[Spm] A question about Image flattening

[Александр Филонов]

I think that every SPM software has a flattening function, with 2nd degree
polynomial – that is what you need.

FemtoScan Online, SPIP, WSxM, Gwyddion – you can easily find them on the
web.

Particularly, in FemtoScan Online software there is a function “planefit”
with 2nd degree polynomial, and “Bspline fit”, which is more flexible – if
you know exactly that surface is flat, It will give you better result.

 

If you’ll scan freshly cleaved mica, you’ll obtain an image of the flat
surface (F). Make a copy of the image and flatten it as much as possible
(F1). If you subtract one from another:

D=F1-F

you’ll have a displacement image, which can be used with newly obtained
images (I): 

I+D=I1

Of course, some artifacts can remain in the displacement image  - like
vertical thermal drift. 

 

 

 

 Sent: Thursday, November 27, 2008 9:41 PM
To: spm at spmlist.di.com
Subject: [Spm] A question about Image flattening

 

Hi
I am scanning with tube AFMs (Mutimode IV and PicoSPM II).
A recurring issue is that the images always have a "bow", meaning that flat
surfaces seem to be curved downwards, (i.e. parabolic).
>From what I've read, this is known issue in tube afms and occurs due to
"cross-talking between x-y and z piezos".

I have 2 related issues, and I would appreciate your advice:  

1. How to flatten the image? polynomial? what degree? what software? How
much does this flattening affects z-axis resolution?

2. How do I calibrate the AFM...? the grids always appear bowed in the
imaging, and It's quite difficult to determine z-axis heights accurately.

Thanks,
Noam.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://spmlist.di.com/pipermail/spm/attachments/20081130/8504f089/attachment-0002.html