Can one get the engineering value out of a calibrated image as one scans down the image and searches for the part edge?

[wdm]

I grab an image containing a laser line that follows a curve. I then capture another image after I turn off the laser, subtract the 2 images displaying only the line, make it a binary file, and then search row by row for the ones and record the x/y coordinates for each. I then want to plot the x/y coordinates in engineering units. I currently do a rough calibration of the image and use equations developed from the calibration to convert the x/y to engineering units. I would like to use the calibration function in IMAQ 6.0 to do the calibration and then extract the engineering units right from the binary image with out having to convert th
e pixels.

[oliver_sidla]

Hi,

the way you describe your procedure seems basically ok.
What we usualy do to calibrate is the folllowing procedure:

* Ceate a calibration target with dot patterns,
hold it as precisely as possible into the plane which is created by the laser. Take care that the target covers the field of view of the camera.
* Take an image
* Having the image of the blobs you can use it to calibrate using the IMAQ functions - in this case calibrate for perspective distortion and possibly lens distortion.
The IMAQ function accepts either an image and tries to locate the calibration points automatically (which will probably not work in your case) or it accepts a list of point coordinates which you have to create on your own (we use a custom tool for that). The struc
ture of the list is described in the IMAQ manuals respective VI documentation. It contains basically the image coordinates of blobs (measured from the calibration image) and their associated world (engineering) coordinates (known in advance).

* Once the calibration is finished you can use it to convert from image to world (engineering) coordinates.

Its still a little bit complicated but I hope this helps.

Regards Oliver