Re: using a scanner as a densitometer?

I used MATLAB because that's my language, and price has not been an
issue because I use MATLAB for a LOT of other things (such as
computing odds for Texas Hold'em). For the sort of applications
discussed here, you only need a student version (USD100) and image
processing toolkit. Their statistics toolkit and signal processing
toolkit are useful but one can program necessary piece in not much
time, I suppose.

Anyway, the point in my previous post was that, without averating of
multiple pixels, the measurement error may be acceptably high. With a
reasonably large measurement aperture, I get repeatable reading to
0.01 density unit. How you manipulate the numerical data is
unimportant as long as the data are processed correctly.

I have no experience, btu Octave is a well known knock-off for MATLAB.

From: Yves Gauvreau <gauvreau-yves@sympatico.ca>
Subject: Re: using a scanner as a densitometer?
Date: Tue, 21 Feb 2006 15:13:37 -0500

> Ryuji,
>
> instead of Matlab (not cheap) you can use R which is free and loaded with
> statistical package to do just the kind of regression you suggest. The
> learning curve for both is relatively steep though. You could probably use a
> Excel or similar as well and maybe this page could help others as well
> (http://www.normankoren.com/color_management_4.html) for formulas amoung
> others.
>
> Regards
> Yves
Received on Thu Feb 23 00:21:46 2006