改进的线性预测滤波算法在机载LiDAR点云中提取DEM的应用

Application of the Improved Linear Prediction Filtering Algorithm in Extracting DEM from Airborne LiDAR Data

在机载Li DAR点云数据处理中,如何从原始点云中提取地面点进而生成DEM(Digital Elevation Model)成为当下机载Li DAR点云数据处理的重要研究方向。针对传统线性预测算法存在的不足,对传统线性预测滤波算法在噪声点剔除、数据格网化分块、迭代终止条件判断、重要滤波参数设置等方面进行了改进,并利用国际摄影测量与遥感协会提供的实验数据,对4块实验区进行改进后的算法验证,结果表明,改进后的滤波算法提高了滤波的总精度和地面点的精度。

Extracting ground points from the original point cloud has become an important research direction in current airborne LiDAR data process. This paper improves the traditional linear prediction algorithm in noise elimination, the divided blocks in the data grid, filtering parameter setting and iteration termination conditions. With the experimental data provided by ISPRS, the improved algorithm is verified and applied in four experimental areas. And the results show that the total filtering accuracy and the accuracy of the ground point are both improved.