摘要
针对经典滤波算法易受参数、阈值选取影响以及传统偏度平衡方法滤波结果不理想等问题,提出一种联立偏度与峰度变化曲线的机载LiDAR点云二次滤波方法。该方法不受参数与阈值选取影响,在联立初始LiDAR点云二者基础上,通过在其变化曲线上寻找最优偏度平衡点完成一次滤波;然后对初次滤波后获取地面点进行多项式曲面拟合,根据拟合后的高差统计值进行偏度平衡二次滤波。实验结果表明:该方法能保证变化很小的前提下,减少和总误差,滤波效果更好。
In order to solve the problems which the classic filtering algorithms are easily affected by the selection of parameters and threshold values and the results of the traditional skewness balance filtering method are not very good, so a filtering method of airborne LiDAR point cloud based on the curve of skewness and kurtosis change was put forward in this paper. This method is based on the skewness and kurtosis of original LiDAR point cloud,through the changing curve to find the optimal point of skewness balance. Then the secondary filter is based on the skewness balance of elevation difference values, which get from polynomial surface fitting. Finally, the experimental results showed that this method could reduce the class errors and the total errors under the premise that the class error change was very small, so the filtering results are better.
出处
《测绘科学技术学报》
CSCD
北大核心
2016年第1期48-52,共5页
Journal of Geomatics Science and Technology
基金
"十二五"农村领域国家科技计划课题(2012BAJ22B00)
江苏高校优势学科建设工程项目(SZBF2011-6-B35)
