svn co https://svn.janelia.org/penglab/projects/vaa3d/trunk/released_plugins/v3d_plugins/lobeseg

   Usage: v3d -p <lobeseg_plugin_name> -n lobeseg -i <input_image_file> -o <output_image_file> -par 
      "#c <channalNo_reference> #A <alpha: image force> #B <beta: length force> #G <gamma: smoothness force> #n <nloop> 
      #w <local win radius> [#s #p <position> #k <direction> #N <num_of_controls>]"
	 -i <input_image_file>              input 3D image (tif, or Hanchuans RAW or LSM). 
	 -o <output_image_file>             output image where the third channel is a mask indicating the regions. 
	 #c <channalNo_reference>           the ID of channel for processing (starting from 0). If unspecified, then initialized as 0.
	 #A <alpha>                         the alpha coefficient for the image force. Default = 1.0.
	 #B <beta>                          the beta coefficient for the length force. Default = 0.5.
	 #G <gamma>                         the gamma coefficient for the smoothness force Default = 0.5.
	 #n <nloop>                         the number of maximum loops of optimization. If unspecified, then initialized as 500.
	 #w <local win radius>              the radius of local window for center of mass estimation. The real win size is 2*radius+1. The default radius = 20. 
	 
	 [#h]                               print this message.
	 [#s]                               lobe segmentation on one side, -p -k -N parameters is needed.
	 [#p]                               set the relative initial position, format X1xY1+X2xY2, e.g. 30x0+100x50, top left is the origin.
	 [#k]                               specifiy which direction will be kept, left or right.
	 [#N]                               the number of control points.
	 [#S]                               output the seperating surface.

   e.g.	 v3d -x liblobeseg_debug.dylib -f lobeseg -i input.tif -o output.tif -p "#s #N 20 #p 30x0+100x30 #k left #S"