Research On Height Reconstruction Method Of Forests And Buildings Based On Polarimetric SAR Interferometry

The stereoscopic visual effect of the ground target can provide a great help to urban planning and forest monitoring.Synthetic aperture radar(Synthetic Aperture Radar,SAR)interferometric imaging technology can obtain ground height information through remote sensing.Therefore,target height reconstruction is a research hotspot in the field of remote sensing.At present,for buildings,the height reconstruction using single SAR image is limited by the image quality and building trend and the reconstruction using Tomo SAR technology is limited by the cost and the amount of data while the accuracy is high.Besides,the height reconstruction using InSAR technology is affected by the layover phenomenon and height jump.Therefore,considering the compromise,it is an important direction to use PolInSAR technology for building height reconstruction.In addition,the research on forest height inversion is mainly based on the study of coherent scattering models.RVo G model is a classical scattering model,but its assumption is too simple,which has a certain gap with the real scattering process in forest areas.Therefore,the establishment of a complete scattering model and the corresponding inversion algorithm is an important research direction at present.This paper focuses on the height reconstruction of buildings and forests in PolInSAR images of large scenes.For the height reconstruction of buildings,firstly,the single-polarization InSAR height reconstruction technology was studied and the effect of the layover phenomenon on the single-polarization InSAR height reconstruction was analyzed and the experimental verification is carried out using experimental data.Then for the above problem,the PolInSAR technology is used to reconstruct the building height.For this part,the phase separation method based on Pauli basis and PD coherence optimization method were studied and experiments were conducted using simulation data and real ESAR data.At the same time,for the height inversion of forests,this paper focuses on the method based on coherent scattering model.Firstly,the three-stage inversion algorithm based on RVo G model is studied,and then the scattering process in forest area is analyzed in detail.And based on the above analysis,STLS model is derived from the perspective of vertical structure function and the coherence distribution of STLS model is analyzed in detail.Finally,the threestage inversion algorithm is improved based on STLS model,and experiments are carried out with simulation data and real SIR-C data.The experimental results show that for building targets,the phase separation method based on Pauli basis and the PD coherence optimization method can effectively classify the phases of different scattering centers and solve the problem of overlay in building area;for forest targets,compared to RVoG model,the results of STLS scattering model and the improved three-stage algorithm on terrain with different slopes are more stable and accurate,proving that STLS model is closer to the true scattering process in forest areas and eliminates the influence of terrain slopes.