Computer Vision for robotics and autonomous driving

We teach computers to see like humans do
... just based on cameras & Artificial Intelligence by offering:

  • an innovative 3D-depth sensor that is cost-effective, real-time capable and with a big field-of-view under high-resolution

  • a robust Freespace Detection against arbitrary obstacles

  • a full collision avoidance system for robotics and autonomous driving

Our perception system
Our perception system will extract the 3D-depth and further information from two camera sensors only. This teaser demonstrates the free-drivable space (green) that is extracted in real-time.

* actual DeeVisio Freespace Detection in real-time

 
What we do

We offer a solution that uses camera sensors combined with Artificial Intelligence to enable a low-cost perception system that is able to identify the environment in a detailed and reliable way.

Thereby we offer an innovative 3D-depth sensor that is based on two camera sensors only. This enables depth perception similar to real 3D-LiDAR sensors, but with the advantage of being much more cost-effective, scalable, real-time capable and with a big field-of-view under high-resolution. Furthermore, it can be used as a redundancy against other 3D-sensors for safety critical applications.

In combination with our robust Freespace Detecion System we can enable a full collision avoidance system that detects free-drivable space and all arbitrary obstacles in the environment for robotics and autonomous vehicles.

DeeVisio is funded by the Federal Ministry of Economics and Energy and the European Social Fund as part of the EXIST start-up grant