【作者】 王涵； 刘云鹤； 殷长春； 王浩蔓； 李俊； 凌宏； 柯智渊； 蔡晶； 陈健；

【Author】 Wang Han;Liu Yunhe;Yin Changchun;Wang Haoman;Li Jun;Ling Hong;Ke Zhiyuan;CaiI Jing;Chen Jian;College of Geoexploration Science and Technology,Jilin University;College of Instrument Science and Electrical Engineering,Jilin University;

【机构】 吉林大学地球探测科学与技术学院； 吉林大学仪器科学与电气工程学院；

【摘要】 传统瞬变电磁工作流程通常是在野外采集数据后回到实验室进行数据处理和解释,无法进行现场实时成像,因此施工人员无法实时了解重点区域异常特征。本文基于深度学习理论,自主研发了一套新的瞬变电磁实时成像系统。数值模拟结果表明,该系统具有高效快速的特点,可以划分出重点异常区域的位置。更多还原

【Abstract】 The traditional transient EM method usually collects data in the field and then perform data processing and interpretation in the laboratory.The real-time imaging cannot be performed on site,which makes it impossible for construction personnel to locate key abnormal areas in the first time and improve the accuracy of the detection.In this paper we develop a new real-time TEM imaging system based on the deep learning.Numerical simulation results show that the system has the characteristics of high efficiency and can identify the locations of key abnormal areas.更多还原