Shale Microstructure Prediction with Nonsmooth Sparse Regularizing X-Ray CT Imaging. (25th October 2020)

• Record Type: Journal Article
• Authors: Ma, Qinghua; Xu, Zuoliang; Wang, Yanfei; Wang, Yu; Wang, Lihua; Wang, Jianqiang
• Is Part Of: Numerical functional analysis and optimization. Volume 41:Number 14(2020)
• Page Start: 1717
• Page End: 1727
• Publication Date: 2020-10-25
• Subjects: Data-based image analysis -- micro-CT imaging -- sparse regularization; 65J22 -- 86-08 -- 86A22; Functional analysis -- Periodicals; Numerical analysis -- Periodicals; Mathematical optimization -- Periodicals; Numerical Analysis, Computer-Assisted; 515.705
• DOI: 10.1080/01630563.2020.1813168
• Languages: English
• ISSNs: 0163-0563

Abstract

Abstract: Microscopic analysis of mineral distributions in shale receives much attention in recent years. Traditional methods such as optical and scanning electron microscopy are common tools for providing valuable information of microstructures; however, they belong to surface observations, which damage the shale sample and only generate a 2D image. We consider using the X-ray computerized tomography (CT) based on synchrotron radiation (SR) to generate 3D image reconstructions. In addition, with the CT imaging data, we are required to analyze compositional microstructures of the shale. Hence, we are faced with two kinds of inverse problems: model-based and data-based inverse problems. For the X-ray tomographic imaging, the filtered backprojection is the well-adopted conventional algorithm, but it is sensitive to noise and is inflexible. So, we will develop the nonsmooth sparse regularizing technique for this issue. For the later one, we need to consider the CT-data constrained minimization technique. Solving methodologies will be addressed. Experimental tests for both inverse problems will be given.