[Therion] Loop closure / deviation of broken clino

[Xavier Pennec]

Hi Stelios,

if you believe that your clino measure is plus or minus 10 degrees (in 
standard deviation, meaning that 95% of the measures at p/m 20 degrees), 
then you state before your data:
   sd clino 10 degrees

You can put a higher value, for instance 1000 degrees or even more, 
which means that all measurements (which will be between -90 and 90) are 
in the flat part of the Gaussian, i.e. they have the same probability. 
With such a sd, the (weighted) least-squares  loop closure algorithm 
will preferably change the clino measurement than any other (provided 
that this allows closing the loops). Now therion will report the closure 
error in meter and percentage error and the only difference should be a 
decrease of error in X and Y and an increase in Z (but usually little 
change on the norm of the 3D error).

Hope that helps

Xavier

Le 04/03/2014 15:52, Stelios Zacharias a écrit :
> Dear all,
>
> I recall a discussion here about using a fairly broad sd value to 
> allow loop closure when trying to transfer paper surveys into Therion 
> - I may be misremembering.
>
> I recently surveyed a closed loop with a broken clinometer (spinning 
> wildly with no fluid to damp the motion), so for each leg of the 
> survey I have distance and azimouth accurately and slope as an 
> approximation. How would I set up the sd (if indeed this is how I 
> would do it) to close the loop and correct my approximate / estimated 
> slopes?
>
> I look forward to hearing from you.
> Stelios Zacharias
>
> _______________________________________________
> Therion mailing list
> Therion at speleo.sk
> http://mailman.speleo.sk/mailman/listinfo/therion

-- 
> -------------------------
> Xavier Pennec
> Senior Research Scientist / Directeur de recherche
> Asclepios project-team, INRIA Sophia-Antipolis
> 2004 Route des Lucioles, BP93
> F-06902 Sophia-Antipolis Cedex, France
> +33 4 92 38 76 64
> +33 6 78 35 16 90
> http://www-sop.inria.fr/asclepios/
> -------------------------------