DLO_following

[Project Website]

Repository for the IROS 2024 paper "In-Hand Following of Deformable Linear Objects Using Dexterous Fingers with Tactile Sensing".

In this repo, we provide the code for

The optimization-based inverse kinematics (IK) solver.
The tactile-based in-hand 3-D DLO pose estimation approach.

We do not recommend developing your project directly based on our code as it has not been optimized. The code is mainly released for reference or a quick test run.

Installation

git clone https://github.com/Mingrui-Yu/DLO_following.git

In your python3 env:

pip install matplotlib scikit-learn open3d pybullet imageio opencv-python

Usage

Test the IK solver

cd DLO_following

python test_ik.py global_rotation
# or
python test_ik.py v_shape_rotation

The global_rotation and v_shape_rotation are corresponding to the experiments in the paper.

Test the in-hand 3-D DLO pose estimation

We provide the data of two sequences for a quick test run. The two sequences are the same as those shown in the supplementary video.

cd DLO_following

python test_cable_sensing.py seq_0
# or
python test_cable_sensing.py seq_1