论文部分内容阅读
在数字植物虚拟仿真场景中,植物三维模型之间容易出现交叉穿透现象,有必要进行快速有效的碰撞检测。选择以玉米叶片为研究对象,针对叶片形态提出了一种基于AABB-OBB混合包围盒的碰撞检测方法。首先通过AABB盒相交测试剔除部分未相交对象;然后根据叶片形态特点,对可能相交对象构造OBB盒二叉树。依据玉米叶片交叉常态,改进了分离轴测试顺序,提出了基于优先权值的节点搜索方式,并优化了特殊情况的图元测试过程。实验结果表明,该方法能够准确快速地检测到玉米叶片的交叉状况。
In the digital plant virtual simulation scenario, cross-penetration phenomenon easily occurs between plant three-dimensional models, and it is necessary to carry out rapid and effective collision detection. In this paper, maize leaves were selected as the research object. A collision detection method based on AABB-OBB hybrid bounding box was proposed for leaf morphology. At first, some un-intersecting objects are removed through the intersection test of AABB box. Then, according to the morphological characteristics of the leaves, an OBB box binary tree is constructed for possible intersecting objects. Based on the cross normality of maize leaves, the test sequence of the separation axis was improved, the search method of nodes based on priority values was proposed, and the process of primitives was optimized. Experimental results show that this method can accurately and quickly detect the cross of corn leaves.