全波形LiDar数据分解方法的研究

Study of decomposition of full waveform airborne laser scanner data

研究了激光波形数据高斯函数分解的理论基础,提出了利用广义高斯模型函数拟合脉冲波形,并提取脉冲波形重要参数的方法。实验表明:利用广义高斯模型能较好地分解脉冲波形,该参数提取方法在分解振幅较高的脉冲波形时鲁棒性较好。该波形分解方法不但能增加点云的数量,而且能增加点云的层次感,同时能提取出激光点云的振幅、距离、脉冲宽度以及背向散射截面等重要的参数,这些特征对后续的地物分类研究提供了基础。

The basic theory of Gaussian function decomposition of laser waveform data was studied,and the method of fitting pulse waveform by using generalized Gaussian model function was proposed,and the way of extracting important parameters of pulse waveform also was put forward.The testing results show that the Gaussian function of the laser waveform data can better decomposed the pulse waveform,and the parameters extraction method had better robustness when the waveform with higher amplitude was decomposed,and not only increased the number of point cloud,but also increased the level of cloud sense,and the important parameters can be extracted of the laser point cloud,distance,pulse width and backscatter cross sections.The eigenvalues extracted provided a basis for subsequent classification of surface features.