Login

Proceedings

Find matching any: Reset
Add filter to result:
Spatial-temporal Evaluation of Plant Phenotypic Traits Via Imagery Collected by Unmanned Aerial Systems (UAS)
S. Varela, G. Balboa, V. Prasad, A. Ferguson, T. Griffin, I. Ciampitti
Kansas State University

Unmanned aerial systems (UAS) and a stereovision approach were implemented to generate a 3D reconstruction of the top of the canopy. The 3D reconstruction or CSM (crop surface model) was utilized to evaluate biophysical parameters for both spatial- and temporal-scales. The main goal of the project was to evaluate sUAVs technology to assist plant height and biomass estimation. The main outcome of this process was to utilize CSMs to gain insights in the spatial-temporal dynamic of plants within the experiment.  Four experiments were carried out during 2015 corn growing season and utilized for ground-truth validation. The experiments were located at Ashland Bottoms Research Farm, Kansas State University (Manhattan, KS). Flight missions and ground-truthing were accomplished at two critical stages of biomass accumulation. At known locations individual plant height and biomass were measured and correlated to plant height estimation at the same locations in the CSMs. Same samples plants were then compared to a stem volume estimation by mixing estimated plant height from the CSM and ground stalk diameter from the same plants and locations. Plant height correlation was stronger at flowering stage than 2 weeks –prior flowering. Plant biomass estimation became stronger by adding ancillary field data. Per-plant basis results suggested that the CSMs could assist prediction of biophysical variables. Outcomes from this study confirmed that UASs could assist predicting “key” traits (“on-farm rapid phenotyping”).

Keyword: remote sensing, unmanned aerial systems, site-specific, corn, plant phenotyping