LabVIEW

cancel
Showing results for 
Search instead for 
Did you mean: 

Image type accuracy/ pattern matching

I am currently in the beginning process of writing a code for pattern matching. In that, I started to create a code to convert an image to an array. However, the pixels are slightly off. For instance, I might check the pixel at the 400 400 index and the RBG values are slightly off. 

 

In the code I converted the jpeg image to a U8. Is this where my error is coming from? If so, which image should I convert it to so I can maximize accuracy? If not, what are potential sources of the error? And because it is only slightly off, will this cause a huge issue while pattern matching? I am assuming so.

 

Finally, my current code is only outputting one pixel when I give it a certain index. How can I make it a whole array to output all the pixels?

 

Thanks.

0 Kudos
Message 1 of 2
(288 Views)

We can't really help unless you show us your code (Attach your VI and some typical data).

 

  • How much is "slightly off"? (remember that the first element is at index=0, so if you call that "1", you'll be off by one).
  • The target conversion depends on the image type. If it is a greyscale image, U8 is probably sufficient.
  • What kind of "patterns" are you trying to match?
  • If you give it an index, it will give you that element. If you want the entire array, just don't use index array, right? If you want a smaller subset, use "array subset".
0 Kudos
Message 2 of 2
(274 Views)