2021年1月に山口特任研究員,水谷准教授らの論文”Mapping Subsurface Utility Pipes by 3D Convolutional Neural Network and Kirchhoff Migration using GPR Images”がIEEE Transactions on Geoscience and Remote Sensing (IF: 5.9)に掲載されました.本研究では,深層学習と逆解析を組み合わせ,地中レーダーデータから埋設管位置を自動推定して三次元空間上にマッピングする,世界初の画期的な手法を提案しました.
2021.3. Dr. Yamaguchi, Prof. Mizutani et al. published “Mapping Subsurface Utility Pipes by 3D Convolutional Neural Network and Kirchhoff Migration using GPR Images” in IEEE Transactions on Geoscience and Remote Sensing(IF=5.9). 3D positions of subsurface pipes are mapped by deep learning and inverse analysis, in which the first and impactful algorithm are proposed and demonstrated. Early access: https://ieeexplore.ieee.org/abstract/document/9244614