基于α-shape的三维激光点云计算树冠体积的研究

Research on the calculation of tree crown volume by 3D laser point cloud based onα-shape

针对树冠形状各异、结构复杂、轮廓难以提取和体积难以精确计算等问题,文中基于三维激光扫描得到的树冠点云,提出一种综合树冠点云切片算法、α-shape点云轮廓提取算法和散乱轮廓点云排序算法的树冠体积计算方法。首先对树冠点云按照一定的间隔进行切片,将树冠点云分成多个类似台体和锥体部分,然后利用α-shape算法对每一层树冠点云进行轮廓点的提取,再使用点云排序算法对α-shape算法提取到的轮廓点集进行逆时针排序,利用多边形面积的行列式计算公式分别计算每层点云的面积;并使用台体和锥体的体积计算公式计算每部分的体积,最终累加各部分体积值得到树冠的最终体积。通过对利用TrimbleTX8地面三维激光扫描仪得到的九棵不同形状和品种的树木树冠点云进行计算与分析,并使用传统的体元法和凸包法对本文提出的算法进行了验证。结果表明本计算方法能很好地提取树冠的外包轮廓,进而得到准确的树冠体积值,在准确性和精度上优于两种传统方法,可应用于林业资源的调查与分析。

Aiming at the problems of different crown shapes,complex structures,difficult contour extraction and volume calculation,this paper proposes a new calculation method of crown volume based on the crown point cloud obtained by three-dimensional laser scanning,which integrates the algorithm of crown point cloud slicing,algorithm ofα-shape point cloud contour extraction and of scattered contour point cloud sorting.The crown point cloud is firstly divided into several similar platforms and cones after slicing in accordance with a certain interval.Then the contour points of each layer of crown point cloud are extracted by usingα-shape algorithm.Afterwards,the point cloud sorting algorithm is used to sort the contour point group in counter-clockwise order extracted byα-shape algorithm,and the area of each layer of point cloud is calculated by using the determinant formula of polygon area.The volume of each part is calculated by using the volume calculation formula of platforms and cones,and the final volume of crown is obtained by accumulating the volume value of each part.By calculating and analyzing the crown point clouds of nine trees with different shapes and varieties obtained by trimblet TX8 laser scanner,the algorithm proposed in this paper is verified by using the traditional voxel method and convex hull method.The results show that the algorithm can extract the contour of the crown and obtain the accurate crown volume value,which is superior in accuracy compared with the two traditional methods.It can be applied to the investigation and analysis of forestry resources.