LabVIEW

cancel
Showing results for 
Search instead for 
Did you mean: 

Image noise

Hi,

I have some issues on image processing without any use of Vision series, just purely Labview Professional Development System.

 

First Problem:

I am able to convert an image from RGB format to a Greyscale but somehow the image is noisy. I meant noisy is that when there is a line in the image, it does not has a clean line. It has some other noise pixel surrounding the line.

So I would like to clean off the noise pixel surrounding the line.

 

Second Problem:

If I implement a hard threshold to clean the images, it will clean off the some of the slightly dark pixel as compared to the white pixel which is approximately in greyscale's 255 which may be representing the yellow colored pixel.

 

How can I not cleaning off the yellow colored pixel?

So I hope someone can direct me to some other Community Discussion which has such example which can clear my doubts, or someone can enlighten me for such cases.

Here is the .jpg file which I would like to have a clean background which has a 255 greyscale background.

Image.jpg

 

 

 

0 Kudos
Message 1 of 5
(526 Views)

Thank you for attaching your code (meaning the LabVIEW code that "doesn't work" and your example image.  We can clearly see the problem ... oops, no, we have no idea what you did because we cannot read minds.

 

Please help us to help you.  Attach all of the LabVIEW code you are using, and the example Image.

 

Bob Schor

0 Kudos
Message 2 of 5
(492 Views)

Your image is RGB, i.e. 8 bits for each of the three colors. You can filter any way you want, for example you can eliminate all pixels where the BLUE component is above a certain adjustable threshold. This will eliminate all lighter grey pixels except for the yellow. (You can make it even more selective with a few more rules, so see how far you get and modify as needed.)


LabVIEW Champion. It all comes together in GCentral GCentral
What does "Engineering Redefined" mean??
Message 3 of 5
(474 Views)

Your question is how to clean up the output from a vi that is meant to turn an image to greyscale, but it seems that maybe you need to revisit the vi you're using that is creating this noisy greyscale to begin with. It sounds as though it is not properly filtering out color as it should.

 

Presumably, you're splitting your color values of your image into RGB and then adjusting them somehow based on how the eye perceives intensity, correct? It would help to see your original code. 

However, this thread may provide the answers you seek:

https://forums.ni.com/t5/LabVIEW/converting-RGB-to-grayscale/td-p/2050804

Using Labview 2016/2019, Windows 10
0 Kudos
Message 4 of 5
(371 Views)

5 years ago, someone asked almost the exact same question, with the same input picture

in Remove Image Noise - ( ‎2015-03-21 )

 

must be some kind of assignment

Message 5 of 5
(346 Views)