IMAQ fit Ellipse 2

[xgatc]

Hi,

when triing IMAQ Fit Ellipse 2.vi, from LV2010SP,( create a ellipse array then fit it) I found that the Valid Fit flag will alwys==F if the Reject Outliers==T.

also, change Reject Outliers=T or F will casue the output Ellipse to swap its Major Axis Point1 and Point2.

has anyone noticed this ?  is it a bug ?

cheers

xg

[Andrey_Dmitriev]

Hi,

Valid or not - it depends from Minimum Score parameter.

With axis swap - unable to reproduce described behavior (see attachment)

Andrey.

[xgatc]

Hi,

thanks

I could not figure out why you set the minumimscore==0.  if you set it !=0, thing will be different. I added my test code, see the attached.

by changing the Reject Outliers, you will see the behavior I mentioned.

cheers

xg  

[Andrey_Dmitriev]

Yes, you're right, with "ideal ellipse" it looks as described. This might be a bug. I don't see big problems with "swapped" points (it depends how algorithm perform the best fit), but for ideal ellipse we should get valid fit anyway.

[xgatc]

thanks, Andrey

I am interested in how you created the contours. can you eloborate  a bit?

cheers

xg

[Andrey_Dmitriev · Accepted solution]

---

@xgatc wrote:

thanks, Andrey

I am interested in how you created the contours. can you eloborate  a bit?

cheers

xg

---

1. Create IMAQ Image, set size, connect it to Image indicator on the front panel, run VI once.

2. Draw ROI manually on image indicator (Closed freehand tool was used)

3. Use ROI property (read access) for obtain contour (connected to Indicator on the front panel) Run VI again - you will get ROI.

4. Convert indicator to constant, set ROI property to write.

Profit!

[xgatc]

many thanks

[DeNY]

Hi,

could me anyone please describe the algorithm IMAQ fit Ellipse is using?

Thanks for reply in advance.

Dan