A comparison of bivariate classification and segmentation approaches to delineating and interpreting grain yield-protein management units
Résumé
A univariate segmentation algorithm has recently been developed for precision agricultural applications. This is adapted to a bivariate analysis to invest zoning based on yield and protein response in an eastern Australian wheat field. The intention is to be able to provide a zone-by-zone interpretation of the agronomic response to N. The segmentation algorithm provided comparable management zone results with the more widely used k-means classification. The algorithm is still under development but allows expert-knowledge to be incorporated into the zone delineation process