Tawny : Différence entre versions

De MicMac
Aller à : navigation, rechercher
(Example)
Ligne 3 : Ligne 3 :
 
The job of Tawny is essentially to merge these data and to optionally do some radiometric equalization.
 
The job of Tawny is essentially to merge these data and to optionally do some radiometric equalization.
  
For the radiometric equalization, Tawny will compute for each individual ortho image O_i a polynom P_i such that, ∀(i, j, x, y) where ortho image O_i and O_j are both defined in x, y we have the relation:
+
For the radiometric equalization, Tawny will compute for each individual ortho image <math>O_i</math> a polynom <math>P_i</math> such that, <math>∀(i, j, x, y)</math> where ortho image <math>O_i</math> and <math>O_j</math> are both defined in at the ground coordinates <math>(x, y)</math> we have the relation:
* O_i(x, y)*P_i(x, y) = O_j(x, y)*P_j(x, y)
+
*<math>O_i(x, y)*P_i(x, y) = O_j(x, y)*P_j(x, y)</math>
  
 
The problem with such formula is that it can lead to important drift in radiometry. So there is also a global polynom R that is computed, this polynom is such that:
 
The problem with such formula is that it can lead to important drift in radiometry. So there is also a global polynom R that is computed, this polynom is such that:
* O_i(x, y)*P_i(x, y)*R(x, y) = O_i(x, y)
+
*<math>O_i(x, y)*P_i(x, y)*R(x, y) = O_i(x, y)</math>
  
 
The radiometry of each image used for the ortho photo will finally be O i (x, y)P i (x, y)R(x, y). Of course for equation 3.1 and 3.2, there is much more observations than unknowns and they are solved using least mean square. The user can control the radiometric equalization by specifying the degree of the polynom.  
 
The radiometry of each image used for the ortho photo will finally be O i (x, y)P i (x, y)R(x, y). Of course for equation 3.1 and 3.2, there is much more observations than unknowns and they are solved using least mean square. The user can control the radiometric equalization by specifying the degree of the polynom.  

Version du 27 janvier 2016 à 17:33

Description

The simplified tool for generating ortho mosaic is Tawny, it is an interface to the Porto tool. The use of Tawny is quite simple because it assumes that the data have been correctly prepared and organized during the matching process. Practically this is done when the matching has been made using Malt and it is recommended to only use Tawny in conjunction with Malt. In Ortho Mode, Malt has created a set of individual ortho images, associated mask, incidence image, . . . in a directory Ortho-MEC-Malt/ The job of Tawny is essentially to merge these data and to optionally do some radiometric equalization.

For the radiometric equalization, Tawny will compute for each individual ortho image [math]O_i[/math] a polynom [math]P_i[/math] such that, [math]∀(i, j, x, y)[/math] where ortho image [math]O_i[/math] and [math]O_j[/math] are both defined in at the ground coordinates [math](x, y)[/math] we have the relation:

  • [math]O_i(x, y)*P_i(x, y) = O_j(x, y)*P_j(x, y)[/math]

The problem with such formula is that it can lead to important drift in radiometry. So there is also a global polynom R that is computed, this polynom is such that:

  • [math]O_i(x, y)*P_i(x, y)*R(x, y) = O_i(x, y)[/math]

The radiometry of each image used for the ortho photo will finally be O i (x, y)P i (x, y)R(x, y). Of course for equation 3.1 and 3.2, there is much more observations than unknowns and they are solved using least mean square. The user can control the radiometric equalization by specifying the degree of the polynom.

Syntax

The global syntax for Tawny is
mm3d Tawny Directory NamedArgs

Help

A basic help can be asked with mm3d Tawny -help


Mandatory unnamed args :

  • string :: {Data directory}

Named args :

  • [Name=DEq] INT :: {Degree of equalization (Def=1)}
  • [Name=DEqXY] Pt2di :: {Degree of equalization, if diff in X and Y}
  • [Name=AddCste] bool :: {Add unknown constant for equalization (Def=false)}
  • [Name=DegRap] INT :: {Degree of rappel to initial values, Def = 0}
  • [Name=DegRapXY] Pt2di :: {Degree of rappel to initial values, Def = 0}
  • [Name=RGP] bool :: {Rappel glob on physically equalized, Def = true}
  • [Name=DynG] REAL :: {Global Dynamic (to correct saturation problems)}
  • [Name=ImPrio] string :: {Pattern of image with high prio, def=.*}
  • [Name=SzV] INT :: {Sz of Window for equalization (Def=1, means 3x3)}
  • [Name=CorThr] REAL :: {Threshold of correlation to validate homologous (Def 0.7)}
  • [Name=NbPerIm] REAL :: {Average number of point per image (Def = 1e4)}
  • [Name=L1F] bool :: {Do L1 Filter on couple, def=true (change when process is blocked)}
  • [Name=SatThresh] REAL :: {Threshold determining saturation value (pixel >SatThresh will be ignored)}
  • [Name=Out] string :: {Name of output file (in the folder)}

Example

For example in the Mur Saint Martin dataset (or whenever "Malt Ortho" was used without the "DirMEC" option changed), you can launch :
mm3d Tawny Ortho-MEC-Malt/