Field phenotyping has become of vital importance for the evaluation of plant adaptations to an always changing environment. Monitoring vegetation under these conditions requires the development of new flexible tools capable to assess different plant traits in a fast and non-invasive manner.
The FieldLift (Fig. 1) offers the unique opportunity to perform combined measurements of structural und functional properties of the canopy based on different imaging and measurement techniques, like 3-d, hyperspectral or thermal imaging. In addition, this set up opens up new perspectives for the development of tools able to merge registered data from different sensors (sensor fusion). The FieldLift is capable to perform measurement from 1 to 10 meters above the ground and its articulated arm provides great flexibility to cover different positions above the canopy. So far, the FieldLift has been applied with a stereo camera setup and a hyperspectral imaging system, but others sensors may be added as well up to a payload of 200 kg.
The hyperspectral camera system (Fig. 2) consists of two cameras and provides spatially and spectrally resolved information between 400 to 2500 nm with a nominal spectral resolution of 2.8nm for the VIS/NIR and 10nm for the SWIR. Among the outputs of the hyperspectral images there are different maps of vegetation indices used to estimate plant biophysical properties like chlorophyll and water content. In addition, dynamics adaptations of plant photosynthesis can be assess through the estimation of sun-induced fluorescence at 760nm (a fluorescence map of sugar beet plants is depicted in Fig. 2).
Structural properties of the canopy are retrieved by the stereo imaging system (Fig. 3). Acquired images are used to compute depth maps (a depth map of sugar beet plant populations is illustrated in Fig. 3) which are further used to estimate 3-d models of the canopy with software developed in-house that also provides information on leaf orientations and leaf angle distributions.