小麦群体生长状态实时绘制技术及实现

Real-time rendering of wheat population growth status and its realization

为了有效提高小麦群体可视化的真实感与实时性,该文提出一种基于多技术融合的小麦群体生长实时绘制的方法。首先,根据小麦群体形态建成规律,定量分析了小麦群体中个体间差异性。其次,在改进群体显示实时性方面,基于已有的小麦个体生长可视化,采用细节层次(LOD)模型,根据视点、视角和视线来动态指导空间剖分以及细节层次的选择;采用视域裁剪技术,通过减少被检测节点的数量和提高单个节点视域裁剪的处理速度来提高系统的渲染速度;基于小麦群体中个体器官形态的组成,结合树形多级显示列表技术,通过改进数据结构设计,提高场景的显示效率;在增强群体可视化真实感方面,基于已有的小麦形态模型输出,通过小麦群体中个体间差异性分析研究,采用外部引用和实例化技术,创建互异的个体实例样本,群体绘制时进行随机引用,提高了显示的真实感;最后,进行基于不同技术小麦群体绘制的效率与实例分析比较,结果表明,基于多技术融合场景绘制算法的小麦群体生长可视化,具有较好的实时性和真实感。研究结果为进一步建立小麦生长可视化系统奠定了基础。

To effectively improve realistic and efficiency of visualization of wheat population, a real-time rendering algorithm based on multi-technology integration was developed for visualization of wheat population growth. Firstly, based on the rule of morphological formation of wheat population, differences between individual architectures in population were analyzed. Secondly, in order to improve rendering efficiency, level of detail （LOD） model was adopted, space dividing and selection of level of detail were decided by sight position, sight angle and sight line. View frustum culling was used to reduce the number of detecting point and improve speed of sigle point culling. Tree-structure display list was used to improve data strcture design. Thirdly, external reference and instance technology was used to improve realistic of visualization by generate several diffferent samples which would be used when wheat population were rendering. Finally, a simulation experiments for comparison showed that the rendering algorithm based on multi-technology integration improved efficiency of real-time visualization of wheat population, reduced memory requirement, the techologies for visualization of wheat population growth had a good performance on dynamic and realistic characteristics. These results are helpful for development of a visualization wheat growth system.