摘要
密集匹配点云作为一种新型数据,被广泛应用于各种地形图测绘生产,针对现阶段密集匹配点云滤波技术尚不成熟的问题,本文利用一种改进的不规则三角网加密滤波算法,通过建筑物立面选取地面种子点,结合高程排序构建不规则三角网进行地面点云滤波,并与传统不规则三角网加密滤波算法进行对比,结果表明:改进的算法在地面滤波过程中,选取的地面种子点更加准确。对于不同的地形,此算法具有一定的适用性,在地形平坦区域总误差为9.3%,地形复杂区域总误差为12.9%,匹配精度提升了20%~30%,取得了较好的滤波效果。
As a new type of data,the dense matching point cloud is widely used in various topographic map mapping production.For the problem that the filtering technology of dense matching point cloud is still immature at this stage,this paper uses an improved irregular triangular network encryption filtering algorithm to select ground seed points by building elevation and construct irregular triangular network for ground point cloud filtering by combining with elevation sorting,and compares with the traditional irregular triangular network encryption filtering algorithm The results show that the improved algorithm selects more accurate ground seed points in the ground filtering process.For different terrain,this algorithm has certain applicability,and the total error is 9.3%in the flat terrain area and 12.9%in the complex terrain area,and the matching accuracy is improved by 20%-30%,which achieves a better filtering effect.
作者
成枢
孙昂
程旭明
CHENG Shu;SUN Ang;CHENG Xuming(Shandong University of Science and Technology,College of Geodesy and Geomatics,Qingdao 266590,China)
出处
《城市勘测》
2023年第2期85-89,共5页
Urban Geotechnical Investigation & Surveying
基金
国家自然科学基金资助项目(42001414)
山东省自然科学基金资助项目(ZR2019BD033)。
关键词
密集匹配点云
点云滤波
不规则三角网
高程排序
dense matching point cloud
point cloud filtering
irregular triangle network
elevation ordering
