ami-iit / mvnx-to-urdf Goto Github PK

• Installation
• Prerequisites
• Una-tantum procedure
• How to generate model
• What you need
• How to use it

• Matlab
• iDynTree: a library of robots dynamics algorithms for control, estimation and simulation.

• Install the robotology-superbuild.

  git clone https://github.com/robotology/robotology-superbuild.git
  mkdir build
  cd build
  

  Follow installation instructions here. The required library iDynTree will be automatically install via the superbuild.
• Enable the Matlab bindings for iDynTree.

  cd robotology-superbuild/build/src/iDynTree
  cmake ..
  ccmake .
      ROBOTOLOGY_ENABLE_MATLAB    ON
      IDYNTREE_USES_MATLAB        ON
      Matlab_ROOT_DIR             root-to-your-application-Matlab
  make install
  

  • In your .bashrc add export Matlab_ROOT_DIR="root-to-your-application-Matlab".
  • Open Matlab and browse to the folder your-path/robotology-superbuild/build.
  • Run the script Startup-robotology-superbuild.m.

• A data folder (e.g., data) containing your dataset, per each subject (e.g., S00X) with
  • a .mvnx file for the human whole-body acquisition.
  • a .mat file with markers acquisition for the hand (if you wantto build the model with articulated hands)
• The folder templates containig human templates for URDF models.

• Move to the folder where you clone the repo.

• Run the script computeHumanURDF.m

• Fill the dialog box with required info.

  

• If you do not want to include the articulad model of the hand into the URDF, insert 'n'. This means that the hand will be represented only by a unique rigid box in the URDF.

• If you want to include the articulad model of the hand into the URDF, insert 'y'. This will open another dialog box, as follows:

• With the second dialog box you can choose if building
  • an anthropometric model of the hand. This model does not require the markers acquisition, but only the .mvnx.
  • a marker-driven model of the hand. This model also requires an acquisition of markers properly positioned on top of the hand with a standard optical cameras tracking (e.g., Vicon).

Given

• the subject number (e.g., 1)
• the subject mass (e.g., 60)
• the subject height (e.g.,1.80)

OPT	48 DOF	66 DOF	ANTHROPOMETRIC HAND	MARKER-DRIVEN HAND	GENERATED URDF
1	✔️
2		✔️
3	✔️		✔️
4	✔️			✔️