多基线干涉层析SAR提取森林树高方法研究

Approach for Forest Height Extraction Using Multi-baseline Interferometric Tomographic SAR

应用瑞典Raminstorp研究区Bio SAR 2007 P-波段多基线In SAR数据,研究了基于多基线干涉SAR数据的层析方法,成功提取了可代表森林垂直结构信息的雷达后向散射功率垂直分布信息,并基于该信息提取了树高。应用地面实测样地对树高提取精度进行了检验,结果表明:HH极化树高提取精度最高(R2为0.65,RMSE为2.35 m,相关系数为0.80),HV其次,VV最差。

Applying P-band multibaseline InSAR data from BioSAR 2007 campaign in Raminstorp test site in Sweden by Multi-baseline Interferometric Tomographic SAR （MBInTomo SAR） technique to extract forest vertical structure information and estimate forest height from them,and validating the forest height with in situ data.The resuhs showed that the values of R^2 were 0.65,0.55,0.34,the values of RMSE were 2.35 m,3.27 m,5.13 m and the values of correlation coefficient （R） were 0.80,0.74,0.58 among the estimated forest heights using HH,HV and VV MBInTom SAR and Lidar H80,respectively.The conclusion is that the P-band multi-baseline InSAR data can beused for forest vertical structure information extraction and forest height estimation.And traditional forest scattering model and forest height estimation approach may not be suitable for P-band.