Simulation and characterization of tree branching morphologies based on real tree images

Abstract

L-systems was used to generate ideal and realistic tree branching. Morphological symmetry and self-similarity were exploited to show global development in terms of local rules. A method to simulate growth of ideal and realistic 2D trees with parameters branching angle a and branching ratio b is presented. We adopt a technique for morphological characterization of tree foliage by means of spatial analysis. From ideal trees, a-dependence of foliage profile and b-dependence of foliage mapping was exhibited. Modeling realistic trees was then explored by setting distribution-based parameters. Images of real Acacia trees were sampled and characterized to obtain P_a and P_b probability distributions implemented to realistic tree simulations. Incorporating internal and environmental factors can still be made to provide deeper understanding of principles governing branching biological systems.