Water stress assessment can be performed by analyzing thermal images taken onboard Unmanned Aerial Vehicles (UAVs). This study focuses on the acquisition and data extraction of high-resolution UAV-sensed thermal images. The datasets obtained, through computation of spectral indices and image classification, allowed to assess the response to drought of an apple hybrid population submitted to different water regimes. Studies were performed in an experimental apple orchard located in southern France. Flights were planned at solar noon on four successive dates during summer 2013 (40 m altitude, 0.10 m spatial resolution) and at five successive hours of the day, once water stress was established. The high temporal and spatial resolution of images allowed acquiring data at canopy and intra-canopy scales, with a short revisit time. As the miniaturized uncooled thermal camera carried onboard the UAV needs careful correction of radiometry, this was performed by continuous reference to ground thermal targets, while an automatized image processing was carried out. Thanks to the high resolution of the remote images obtained, and the capacity to efficiently delineate each individual tree within the whole trial, it was possible to analyze inter- and intra-canopy thermal variations. Indices extracted from thermal images showed significantly higher canopy temperatures in water restricted trees compared to well-irrigated ones. These differences were related to the severity of water deficit. However, responses also varied significantly according to the genotype. The image-based variables in apple trees constitute a basis for a further finely tuned analysis of the differential response to water stress.