透過您的圖書館登入
IP:18.223.205.84
  • 期刊
  • OpenAccess

Ground-Based Scatterometer Measurements and Inversion of Surface Parameters by Using Neural Networks

並列摘要


In this study, a multi-band FM-CW ground-based scatterometer is used to measure the backscattering coefficient of bare soil surface. Combined with the neural network (NN) and the advanced integrated equation model (AIEM), it inverses all the soil parameters, including dielectric constant, root mean square (RMS) height and correlation length. According to different data sources, the NN can be constituted as different input-output mapping mode. The NN is trained by using theoretical data that are simulated by the AIEM. The measured data are an input to the NN to inverse the soil parameters. The inversion results have a better consistency with the sampling parameters of soil. More redundancy scattering data can improve accuracy and stability of the inversion. In addition, the scattering measurement experiment is an effective mean of studying scattering model and the inversion algorithm of microwave remote sensing.

延伸閱讀