Modelling of tree crowns with realistic morphological features: New reconstruction methodology based on Iterated Function System tool

For a large number of environmental studies, it is necessary to modelize numerical vegetal sets as trees or tree crowns. Studies on forest fire propagation or analysis of the photosynthesis in vegetation canopy are some examples. The first step for these studies is to construct vegetation like shapes. The aim of this paper is to propose a new methodology allowing to generate a vegetal set by a fractal approach. Iterated Function Systems (IFS) are associated with a genetic algorithm in order to determine the IFS parameters with respect to some morphological properties of real tree crowns as the Leaf Area Index (LAI) vertical distribution. This new methodology is applied and validated on fifteen deciduous tree and two of them are fully developed: Quercus prinus and Populous tremuloides. A sensitivity study is carried out for the IFS parameters to better understand the influence of each ones on the tree reconstruction. This new methodology has been applied on various tree species in order to provide a data base summarizing the IFS parameters to easily generate realistic tree crowns for further works.