IBVS4

RTSX Help: Image-Based Visual Servoing for 4 feature points

IBVS4

Syntax

  • [pmat,pt,vmat,emat,Tcmat,jcond,zvec]=IBVS4(cam, options)

Input Arguments

  • cam –a camera model

Output Arguments

  • pmat, pt — point trajectory data
  • vmat — velocity data
  • emat — point error data
  • Tcmat — camera pose data
  • jcond — Jacobian condition number data
  • zvec — depth estimation data

Options

Note : options in [ ] must be passed to the function in pairs.

  • [ ‘niter’, N] — Maximum number of iterations.
  • [ ‘eterm’, E] — Terminate when norm of feature error less than E
  • [ ‘lambda’, L] — Control gain, positve scalar
  • [ ‘T0’, T] — Initial pose
  • [ ‘P’, p] — Set of world points (3 x N)
  • [ ‘pstar’, P] — The desired image plane coordinates
  • [ ‘depth’, D] — set dept of points to D
  • ‘depthest’ — run a simple depth estimator

Description

Use IBVS4( ) shows how to implement IBVS (Image-Based Visual Servoing) for 4 feature points. Results are plotted and given as outputs.

Examples

-->cam = CentralCamera('default');
-->P = mkgrid(2, 0.5,'T',transl(0,0,2)); 
-->pStar = [ 312   312   712   712;
   312   712   712   312];
-->Tc0 = transl(1,1,-3)*trotz(0.6);
-->ibvs4(cam,'T0',Tc0, 'P', P,'pstar',pStar,'depth',1);
-->ibvs4(cam,'T0',Tc0, 'P', P,'pstar',pStar,'niter',100,'depthest');
// next commands show how rotation about camera Z axis causes camera retraction problem
-->pStar = camproject(cam,P,'Tcam',trotz(0.9*pi));  
-->ibvs4(cam, 'P', P,'pstar',pStar,'depth',2);

See Also

  • CentralCamera — create a central-projection perspective camera model
  • CamProject — project world points
  • CamPlot — plot projection of world point on image plane