论文部分内容阅读
针对光学遥感图像记录的地表物体反射太阳辐射能量的高低不能较好地反映地表林木冠层高度这一问题,该文利用小波去噪对激光雷达数据进行去噪处理,分别计算小波去噪和高斯去噪后信号的信噪比(SNR)、均方根误差(RMSE);评价去噪效果,发现利用小波去噪得到的信号SNR均大于13db、RMSE均小于0.007 0,去噪效果优于高斯去噪;根据树木垂直结构以及回波原理利用回波分析法,确定回波开始以及地面回波位置来反演树木冠顶高度;将反演得到的树高进行对比分析,并利用实测数据对反演结果进行检验,发现本文反演结果与实测数据具有较好的相关性,反演结果稳定、精度较高。
In view of the fact that the ground surface reflected by the optical remote sensing image can not reflect the height of the forest canopy, this paper uses the wavelet denoising to denoise the lidar data and calculates the wavelet denoising and Gaussian The signal-to-noise ratio (SNR) and root mean square error (RMSE) of the denoised signal are evaluated. The denoising effect is evaluated. The SNRs of the signals obtained by wavelet denoising are both greater than 13db and RMSE less than 0.007 0. Denoising; According to the vertical structure of the trees and the principle of the echo, using echo analysis method to determine the echoes start and the position of the ground echo to invert the height of the tree crown; comparative analysis of the tree height obtained by inversion, and using the measured data pairs The inversion results are tested and found that the inversion results in this paper have good correlation with the measured data, the inversion results are stable and the accuracy is high.