CHARACTERIZING OLIVE TREE GEOMETRIC FEATURES USING UNMANNED AERIAL VEHICLE (UAV) IMAGES J. Torres-SÆnchez 1*, , F. Lpez-Granados 1 , N. Serrano 2 , O. Arquero 2 , R. FernÆndez- Escobar 3 , J.M. Peæa 1 1. Department of Crop Protection, IAS-CSIC, 14080 Crdoba, Spain 2. IFAPA Alameda del Obispo, Avda MenØndez Pidal s/n 14071, Crdoba, Spain 3. Department of Agronomy, UCO, 14080 Crdoba, Spain Abstract Olive tree geometric features such as height, diameter, area and volume serve for monitoring crop status and are input variables in crop production models. Traditionally, these variables are estimated after an intensive field work and applying equations that treat the trees as geometric solids, which may produce inconsistent results. As an alternative, this work present an innovative automatic method for olive tree characterization based on two phases: 1) close range photogrammetry from Unmanned Aerial Vehicles (UAV) and 2) use of object based image analysis (OBIA) techniques. The presented methodology was tested in two olive orchards in southern Spain. 100% of the olive crowns were reconstructed by the photogrammetric software, and 100% of the olives trees were detected by the OBIA algorithm.This method could be adapted to other similar woody crops allowing a notably reduction in time and labor of measuring tasks. Keywords Olea europaea, stereoscopy, tree morphology, remote sensing, OBIA Introduction Olive tree geometric features such as height, diameter, area and volume are used by olive researchers and producers for inventory and crop monitoring. They serve as input variables for modelling crop production and a number of agronomic tasks, e.g., water requirements, fertilization, design of pruning, among others. Tree height and diameter are traditionally measured after an intensive field work and, subsequently, apparent crown volume and surface are estimated from these measures by using equations that treat the trees as geometric models (Pastor Muæoz-Cobo, 2005), which may produce inconsistent results. An alternative to the traditional field-based methods is using Digital Surface Models (DSMs) derived from information of sensors installed on remote platforms. Most of researcherses have applied remote sensing techniques to derive height of forest trees by using passive imagery from airborne platforms (Wallerman et al., 2012) and satellites (Takahashi et al., 2012), as well as active laser systems (Kaartinen et al., 2012). By using a similar approach, Zarco-Tejada et al. (2014) estimated olive tree height from DSMs generated with remote images captured with an Unmanned Aerial Vehicle (UAV). Several investigations have demonstrated the advantages of the UAV platforms in comparison with airborne or satellite missions, regarding a minor cost and