Manual Thresholding and Saving 8-bit binary

[VeMo]

Hello,

I am trying to understand some programming and thresholding. I have an image (attached) that I want to convert to 8 bit binary (0-black and 255-white pixel intensities only) but I want to use manual loops and not the available vi's so I can understand the process and manipulate it in my way. I tried doing somethings but seems like it does not work properly. Can somebody please help and let me know how to perform thresholding using manual loops in LabVIEW?

The other reason to do this is I tried using thresholding vi and converted an image to binary (0 and 255) but after I saved the file, it was not binary with only blacks and whites anymore. It has some grey pixels too so the system did not save the binary image as binary. I would also like to learn how to force the system to save the pixels as only 0 and 255 even if they are not?

 

Thanks

[altenbach]

Your output image will turn black for all pixels that are >100 and keep their values otherwise. Maybe you want to swap your cases and either output a zero or a 255, depending on the comparison.

Here's what I would do. Arguably quite a bit simpler 😄 

[VeMo]

I was able to make my program work with minor corrections/edits. 

I tried your suggestion too but I wasn't able to make it work. I am still new so I may have made some mistakes in trying your suggested idea. Thank you for the suggestion. 

[altenbach]

Both should work exactly the same, but with simpler code there are many fewer places for bugs to hide. Can you show us what you did?

[VeMo]

Yes sure I can share. Below program is doing what I was trying to achieve and understand. I do have one more question that I am going to post separately in a new thread. 

[altenbach]

Here's a quick demo:

[altenbach · Accepted solution]

---

@VeMo wrote:

Yes sure I can share. Below program is doing what I was trying to achieve and understand. I do have one more question that I am going to post separately in a new thread. 

---

There are some glaring overcomplications with your code.

• The "value" terminal belongs before the loop so it can be constant folded for the duration of the loop stack. If it is on top of the loop stack, the compiler must assume that it can change at any time, requiring polling at each iteration of the inner loop. Very inefficient.
• Why all the math in the case structure if you could just use two diagram constants (0, 255)? For example this would be the same (make sure all diagram constant have the correct representation, U8):
•  
• Simplify!!!!

[VeMo]

Thank you for your suggestions. I learned now two new ways to approach the same thing and use the best possible method. 

I am going to post another thread now which is what I am after in manipulating a simple image of a circle but with some noise around it.