Title:
機械学習方法、及び地表変動判定方法
Document Type and Number:
Japanese Patent JP7102106
Kind Code:
B2
Abstract:
To automatize determination of the presence or absence of a ground deformation based on the interference SAR image.SOLUTION: Component image generation means 221 generates component images of n kinds (n≥2) defining p(θ)(1≤k≤n) a pixel value from a phase difference θ of each pixel of the interference SAR image. Learning data generation means 222 generates learning data combining the component images of n kinds and a reading result acquired from a user. Model generation means 223 generates fluctuate determination mode by the machine learning using the learning data. Each function p(θ) defines a section D of a width 2πrad corresponding to the range of the phase difference as α-π α+π and continuation in D and P(α-π)=P(α+π). When the coordinates defined by a set of p(θ) with respect to the whole k is represented by P(θ) P(θ)≠P(θ) for Aθ, θεD is θ≠θother than both ends of D.SELECTED DRAWING: Figure 1
Inventors:
Keisuke Nemoto
Tomoyuki Imaizumi
Shuhei Hikosaka
Tomoyuki Imaizumi
Shuhei Hikosaka
Application Number:
JP2017097563A
Publication Date:
July 19, 2022
Filing Date:
May 16, 2017
Export Citation:
Assignee:
Pasco Co., Ltd.
International Classes:
G01S13/90
Domestic Patent References:
| JP2010175381A | ||||
| JP2003500658A | ||||
| JP2009289111A | ||||
| JP20063302A |
Foreign References:
| WO2008016034A1 | ||||
| US20160019458 | ||||
| US5835055 |
Other References:
津田 覚之,"合成開口レーダ情報のニューラルネットワークによる構造化の提案",電子情報通信学会技術研究報告,Vol.115,No.200,日本,一般社団法人電子情報通信学会,2015年08月20日,pp.31-36
Attorney, Agent or Firm:
Haruka International Patent Office