0 Kudos
Message 7 of 8
(3,290 Views)

your camera lens is not focus on target it is one of your problem also your top edge that you find is not accurate because of shadow of edge com from unfocus position of lens

0 Kudos
Message 8 of 8
(3,244 Views)