IGARSS 2022 - 2022 IEEE International Geoscience and Remote Sensing Symposium · 2022 · 13 citations · 7 references
AgricultureAgronomyCrop SciencesCrop SensingField PhenotypingRemote SensingAgricultural ApplicationsHyperspectral ImagingRemote Sensing For PhenotypingSensor Data AcquisitionVegetation IndicesVegetation MonitoringHyperspectral Vegetation IndicesMultispectral Image SensorsMultispectral Imaging SensorsMultispectral Vegetation IndicesPhenological Vegetation IndicesPhenology ObservationUav-based PhenotypingUnmanned Aerial Vehicle Sensing
In-situ monitoring of seasonal crop development is important to provide improved decision-making tools for farmers. In recent years, low altitude unmanned aerial vehicles (UAVs) coupled with ultra-high-resolution sensors have continued to advance the provision of detailed characterisation of in-season crop dynamics at a field scale. However, a ‘bottleneck’ still exists in many high-throughput plant phenotyping (HTPP) approaches to allow for accurate and timely extraction of information from multispectral sensors onboard UAVs. Here, we propose a pipeline for UAV data extraction using an executable software code i.e., ‘Xtractori’: The Xtractori was designed to handle BIG DATA generated from hyperspectral and multispectral sensors on board mobile tractors and/or drones. We used data from a high-resolution, six- band MicaSense Altum camera to capture images at three different winter crop sites during the 2021 cropping season across Australia. Vegetation indices (VIs) and canopy temperature were then extracted using Xtractori to evaluate the phenological development of different winter crops (wheat, barley, chickpea, canola, and oats). Results indicated a clear discrimination of phenological development including peak and saturation of different crops across environments. The VI-based phenology estimation was also closely related to the field phenology observation. Furthermore, VIs at peak crop development had a significant negative association with ΔT (canopy temperature - air temperature) (r= −0.45 to −0.88), suggesting this is likely an important crop development stage to differentiate crop stress. The study showed that including Xtractori in HTPP framework enabled the rapid and accurate extraction of sensing metrics and dynamic crop traits from multispectral data. This will aid researchers in enhancing their understanding and knowledge of seasonal crop growth dynamics across different environments.
7
Overview of the radiometric and biophysical performance of the MODIS vegetation indices
Alfredo Huete, Kamel Didan, Tomoaki Miura et al. · Remote Sensing of Environment · 2002 · 9.6K citations
Agriculture, Ecology, Geography +14
Optimization of soil-adjusted vegetation indices
G. Rondeaux, M. D. Steven, Frédéric Baret · Remote Sensing of Environment · 1996 · 2.8K citations
Automated crop plant counting from very high-resolution aerial imagery
João Valente, Bilal Sari, Lammert Kooistra et al. · Precision Agriculture · 2020 · 109 citations · Full text