LabVIEW

cancel
Showing results for 
Search instead for 
Did you mean: 

[Vision Module] Create a Template from Webcam

Solved!
Go to solution

Hello guys!

I'm working with LabVIEW 2013, Visión Development Module 2013 and IMAQ 4.8

I have a webcam installed and I'm trying to save a template into a folder and leaving a record inside a TXT with the format:

 

templatename [TAB] imagename.png

 

The thing is that when I try to create a template from the image (using a rectangle from ROI tool) it doesn't work... I use the Vision Assistant Express to create a template the image is like 290kbs meanwhile when I save a template from the webcam image is like 3kbs...

 

I'm adding the file that I have now.

 

PS: the picture saving and reading (from the txt file) is perfect... just doesn't recognize the template.

0 Kudos
Message 1 of 9
(3,017 Views)

I was thinking that it was a resolution stuff... but I put the max res (1280x960) in my VI and it takes a 4kb picture just like with 640x480... :S I'm all out of ideas

0 Kudos
Message 2 of 9
(2,991 Views)

I didn't put a loop nor a ROI tool to the Vi that I upload earlier so here it goes... I hope someone could help me 😛

0 Kudos
Message 3 of 9
(2,987 Views)

Well, I'm not at all certain I understand what you are trying to do.  What is a "template"?  Your code certainly can take images (though I don't see any Rectangle Tool in the Image).  When I run it, I do see an Image and a smaller Template, but, again, I don't understand the purpose of it all.

 

I also do not understand what problem you are having.  In my hands, I get errors, but then I don't specify file names, am not sure what you are trying to write, etc.

 

Sometimes it is a good idea to take little pieces of your code and try them out.  Create a single Image (snap a picture) to have some "known data", then exercise the code that writes the Image to disk.  Use Constants for file paths (make everything explicit -- this is a Test, after all, to see if the code works).

 

If you can narrow things down to "what is failing" and post that, we may have a better chance to help you.

 

Bob Schor

0 Kudos
Message 4 of 9
(2,963 Views)

I think I miss an important point of my issue...

The template that I'm saving I'll use it after to recognize it via webcam using Color Pattern Matching with the Vision Assistant Express, here is when the template doesn't work (even if I open the panel of Vision Assistant Express and load the template there it prompt me that is not a valid template format, but when I create a template there and use it on my VI, the recognition works)

foro1.png

When I try to add the green template (that I took from my VI) in the express panel, this is the message:

foro2.png

I don't know where is the difference between the two pictures to have this problem... I'm even using the "NPG with Vision Info" block to save the image...

0 Kudos
Message 5 of 9
(2,959 Views)

So an image of 60 x 56 24-bit pixels is about 1Kb, somewhat smaller than the sizes you mention, but certainly closer to 3Kb than 196Kb.

 

Too bad you are using the "Information-obscuring" Express VIs instead of working directly with IMAQ and IMAQdx, where we could more clearly see what you are doing.

For example, this code writes a 60 x 56 32-bit RGB image (here called Template.png) and, with the default PNG compression, the file is 236 bytes.

Template Code.png

Bob Schor

0 Kudos
Message 6 of 9
(2,952 Views)

One of my if-time-is-right-to-do-list items is to transform the VI Express to a diagram with IMAQ blocks, but for now I'm using this and I have this problem, I don't know if I use this blocks the problem can be solved... but the thing is that right now the templates doesn't work 😞

0 Kudos
Message 7 of 9
(2,949 Views)

I'm uploading a VI better explained where you can see the result on the screen. I hope it works for someone to help me, please 🙂

0 Kudos
Message 8 of 9
(2,941 Views)
Solution
Accepted by lord_nox

Finally I've found a solution to my problem and I'll leave it here in case someone needs the same thing 🙂

 

I've used the example "Color Pattern Matching" and I've modified it so you can receive webcam information instead of a file...

 

The problem was that you have to "prepare" the learning and the way that the recognition will work, that's what it was missing from my VI.

0 Kudos
Message 9 of 9
(2,917 Views)