摘要
针对目前大多数点云滤波方法的不足,提出基于随机抽样一致(random sample consensus, RANSAC)多模型拟合的隧道点云滤波算法。与单模型拟合滤波方法不同,该算法在每个隧道横断面上使用多个圆模型进行分段拟合,多个模型具有不同的模型参数,将远离所拟合的各圆模型的点作为噪声点进行剔除。实验结果表明,相较于单模型拟合滤波方法,所提出的多模型法具有更低的一类误差和相近的二类误差,对隧道三维激光点云的滤波效果更好。
In view of the shortcomings of the most point cloud filter methods, a tunnel point cloud filtering algorithm based on random sample consensus(RANSAC) multi-model fitting is proposed. Unlike single model fitting filtering method, this algorithm uses multiple circular models in each tunnel cross section to fit segmental approximation. Multiple models with different model parameters can remove the points away from the fitted circle models as noise points. The experimental results show that, compared with the single model fitting filtering algorithm, the proposed multi-mo-del algorithm has lower type I error, similar type II error and better filtering effects on tunnel 3 D laser point cloud.
作者
陈珂
刘华
闫利
乐林株
CHEN Ke;LIU Hua;YAN Li;YUE Linzhu(School of Geodesy and Geomatics,Wuhan University,Wuhan 430079,China;Institute of Aerospace,School of Surveying and Mapping,Wuhan University,Wuhan 430072,China)
出处
《测绘地理信息》
CSCD
2021年第3期60-63,共4页
Journal of Geomatics
基金
国家重点研发计划(2017YFC0803801)。
