王宁^1,,
殷长春^1,,,
高玲琦¹,
苏扬¹,
刘云鹤¹,
熊彬²
1. 吉林大学地球探测科学与技术学院, 长春 130026
2. 桂林理工大学地球科学学院, 桂林 541006

基金项目: 国家自然科学基金项目(41774125, 42030806, 41530320, 41904104, 42074120)及北京市科技计划(Z181100005718001)联合资助


详细信息

作者简介: 王宁, 女, 1997年生, 硕士研究生, 主要从事航空电磁数据处理及正反演研究.E-mail:wangning1509@163.com

通讯作者: 殷长春, 男, 1965年生, 教授, 国家特聘专家, 主要从事电磁勘探理论, 特别是航空和海洋电磁方面的研究.E-mail:yinchangchun@jlu.edu.cn

中图分类号: P631

收稿日期:2019-10-04
修回日期:2020-09-28
上线日期:2020-12-05

Airborne EM denoising based on curvelet transform

WANG Ning^1,,
YIN ChangChun^1,,,
GAO LingQi¹,
SU Yang¹,
LIU YunHe¹,
XIONG Bin²
1. College of Geo-Exploration Science and Technology, Jilin University, Changchun 130026, China
2. College of Earth Sciences, Guilin University of Technology, Guilin 541006, China


More Information

Corresponding author: YIN ChangChun,E-mail:yinchangchun@jlu.edu.cn

MSC: P631

--> Received Date: 04 October 2019
Revised Date: 28 September 2020
Available Online: 05 December 2020


摘要
摘要:航空电磁法作为一种地形复杂地区资源探测的有效方法，近年来得到了广泛的应用.然而，由于系统所处的动态环境，噪声干扰严重.为了改善航空电磁数据质量，提高地下电性反演的准确性，需要研发相关去噪技术.传统航电去噪大多针对特定噪声或单一测线上的信号进行处理，难以兼顾相邻测线之间观测信号的相关性.本文采用曲波变换进行二维航空电磁数据去噪.由于曲波变换具有多尺度和多方向性特征，可以在对噪声精细分析的基础上进行有效去除，同时还保证了整个测区内信号的相关性.进而，我们提出Sigmoid阈值函数对传统阈值函数进行改进，以进一步改善去噪效果.为了验证曲波变换方法对航空电磁数据去噪的有效性，将曲波变换和传统去噪方法分别应用于理论模型和实测数据进行对比.试验证明本文曲波变换用于航空电磁数据去噪具有明显的优越性.
关键词: 航空电磁法/
曲波变换/
去噪/
多尺度分析

Abstract:The airborne electromagnetic (AEM) method has become an effective tool for exploration in areas with complex topography and is widely applied in the world. However, most current AEM systems are mounted or towed in a dynamic environment, resulting in big noise interference. To improve the quality of AEM data and interpretation, the denoising technique needs to be developed. Traditional denoising methods work mostly on special noise or single survey lines, without taking into account the correlation of signal at neighboring survey lines. In this work, we develop a denoising technique based on the curvelet transform for AEM data. As curvelet transform has the characteristics of multiple-scale and multiple-direction, it can remove the noise based on detailed analysis to the signal, while at the same time it considers the signal correlation between neighboring survey lines. Further, we improve the quality of denoising results by introducing the Sigmoid threshold function. We test our method by denoising both synthetic and survey data. The numerical results show that the curvelet-based method has obvious advantage in denoising AEM data.
Key words:Airborne EM/
Curvelet transform/
Denoising/
Multi-scale analysis



PDF全文下载地址:

http://www.geophy.cn/data/article/export-pdf?id=5fed39e2bfedb50717e0ce5f