胡双贵^1,2,,
汤井田^1,2,
任政勇^1,2,
周聪^1,2,,,
肖晓^1,2
1. 中南大学地球科学与信息物理学院, 长沙 410083
2. 中南大学有色金属成矿预测与地质环境监测教育部重点实验室, 长沙 410083

基金项目: 国家高技术研究发展计划（2014AA06A602），国家自然科学基金（41574120），中南大学中央高校基本科研业务费专项资金（2017zzts177）和中南大学创新驱动计划（2016CX005）联合资助


详细信息

作者简介: 胡双贵, 男, 1988年生, 博士研究生, 从事重磁数据处理与解释研究.E-mail:hushuanggui808@csu.edu.cn

通讯作者: 周聪, 男, 1987年生, 博士后, 从事电磁法数据处理与解释研究.E-mail:zhoucong_522@163.com

中图分类号: P631

收稿日期:2017-10-16
修回日期:2018-05-29
上线日期:2019-01-05

Normalized facet edge detection and enhancement in potential field sources with the scale-space technique

HU ShuangGui^1,2,,
TANG JingTian^1,2,
REN ZhengYong^1,2,
ZHOU Cong^1,2,,,
XIAO Xiao^1,2
1. School of Geosciences and Info-Physics, Central South University, Changsha 410083, China
2. Key Laboratory of Metallogenic Prediction of Nonferrous Metals and Geological Environment Monitoring, Ministry of Education(Central South University), Changsha 410083, China


More Information

Corresponding author: ZHOU Cong,E-mail:zhoucong_522@163.com

MSC: P631

--> Received Date: 16 October 2017
Revised Date: 29 May 2018
Available Online: 05 January 2019


摘要
摘要:边界识别是位场数据处理解释中的重要环节，传统边界识别方法通常不能均衡深、浅部地质体边界.基于尺度空间技术和归一化的Facet模型检测算子，本文开发了一种带通空间滤波和边缘检测相结合的边界识别方法，有效地提高位场数据边界识别的精度和可靠性.为了验证本文算法的有效性和稳定性，分析了不同尺度空间函数和检测算子对算法的影响，并且对比了传统边界识别方法的效果.理论模拟和实际数据分析表明，利用位场垂向二阶导数进行的基于尺度空间技术的归一化Facet模型边界识别方法不仅算法的稳定性强，而且可以避免高阶导数对噪声干扰放大作用，同时均衡深部和浅部地质体边界，从而可以更精确地识别地质体的形态.
关键词: 边界识别/
尺度空间函数/
归一化Facet模型检测算子/
位场

Abstract:Edge detection and enhancement is an important technique in interpretation of potential field data. Many existing edge detection and enhancement methods are usually unable to enhance the signals of the shallow and deep geological sources. In this study, we present an edge detection method to enhance the weak and noisy signals produced by shallow and deep mass anomalies. This new technique is based on a combination of the normalized cubic facet model and the scale-space technique which uses a band-pass filter taking the differences between two Poisson scale-space representations of the potential-field data. Compared to the existing edge detection algorithms, this new algorithm can more accurately recover the edges and shapes of the underground anomalous bodies with numerical stability. It can also effectively suppress noise as the high-order derivatives are used due to the Poisson scale-space representation which is equivalent to performing an upward continuation of data. Tests on synthetic gravity data and real gravity anomaly data collected from the Witwatersrand Basin in South African and the real aeromagnetic anomaly data from Tongling of China are performed to verity this new edge detection and enhancement algorithm. Results show that different applications of newly developed filter are feasible to interpret other geologic environments with similar challenges of enhancing the field responses of deep and shallow sources. Comparisons with the HGA, tilt angle, NSTD and theta map indicate that this new filter yields superior results and produces better interpretive images than the other filters.
Key words:Edge detection/
The scale-space technique/
Normalized cubic facet model/
Potential field



PDF全文下载地址:

http://www.geophy.cn/data/article/export-pdf?id=dqwlxb_14843