Modelling fuzzy combination of remote sensing vegetation index for durum wheat crop analysis

The application of new technologies (e.g. Internet of Things, mechatronics, remote sensing) to the primary sector will reduce the production costs, limit the waste of primary materials, and reduce the release of polluting compounds into the environment. Precision agriculture (PA) has been growing in the last years thanks to industry efforts and development of applications for diagnostic purposes. Many applications in PA use vegetation indices to measure phenology parameters in terms of Leaf Area Index (LAI). In this context, the correlation of some vegetation indices were analyzed with respect to the durum wheat canopy, evaluating two different phenological stages (elongation and maturity). The results show that for the first stage of growth, the Enhanced Vegetation Index (EVI) was the best-correlated vegetation index with LAI, while the Land Surface Water Index (LSWI) was more reliable for the following stage of growth. Considering trials findings, a fuzzy expert system was developed to combine EVI and LSWI, obtaining a new combined index (Case-specific Fuzzy Vegetation Index) that better represents the LAI in comparison with the single indices. Thus, this approach could give place to a better representative vegetation index of a different biological condition of the plant. It may also serve as a reliable method for wheat yield forecasting and stress monitoring.