Data Organization And Scheduling Optimization Of Virtual Forest Scene For Multi-Level Linkage

In recent years, the development of virtual reality and simulation technology brings new energy for the further research of forest ecosystems. Virtual reality and simulation technology has become a hotspot across ecology, agriculture, forestry, computer science and other disciplines. Virtual simulation technology is used to describe the plant physiological structure and growth process, simulate the dynamic growth, renewal and death of forest. Since the physiological structure and growth process of plant is very complex, lead to the growth model of forest and scene rending process is very complex. How to solve the growth model of forest and render the forest scene quickly is still a difficult problem in virtual simulation.At present, researchers mainly studied plant individual growth model, population growth model, and community growth model; they focused on the growth process of plants in a single spaces-times granularity. However, how to create forest scene data storage structure for the simulations of different granularity, how to create link between different scene data with different granularity, how to schedule scene data when converting scene is still relatively rare.In view of above problems, this paper supports the forest scene multi level interaction studies on the correlation and mapping between different levels, transform the information between different granularity forest scenes. Main research contents of rhis paper including:(1) In this paper, we study the conversion between the differcat granularity scene, analysis the data requirement when forest scene change from large-scale to small-scale (coarse-grained to fine-grained, coarse-grained to medium-grained and medium-grained to fine-grained) and small-scale to large-scale (fine-grained to coarse-grained, fine-grained to medium-grained and coarse grained to fine-gmaied) According to the conversion between di(?)rent granularity scene we build the data structure which supporting multi-level linkage in the external memony.(2)We analysis the solving process of different granularity growth model and the visualization process of results, and organize the scene data in memory. We use improved memory-pool storage structure to organize the scene data in memory to reduce calculation time of the forest growth model.(3)When calculating the growth model of forest and visualizing the growth results in large-scale forest, we use data scheduling method base on quad-tree. The method can reduce the scheduling lime of growth model of forest. Besides, to ensure the smooth visual scene, we use multi-thread technology to visualize the model results.(4)Aceordmg to the above data organization method and data scheduling method, we design and implement the large-scale forest simulation system. Beside, through the time comparison with other methods, verify the effectiveness of the method in this paper.