Simulation Of Petal Deformation And Growth Process Based On 3D Scan Data

Digital plant simulation is the research hotspot of precision agricultural and forestry information technology.At present,the most researches on virtual plants use graphical modeling to simulate the growth and change of plants.The calculation is complicated and the operation is not flexible.With the development of 3D scanning and point cloud reconstruction technology,new methods have been provided for the visualization of complex plant morphology.In view of the complexity of plant morphology and structure,and in order to avoid this complicated modeling process,this paper studied the deformation and growth process simulation of plant petals based on the point cloud data model of 3D scanning.The three-dimensional scanner was used to scan intermittently to obtain model data and growth sequences of 12 flower plants such as Chinese rose and Clivia at 6 growing periods in a 15-day cycle.Based on the measured point cloud data and petal structure characteristics,a method of segmenting petal structure to process data was proposed.And the data fitting method using given function form in MATLAB was used to fit the curve of plant growth.On the Unity platform,the traditional free-form deformation algorithm that relies on interaction was transformed to a data-driven approach.Finally the traditional flower plant growth simulation and free-form deformation of the three-dimensional model was combined and improved.Aiming at plant growth simulation,based on the measured three-dimensional scan data,the paper proposed a method to simulate the dynamic growth of plants through the deformation algorithm.This method can not only retain the complex morphological characteristics of flower plants,but also make deformation control simple,and the fitted growth curve was closer to the measured data.Furthermore,the simulated petal morphology and growth were real and natural.In addition,our method of combining the measured data and the improved free-form deformation algorithm was compare with the physics-based flower opening simulation method,and was evaluated the algorithm precision by fitting regression analysis and measured the geometry error between the real measure data and the reconstructed data.Among them,from the experimental results of the morphology and 6 growing season data of 12 flower plants such as Chinese rose,it showed that the geometry deformation error can be controlled within 2 mm and the determination coefficients of the growth curve fitting were all greater than or equal to 0.79 during the whole period of plant growth.The results showed that the free-form deformation algorithm combined with real growth function can better simulate the growth and deformation of petal structures.It also provided a reference for simulating the growth and deformation of flower plants.