The ability to virtuously combine profitability and quality of productions with increasingly ambitious levels of environmental sustainability is the main long-term challenge of nowadays agriculture. This necessarily requires the application and adaptation of the best technical practices, the use and optimization of the most suitable technologies, and the further development of innovative solutions. This study was conducted on a maize field at the ‘Angelo Menozzi’ experimental farm (Landriano, Italy). The aim of the experiment was to compare different sensing approaches based on proximal and remote sensed data for planning site-specific weed control treatments and evaluating the quality of the results obtained. Remote sensed data were obtained from an RGB camera of an unmanned aerial vehicle (UAV) used to conduct aerial surveys at 40 m altitude on the experimental field. Proximal sensed data of the same field were acquired with a tractor-mounted digital camera, allowing to obtain very high-resolution images of crops and weeds. The comparison between proximal and UAV images was validated through ninety georeferenced images used as ground-truth, that were acquired withstanding very-high resolution and perpendicular to the different areas of the field selected for sampling and annotated manually. The acquired data were processed with custom programs aimed to obtaining weed infestation map of the field based on excess green vegetation index. The results are compared and discussed against the ground-truth.