Higher-rank Minkowski valuations are efficient means for describing the geometry and connectivity of spatial patterns. We show how to extend the framework of the scalar Minkowski valuations to vector- and tensor-valued measures. The properties of these extensions are described in detail. We show the versatility of these measures by using simple toy models as well as real data. Our applications cover the morphology of galaxy clusters, the structure of spiral galaxies, and the geometry of molecules. Furthermore, we consider a physical ansatz closely related to higher-rank Minkowski valuations, the Rosenfeld functional known from density functional theory.