Progress report: Use of satellites and drones for estimation of forage quantity and quality
This body of work can be summarized as follows:
1) Drones and satellite can complement each other. Normalized vegetation difference index (NDVI) as an indication of plant vigor and calculated from satellite and drones can be used in large areas, where manual recording of pastures vigor is not possible or logistically difficult.
2) However, NDVI can be used as a proxy for biomass or height at early stages of plant growth (i.e. during establishment and when leaf growth is greater than stem elongation) but not at later stages.
3) The best classifier for identifying pastures is Random Forest (RF).
4) Ash concentration, standing biomass, crude protein and digestibility can be predicted from high resolution satellite imagery (Planetscope, 3 m pixel) using random forest regressions.
5) Red and NIR bands from satellite sensors (Plantscope) were found the most important bands to predict ash concentration, standing biomass, crude protein
6) Nowadays, phenotyping of shoot traits (cover area, greenness and volume of plants) using drones of large number of plants is routine work
Cardoso Arango, Juan Andrés.