采用车载LiDAR获取点云,根据分层投影算法结合地物特征等先验知识实现各单株行道树点云的提取,通过最小二乘法拟合圆,运用Matlab编程测量行道树胸径和株距.实验结果表明,该方法与人工测量两种测量方法测得的行道树胸径和相邻行道树的间距的均方根误差分别为0.008 7 m和0.010 3 m,相关系数分别为0.968 2和0.999 6.该方法精度高、测量范围大,可为检测行道树分布和生长状态提供优质数据.%Points cloud of street trees were obtained by vehicle LiDAR.The existing hierarchical projection algorithm was combined with heuristic knowledge of object characteristics to achieve the points cloud extraction of each tree.Fitting circle by least squares,the diameter at breast height(DBH) and row spacing of street trees were measured by Matlab programming.The experimental results showed that the root mean square(RMS) error of DBH and row spacing of LiDAR-based measurement and manual measurement were 0.008 7 m and 0.010 3 m,the correlation coefficients were 0.968 2 and 0.999 6.The proposed method had high precision and wide measuring range and provided quality data for the detection of distribution and growth state of street trees.
展开▼
