Satellite and airborne remotely sensed images cover large areas, which normally include dozens of agricultural plots. Agricultural operations such as sowing, fertilization, and pesticide applications are designed for the whole plot area, i.e. 5 to 20 ha, or through precision agriculture. This takes into account the spatial variability of biotic and of abiotic factors and uses diverse technologies to apply inputs at variable rates, fitted to the needs of each small defined area, i.e. 25 to 200 m2 (“micro-plots”). Therefore, using remote sensing for precision agriculture the first step is to isolate the plot image where to implement the desired action, which requires: a) accurate geo-referenciation/ coregistration (accuracy < 0.2 to 0.5 m); b) splitting into small defined areas (individual trees and/ or rectangular micro-plots); c) assessing the adequate agro-environment indicator for the desired operation for each individual tree/ micro-plot. In addition, to make feasible these steps and the development of the corresponding prescription map need all to be implemented in a semi-automatic/ automatic way through specific software/ “ad-on” of the used image processing program. The aim of this paper is to discuss the use of remotely sensed imagery in precision agriculture and to outline the contribution of our research group in its development.