Research project

The analysis of crop interaction with the environment is object of research in agronomy with increasing focus on the observation of individual plants. Besides morphological parameters like number of leaves, their orientation and curvature, the acquisition of the plant architecture, i.e. the composition of the plant from single parts, is of interest. As a manual determination of these parameters is very time-consuming, this projects aims to develop an automatic approach for measuring plants. In this project research will be done on the measurement system as well as the algorithms to process the measured data: The optical measurement of plants in the field is quiet challenging due to bad environmental conditions, which may result in poor quality of the measurement data. Furthermore the measurement setup and selection of the relevant measurements must be dynamically adjusted to the plant's particular growing stage. For the further processing steps it must be considered that plant regions may be occluded and thus cannot be optically acquired. To determine parameters of the plant's habius however, the extrapolation of the occluded areas is required using prior knowledge about the appearance of the observed species. A model-based approach will be developed to hold this knowledge in a geometric plant representation which will be automatically fitted to the measured data to amend missing areas.