Point clouds, mesh representations and Gazebo models of everyday objects.
The object point clouds available on this website were created from low-cost RGBD sensor data without user interaction, see our first publication:
Razvan-George Mihalyi, Kaustubh Pathak, Narunas Vaskevicius, Tobias Fromm and Andreas Birk
"Robust 3D Object Modeling with a Low-Cost RGBD-Sensor and AR-Markers
for Applications with Untrained End-Users"
In Robotics and Autonomous Systems, April 2015
Tobias Fromm, Christian A. Mueller and Andreas Birk
"Unsupervised Watertight Mesh Generation for Physics Simulation Applications Using Growing Neural Gas
on Noisy Free-Form Object Models"
In arXiv.org, https://arxiv.org/abs/1603.00663, September 2016
Each object's point cloud and mesh can be viewed live. We provide zipped .bag files for the objects which were used in our modeling approach as well as .zip files containing: