Publikation

SEEREP: towards a spatio-temporal-SEmantic Environment REPresentation for robots in an agricultural context

Mark Höllmann, Joachim Hertzberg

In: Perception and Navigation for Autonomous Robotics in Unstructured and Dynamic Environments. Perception and Navigation for Autonomous Robotics in Unstructured and Dynamic Environments befindet sich 2021 IEEE/RSJ International Conference on Intelligent Robots and Systems September 27-October 1 Prague Czech Republic unknown 2021.

Abstrakt

This paper presents a concept towards an environment representation that deals with the spatial, the temporal and the semantic domain at once. The objective of the environment representation is to make all three domains available for planning and reasoning algorithms so that robotic systems are enabled to understand a partially unstructured environment that changes over time and to act in a goal-oriented way in this environment. Agricultural environments are taken as application examples here. Robotic systems in such environments have to use their sensors for completing and updating an environment representation with data. Therefore, a focus here is on the sensor data typical for the agricultural context, like images and high-resolution as well as large-scale point clouds.

Projekte

Deutsches Forschungszentrum für Künstliche Intelligenz
German Research Center for Artificial Intelligence