Uncertainties of edge detection

[n 516]

I'm using IMAQ PCI-1408 acquisition board to acquire a picture and run a .vi to detect a edge in the picture by using a few IMAQ functions in LabView. I'm wondering how to decide the uncertainties (including the precision and the bias) of this measurement.

[cld]

Currently,the edge detection function has a 1/4 pixel subpixel accuracy.

[n 516]

In my program, I used some IMAQ functions such as IMAQ creat, IMAQ Snap.vi, IVB Mask from ROI.vi, IMAQ Rake, IMAQ convert Pixel to Real World, IMAQ Set Simple Calibration.... Do I need find out the accuracy for each function? For the total accuracy, do I need just add them together?

[cld]

The only functions here that will be effected is the IMAQ Rake, which will be doing the processing. The convert Pixel to real world will just be a mathmatical function based on what you set as the actual size of a pixel. So your result wil be accurate to 1/4 of the value that you set as the pixel dimension. You may have some other factor provided by the camera vendor on how consistent the pixel width on the ccd is and you would have to factor this in also.
Best Regards,
Chris D