On Thu, Dec 13, 2012 at 10:56:30PM +0000, Andrés Muñiz Piniella wrote:
> There is currently an ISO standard being developed (ISO/CD 13083) to
> find out what is the best method and it seems to be using canny edge
> but using sections lines.
The Canny edge detection algorithm, as I understand it,
(a) may utilise different gradient filters
(b) always results in one-pixel thin lines at the end
So I'm not sure how that would work in the standard. Unless you always
have *two* edges.
> Would you think doing a step presentation or
> RMS and doing sections on that to be a better solution? Or am I
> missing something?
Step gives sharp contours but it may emphasise the small features a bit
too much. Also, it's somewhat eccentric. For general purpose, using
the classic hypotenuse of a horizontal and vertical gradient filter is
a safer bet. For the filters, I'd suggest the optimised Scharr gradient
3/16, 0, -3/16
5/8, 0, -5/8
3/16, 0, -3/16
(and transponsed) that should lead to a more uniform angular response.
> Short from doing hundreds of sections, is there a more efficient way
> to find out the average thickness of the edges in a scan?
I don't know any. Anyway, taking hundreds of sections is easy. The
difficult part is finding automatically where exactly take them.
Regards,
Yeti
------------------------------------------------------------------------------
LogMeIn Rescue: Anywhere, Anytime Remote support for IT. Free Trial
Remotely access PCs and mobile devices and provide instant support
Improve your efficiency, and focus on delivering more value-add services
Discover what IT Professionals Know. Rescue delivers
http://p.sf.net/sfu/logmein_12329d2d
_______________________________________________
Gwyddion-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/gwyddion-users
