sharpen the egdes of an image in labview

[KevinKM]

Hi all,

I have a 2D image, i am trying to extract a suitable ROI from an image, smoothen the extracted image

I am using IMAQ Convolute VI which uses Linear filter for smoothing the image.

I am WA Multiscale Edge Detection VI to extract the edges of the signal.

Using the Edge data i am doing curve fitting (2nd order polynomail curve fitting) using MATLAB script node. 

Finally i comapare the curve fit data with the designers reference data.

The problem i am facing here when i do the comparison, the error difference is large at the edges of the curve when compared to the centre of the image. (accetable error range is 1um). i am getting it in the range of 4um at the edges.

I need help regarding,

How to sharpen the edges of the image?

Can i apply Bilateral Filter OR Anisotropic Diffusion filter using LabVIEW? 

I apolozise since i cannot share any kind of data due to security resaons.

Any help would be greatly appreciated.

Regards,

Manjunath Kamasani

[BruceAmmons]

Why are you filtering the image before finding the edges?  The edge filter will already do its own smoothing, so you are probably filtering out the important data.

Can you even make a simple sketch that shows what you mean?  I am not following when you describe the center and edge of the curve.  Is the object thicker in the center, or uniform vertical thickness?  Are you just scanning across the X axis to find Y coordinates?

Bruce