基于激光点云的含笑树叶重建与形变方法

针对室外含笑树叶存在抖动及树叶间存在遮挡，导致含笑树叶点云数据不易重建与形变的问题。提出了一种含笑树叶重建与形变的方法，首先根据噪声特点和扫描线特性，对点云数据去噪；然后根据每条扫描线边缘点拟合出树叶的边缘，采用双三次广义张量积Bezier曲面拟合叶面，并结合三角剖分算法实现叶面的重建；最后采用基于非线性的有限元形变方法，模拟出真实的含笑树叶形变。实验结果表明，该算法简单高效。

Banana Shrub Tree Leave Reconstruction and Deformation Based on Laser Point Cloud

Due to banana shrub tree leave quivering and blocking between tree leaves, it is not easy to realize reconstruction and deformation of banana shrub tree leaves based on laser point cloud. A method for reconstruction and deformation of banana shrub tree leaves is put forward. First, point cloud data are denoised based on the characteristics of noise and scan lines;then, conduct fitting of tree leaves according to the edge points of each scan line, with bicubic generalized tensor product Bezier surface for leave surface fitting, with triangulation algorithm used to realize reconstruction of leaves;finally, a method based on non-linear finite element deformation is adopted to simulate the real deformation of banana shrub trees. The experimental result shows that this algorithm is simple and efficient.