基于三维激光点云的树木胸径自动提取方法  

Automatic extraction method for tree diameter at breast height measurement based on 3D point cloud

作　　者：王祺[1] 胡洪[1] 吴艳兰[1] 许邦鑫[2] 王浩 薛兴盛[1] 

机构地区：[1]安徽大学资源与环境工程学院,合肥230601 [2]中国能源建设集团安徽省电力设计院有限公司,合肥230601 [3]滨州市技师学院,滨州256500

出　　处：《安徽农业大学学报》2017年第2期283-288,共6页Journal of Anhui Agricultural University

基　　金：国土环境与灾害监测国家测绘局重点实验室开放基金资助项目(LEDM2014B02);安徽省智慧城市与地理国情监测重点实验室开放基金(2016-K-01Y);安徽大学研究生学术创新研究扶持与强化项目(yqh100252)共同资助

摘　　要：胸径是评价林木生长状况的重要参数之一。针对接触式人工测量自动化程度低和基于点云的现有算法提取树木胸径精度不高的问题,提出一种基于点云数据的自动准确获取树木胸径的新方法。该方法以树木点云数据为基础,运用蚁群算法和B样条曲线拟合技术,实现树木胸径的自动准确提取。对实验区树木测量计算,结果表明,利用该方法提取树木胸径的均方根误差为±0.19 cm,平均绝对误差为0.15 cm,相对于基于点云的传统算法提取精度分别提高了50%和60.7%。该方法基于高精度点云数据,实现了树木胸径的无损自动提取,在精准林业领域具有推广价值。The diameter at breast height （DBH） is one of the important parameters to evaluate the growth of forest trees. In view of the low degree of automation of contact-type artificial field measurement, and the low accuracy of existing algorithms based on point cloud, a new method using 3D laser scanning technology to automatically measure tree DBH was proposed. Based on tree point cloud data, using ant colony algorithm and spline curve fitting technique, the DBH was measured by programming automatically. Sample trees in the experimental area were measured and calculated, and the results showed that the root-mean-square error of the algorithm for the extraction of tree diameter was
0.19 cm, and the average-absolute error was 0.15 cm, which were improved by 50% and 60.7%, respectively compared with the traditional algorithm based on point cloud. Based on high-precision point cloud data, this method can achieve the automatic extraction of tree diameter at breast height and showed potential to be popularized in the field of precision forestry.

关 键 词：胸径 三维激光扫描 点云数据 蚁群算法 曲线拟合 

分 类 号：S758.7[农业科学—森林经理学]