Normalized shaping regularization for robust separation of blended data

Zhao, Qiang; Du, Qizhen

doi:10.1190/segam2019-3216416.1

Cited by 2 publications

(2 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…e regularization method was first proposed by Tikhonov [22]. It has become an indispensable part of the inverse problem theory and has been widely used in geophysical problems [23][24][25][26]. e relationship between gravity field T(x, y) and its vertical derivative D m (x, y) (m represents the order) in the wavenumber domain is…”

Section: E Basic Principle Of Tikhonov Regularization Iterationmentioning

confidence: 99%

The Calculation of High-Order Vertical Derivative in Gravity Field by Tikhonov Regularization Iterative Method

Zhang

2021

Mathematical Problems in Engineering

View full text Add to dashboard Cite

In mathematics, statistics, and computer science, particularly in the fields of machine learning and inverse problems, regularization is a process of introducing additional information in order to solve an ill-posed problem or to prevent overfitting. The Tikhonov regularization method is widely used to solve complex problems in engineering. The vertical derivative of gravity can highlight the local anomalies and separate the horizontal superimposed abnormal bodies. The higher the order of the vertical derivative is, the stronger the resolution is. However, it is generally considered that the conversion of the high-order vertical derivative is unstable. In this paper, based on Tikhonov regularization for solving the high-order vertical derivatives of gravity field and combining with the iterative method for successive approximation, the Tikhonov regularization method for calculating the vertical high-order derivative in gravity field is proposed. The recurrence formula of Tikhonov regularization iterative method is obtained. Through the analysis of the filtering characteristics of this method, the high-order vertical derivative of gravity field calculated by this method is stable. Model tests and practical data processing also show that the method is of important theoretical significance and practical value.

show abstract

Section: E Basic Principle Of Tikhonov Regularization Iterationmentioning

confidence: 99%

The Calculation of High-Order Vertical Derivative in Gravity Field by Tikhonov Regularization Iterative Method

Zhang

2021

Mathematical Problems in Engineering

View full text Add to dashboard Cite

show abstract

“…Recently, an alternative method called shaping regularization has been proposed to enhance the quality of imaging [29][30][31]. It can effectively reduce energy leakage when abnormal noise exists.…”

Section: Imaging Algorithmmentioning

confidence: 99%

A 0.2-Terahertz Ceramic Relic Detection System Based on Iterative Threshold Filtering Imaging and Neural Network

Mao

Chen

et al. 2021

Electronics

View full text Add to dashboard Cite

Ceramic cultural relics have an important cultural value, and their detection technology plays a pivotal role in the protection of cultural relics. The traditional detection method of ceramic relics is based on X-ray technology, resulting in damage to the cultural relics. The terahertz-wave transmission is nondestructive; yet, the terahertz imaging has several technical problems, such as complex algorithms and unclear imaging. In this paper, we propose a terahertz-wave imaging system of a 0.2-terahertz operating frequency with a single input and single output and perform the transmission imaging experiment of the ceramic container. This design adopts the iterative threshold segmentation imaging method to obtain the high-quality transmission image and the neural network algorithm in the imaging recognition. Both the imaging and identification results effectively prove excellent performance, such as the high efficiency, high precision, and good reliability of the nondestructive testing system for the ceramic relics. The resolution can reach 2 mm, and the time for imaging identification can be < 10 s, which will provide support for the nondestructive testing of ceramic cultural relics.

show abstract

Normalized shaping regularization for robust separation of blended data

Cited by 2 publications

References 12 publications

The Calculation of High-Order Vertical Derivative in Gravity Field by Tikhonov Regularization Iterative Method

The Calculation of High-Order Vertical Derivative in Gravity Field by Tikhonov Regularization Iterative Method

A 0.2-Terahertz Ceramic Relic Detection System Based on Iterative Threshold Filtering Imaging and Neural Network

Contact Info

Product

Resources

About