Nikolaos’s research interests include vision guided robotic systems and crop detection.
- AgriFoRwArdS CDT Annual Conference (2022): Long-Term Visual Teach & Repeat in Agricultural Environments.
Other Activities and Outputs
- Represented the CDT at the school outreach University of Lincoln British Science Week event 2022.
I received B.Sc. and M.Sc. degrees in Physics and Artificial Intelligence from the University of Patras and the University of Southampton, respectively. From 2020 to 2021, I was a Computer Vision engineer in SAGA Robotics, where I was involved in the FirstFleet, GRASPberry and other UKRI projects. My main research interests include vision guided fruit harvesting and vision based navigate-on in agricultural environments. By joining the CDT I will be able to contribute academically and produce high impact research in the agri-robotics sector.
Long-Term Visual Teach & Repeat in Agricultural Environments
Visual based navigation aims to provide affordable and accurate navigation in GPS-RTK denied environments. The Visual Teach & Repeat approach differentiates itself from other visual based approaches (SLAM) as it does not need to create metric maps to achieve navigation. In the general form of VT&R systems an operator drives the robot along a desired trajectory, and the robot during the repeat phase, attempts to follow the taught trajectory by associating information from on-board visual sensors against a stored map. In order to succeed in long-term navigation, a VT&R system needs to perform the repeat phase successfully at any time of the day despite high illumination variance or even gradual changes in the environment.
Long-Term Affordable Navigation in Unseen Agricultural Environments
Current agricultural navigation systems rely on expensive sensors such as GNSS/RTK and Lidars. Visual-based navigation aims to replace these sensors with cheap monocular cameras. This technology will enable affordable and accurate navigation in GNSS/RTK denied environments.
The Primary Investigator of this research will focus on learning visual general representations of agricultural traversable paths, robust in seasonal and vegetation changes. To conduct his experiment, he will be collecting regular data from the field, and will operate robots.
Among other skills he will acquire practical experience with robotics following industry standards, experience in data and software engineering, computer vision etc.