Research On Key Technologies Of Plant Modelling Based On Intelligent Virtual Organ

As an emerging interdisciplinary field, the research of computer-simulation-based virtual plants combines artificial intelligence, botany and plant physiology as well as other subjects. It is an important part of the national digital agriculture strategic planning, and is the core content of agricultural informationization also, and has great theoretical and practical value. However, there are still some problems and limitations between research and application, such as the function and structure model of poor coupling and difficult to configure complex parameters, being difficult to discover and extract the growth rules automatically, difficult to present complicated functional structure of organs by using the traditional L-system, etc.This thesis focuses on the research of plant modeling and its key technologies to solve the limitations of existing research. The most important feature of this thesis is that the model based on Intelligent Virtual Organ (IVO) is proposed, IVO implements the growth reasoning process to simulate the physiological function and growth deformation independently, the Growth Controller (GC) implements the intelligent growth reasoning process to achieve the growth and development of the virtual plant. Main contributions are as follows:First, collecting physiology data of plant, statistical analysis and Fuzzy Neuron Inference of Physiological Process (FNI-PP) have been used to automatically extract the growth behaviors rules of organs, the physiological growth status of organ and the physical and environmental restrictions for development. Discrete pressure-flow model and graph automata have been used to simulate the material transport and distribution and information exchange between IVOs. The autonomous physiological processes of IVOs have been established to achieve physiological functions.Second, bézier surface and 3D reconstruction technique have been used to create 3D geometric model of organs. Wang Tiles texture synthesis algorithm has been used to obtain 3D rendering vivid images. Adjusting geometrical parameters, 3D Morphing and growth function have been used to implement the growth deformation of virtual organs. IVOs calculate the best angle of new child organ by integrated reasoning process, in which the child organ can maximize light interception.Third, the Growth Controller (GC) has been established to control the growth and development of the virtual plant. The branching network composed of IVOs has been established, to store the virtual plant topology, geometry and physical information, and can directly be resolved 3D visualization image. GC checks the physiological and virtual environmental status of every organ, calls Branching Controller (BC) to control the growth and development of the meristems. BC analysis the branching rules and exports the branching instructions. The Two Dimensional Hierarchical Automata (2DHA) has been used to say and store the branching rules. The branching rules can be extracted by manually observation, or be extracted automatically by image processing and statistical analysis.Finally, the model has been implemented on the platform based on the virtual plant simulation software'PlantLab', which was developed by our laboratory with indigenous intellectual property rights. Several simulations of cayennes'growth and 3D morphological changes under normal, one-way light, lower temperature, water deficit environment and pruning have been conducted using this simulation platform. The comparison between simulated and real growth data shows that the model can effectively and vividly simulate the growth and development of virtual plant and its deformation.