Software

Even the best hardware needs software. Only then can robots process impressions, make decisions, plan tasks, and interact with humans or other robots.

We make some programs developed at the Institute of Robotics and Mechatronics available free of charge (usually as open source software) so that other researchers can use them as well. By doing so, we would like to contribute to the advancement of the robotic research community.

On this website we present a selection of software projects. More projects are published on https://github.com/DLR-RM.

3D Object Tracking

In this repository, we share several algorithms and code of our ongoing research on 3D object tracking.

AugmentedAutoencoder

Augmented Autoencoders (AAEs) are an approach developed at the institute to train neural networks for object pose estimation on synthetically rendered images from 3D models.

BlenderProc

BlenderProc is a software to generate realistic images for the training of neural networks.

DLR CalDe and DLR CalLab

"DLR CalDe and DLR CalLab" is a camera calibration toolbox that implements the well-known method of Zhang, Sturm and Maybank.

cvkit

The cvkit software package is offered for supporting an easy exchange and analysis of computer vision data including images in floating point format and 3D data like point clouds or triangle meshes. The toolkit targets researchers as well as interested groups and individuals. It is also especially useful for working with the Middlebury stereo and multi-view data.

RAFCON

RAFCON (RMC advanced flow control) is a software tool that allows hierarchical state machines to be programmed visually using a feature-rich graphical user interface (GUI).

OAISYS

OAISYS is a simulator that generates images and metadata for unstructured outdoor environments, specifically considering the requirements of planetary robotics.

Stable Baselines3

Stable Baselines3 provides reliable open-source implementations of deep reinforcement learning (RL) algorithms in Python.