Hmm, I didn't think about that. Because now that you mention it, I redid the scenario, but it worked flawlessly. Maybe I picked something that looked white (that was the color I was using). Thanks again for your help. I hope to talk to you again.
did somebody actually tried to implement the scanline algorithm in labview?
iam not that good in labview and couldnt make it work 😕
the floodfill vi from wiebe@carya works for me but it just takes an awful time cause my 2D array is huge.
so therefore i need a much faster algorithm.
I am currently unable to find an implemented scanline algorithm in LabVIEW. However, here is a suggestion to speed up the LabVIEW provided by wiebe@carya.
The link above describes a few ways to increase efficiency, including replacing a while loop with a for loop (if you know the iterations or size of the image).
Firstly thank you for the floodfill contribution by wiebe@CARYA
I required a bit more performance, but only for 8bit images..... and improved the performance a bit.
I then replaced the array manipulation by a fixed array size, and 2 pointers, which improved performance by a factor of 10, but only for large arrays (images from a webcam)
Attached please find both options.
The searth path size is limited to 16*image size..... which occationally fails to fill a complex image.
I was hoping Opencvl would have the floodfill alrorithm included, but not yet....
If someone can beat this performance, please share.
I tried to send the vi in version 9, and then in v13, but the NI web site is saying:
Correct the highlighted errors and try again.
I can only guess that NI do not want to be backward compatible (or alternatively do not recognise their own file format).
Will check the forum to see if there is a work around.