A 3-D simulation with virtual stereo rig for centrifugal fertilizer spreading Bilal Hijazi 1&2 , Jürgen Vangeyte 2 , Frédéric. Cointault 1 * , Michel Paindavoine 4 , Jan Pieters 3 1 AgroSup Dijon – UP GAP – 26, Bd Dr Petitjean – BP 87999 – 21079 Dijon cedex – France 2 Institute for Agricultural and Fisheries Research (ILVO) - Technology and Food Science – Agricultural Engineering – Burg. Van Gansberghelaan 115 – 9820 Merelbeke – Belgium 3 Department of Biosystems Engineering – Faculty of Bioscience Engineering - Ghent University – Coupure Links 653 - 9000 Gent - Belgium 4 LEAD – UMR CNRS 5022 – University of Burgundy – Pôle AAFE – BP 26513 – 21065 Dijon cedex – France *Corresponding author: f.cointault@agrosupdijon.fr Abstract Stereovision can be used to characterize of the fertilizer centrifugal spreading process and to control the spreading fertilizer distribution pattern on the ground reference. Fertilizer grains, however, resemble each other and the grain images contain little information on texture. Therefore, the accuracy of stereo matching algorithms in literature cannot be used as a reference for stereo images of fertilizer grains. In order to evaluate stereo matching algorithms applied to images of grains a generator of synthetic stereo particle images is presented in this paper. The particle stereo image generator consists of two main parts: the particle 3D position generator and the virtual stereo rig. The particle 3D position generator uses a simple ballistic flight model and the disc characteristics to simulate the ejection and the displacement of grains. The virtual stereo rig simulates the stereo acquisition system and generates stereo images, a disparity map and an occlusion map. The results are satisfying and present an accurate reference to evaluate stereo particles matching algorithms. Keywords: fertilizer centrifugal spreader, stereovision, image processing. 1. Introduction With respect to centrifugal fertilizer spreading, many factors, such as construction and calibration of the machinery, particle types and properties, field conditions, etc. influence the distribution pattern in the field (Cointault et al., 2008; Hijazi et al., 2010b; Hijazi et al., 2011; Van Liedekerke et al., 2009). Maladjusted centrifugal spreaders create heterogeneous distributions with economical and environmental consequences. In previous work (Hijazi et al., 2010a; Hijazi et al., 2010c), the feasibility of stereovision to characterize the fertilizer centrifugal spreading process and to control the spreading fertilizer distribution pattern on the ground was already demonstrated. Imaging systems, specifically stereo rig, are lately being more widely used in several domains (Ayache, 1989; Durdle et al., 1998; Hori, 2004; Michoud, 2009). Several procedural steps are necessary in stereovision: camera calibration, image acquisition, image rectification, image matching and depth determination. The major task in stereo computing is image matching. Image matching is the determination of the correspondence between pixels of the right and left images. If matching is accurate, depth calculation will be accurate as well. However, image matching is difficult. Its accuracy depends on several factors such as image resolution, noises, texture, etc. Fertilizer grains all look alike and their images contain little information on their texture. As a result the accuracy of stereo matching algorithm in literature cannot be used as a reference for a stereo images of fertilizer grain. In order to evaluate stereo matching algorithms applied to grain images in this paper a generator of synthetic stereo particle images is presented.