cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
Showing results for 
Search instead for 
Did you mean: 

Community Tip - Want the oppurtunity to discuss enhancements to PTC products? Join a working group! X

Extension of previous 'convolution' Topic

Jbryant61
4-Participant

Extension of previous 'convolution' Topic

Sorry for starting a new thread but i wasn't sure if the experts were still watching an old thread.

Eden, Richard & Jean kindly helped take two functions (psf & pixel response) and convolved them (by oversampling), create beautiful graphics and downsample to give what a CCD would detect. Eden also gave a very nice integration function that allowed the half total integral distance to be determined.


I have taken this further and now allowed for non-point objects - modelled real objects as gaussian for now. Eden commented on the need for recentring when performing convolution, but it appears if the convolution involves 3 functions then this is not needed. Is this correct or is my convolution wrong.

Also, the plot at the end for fractional volume radius displays 2 plots, one is the raw (unconvolved) psf, and the other is the total convolution. For some reason the raw psf (red curve) changes when the pixel value changes which it shouldn't do. I have spent a week trying to see why with no success.

Jason
2 REPLIES 2
Jbryant61
4-Participant
(To:Jbryant61)

Could it be that the Rmax for the integration should be equal to the span_microns / 2 ?

Jason

http://collab.mathsoft.com/~Mathcad2000/read?126410,63.

A long work sheet but no reconciliation.
Pixelation by oversampling of a real image is all exemplified in the attached, that you might have missed. What would be interesting is the coding of the Mathcad Improc built-in zoom(M, hscale, vscale), which function is probably a standard in imagery.

Jean
Announcements

Top Tags