摘要
由于井下环境的复杂性,借助三维激光扫描仪获取的采空区边界三维空间信息点云数据中不可避免包含一些噪声点.为此,提出曲率-弦长比复合判据实现了对点云数据中高频噪声点的过滤处理,并运用随机滤波法去除点云数据中的低频随机噪声点,通过分段低次插值法实现空区模型曲线光顺处理.结果表明:过滤及光顺处理不仅有效去除了采空区点云数据中的噪声点,同时避免了采空区三维模型构建中自相交情况的出现,达到了采空区三维模型精确构建的目的.
Aiming at noise points in point cloud data detected by three-dimensional laser in cavity, curvature-chord ratio composite criterion was put forward for filtering and simplifying significant noise points. Besides, random filter algorithm was applied to reduce low frequency random noise points similar to the change of object. The piecewise low-order interpolation method was applied to fit cavity point cloud based on reduction and simplifying, thus the curve became smoother than before and after treatment. The application results show that the noise points in point cloud data detected by three-dimensional laser in cavity are effectively removed by filtering and smoothing treatment. Furthermore, the self-intersection in model is avoided. Therefore, cavity models generated by processed point cloud data are extremely identical to engineering practice. © 2016, Editorial Department of Journal of Northeastern University. All right reserved.
出处
《东北大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2016年第11期1635-1639,共5页
Journal of Northeastern University(Natural Science)
基金
国家自然科学基金资助项目(51274250)
