Resolution:bar target to sine grating conversion

[Sergey Tauger]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
Post images on http://www.imgur.com and include the link in your posting.
*****

Hello

Could you please recommend an article describing how to convert data
obtained using bar target to response to sine grating of the same period?

The articles I have at hand suggest infinite number of bars, but I remember
that we should somehow compensate for little or insufficient number of bars to
not to overestimate resolution.

Thanks in advance,
Serge

[Sergey Tauger]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
Post images on http://www.imgur.com and include the link in your posting.
*****

I found some literature on the topic:

1. J. W.Coltman, “The specification of imaging properties by response to a
sine wave input”
The first article on sine to bar target conversion. Suggests infinite-length
patterns

2. D. H. Kelly,“Spatial frequency, bandwidth, and resolution”
Shows that 3 and 4 bar targets show overestimated contrast over both
infinite bar and sine targets

3. G.D.Boreman, "Modulation Transfer Function Measurement Using Three- and
Four-bar Targets"
Discusses conversion from 3 and 4 bar targets to infinite sine pattern.

Could somebody please tell how to generalize equations from Boreman's
article to arbitrary number of bars? I can't fully get the math.

Best,
Serge

[Zdenek Svindrych-2]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
Post images on http://www.imgur.com and include the link in your posting.
*****

Hi Serge,
I don't know which equations you mean, but concerning eqns (7)-(10) I put up
a little script to clarify this.

The link to the readable document is:

https://docs.google.com/file/d/0B5vWyBYrDvcJYzVERXdVSzlobG8/edit
and the (unreadable) source is here:
https://docs.google.com/file/d/0B5vWyBYrDvcJcy1YN3FEaG1vOTQ/edit

But I think the main message of the paper is that you should use Fourier
transform and divide the spectral power of your image by the spectral power
of the ideal bar pattern (adjusted so it fits the image pixel-by-pixel).
Then you look at the frequencies characteristic for your pattern (i.e. peaks
of the original pattern spectrum or the theoretical positions of the delta-
functions of the corresponding infinite pattern; don't get distracted by the
divide-by-zero noise at frequencies where your pattern has low spectral
density).

Good luck!

zdenek





---------- Původní zpráva ----------
Od: Sergey Tauger <[hidden email]>
Komu: [hidden email]
Datum: 5. 12. 2014 18:28:58
Předmět: Re: Resolution:bar target to sine grating conversion

"*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
Post images on http://www.imgur.com and include the link in your posting.
*****

I found some literature on the topic:

1. J. W.Coltman, “The specification of imaging properties by response to a
sine wave input”
The first article on sine to bar target conversion. Suggests infinite-length
patterns

2. D. H. Kelly,“Spatial frequency, bandwidth, and resolution”
Shows that 3 and 4 bar targets show overestimated contrast over both
infinite bar and sine targets

3. G.D.Boreman, "Modulation Transfer Function Measurement Using Three- and
Four-bar Targets"
Discusses conversion from 3 and 4 bar targets to infinite sine pattern.

Could somebody please tell how to generalize equations from Boreman's
article to arbitrary number of bars? I can't fully get the math.

Best,
Serge"