A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
Use of machine learning based technique to X-ray microtomographic images of concrete for phase segmentation at meso-scale
https://orcid.org/0000-0002-0468-6543
Highlights Segmentation of voids, aggregates and mortar using X-ray microtomographic image. Gray value thresholding technique used for void segmentation. Logistic regression method used for mortar and aggregate segmentation. Proposed method shows accuracy in aggregate detection.
Abstract The paper discusses the technical limitation of the gray value thresholding technique to detect the voids, aggregate and mortar phases. A two-stage image processing methodology is proposed for the segmentation of the three phases of concrete using the X-ray microtomographic images. In the first stage, the gray value thresholding technique is used to detect the voids. A machine learning based technique is proposed in the second stage for the segmentation of aggregate and mortar. The training data is used to model a planar decision boundary using the logistic regression method. For this, the radial distance from the centre of the image, gray value, and gray value of the filtered embossed image features are considered. The accuracy of the model to quantify the voids is validated with the commercial software. The machine learning model based on logistic regression method exhibits very good accuracy () in detecting the aggregate.
Use of machine learning based technique to X-ray microtomographic images of concrete for phase segmentation at meso-scale
https://orcid.org/0000-0002-0468-6543
Highlights Segmentation of voids, aggregates and mortar using X-ray microtomographic image. Gray value thresholding technique used for void segmentation. Logistic regression method used for mortar and aggregate segmentation. Proposed method shows accuracy in aggregate detection.
Abstract The paper discusses the technical limitation of the gray value thresholding technique to detect the voids, aggregate and mortar phases. A two-stage image processing methodology is proposed for the segmentation of the three phases of concrete using the X-ray microtomographic images. In the first stage, the gray value thresholding technique is used to detect the voids. A machine learning based technique is proposed in the second stage for the segmentation of aggregate and mortar. The training data is used to model a planar decision boundary using the logistic regression method. For this, the radial distance from the centre of the image, gray value, and gray value of the filtered embossed image features are considered. The accuracy of the model to quantify the voids is validated with the commercial software. The machine learning model based on logistic regression method exhibits very good accuracy () in detecting the aggregate.
Use of machine learning based technique to X-ray microtomographic images of concrete for phase segmentation at meso-scale
https://orcid.org/0000-0002-0468-6543
Highlights Segmentation of voids, aggregates and mortar using X-ray microtomographic image. Gray value thresholding technique used for void segmentation. Logistic regression method used for mortar and aggregate segmentation. Proposed method shows accuracy in aggregate detection.
Abstract The paper discusses the technical limitation of the gray value thresholding technique to detect the voids, aggregate and mortar phases. A two-stage image processing methodology is proposed for the segmentation of the three phases of concrete using the X-ray microtomographic images. In the first stage, the gray value thresholding technique is used to detect the voids. A machine learning based technique is proposed in the second stage for the segmentation of aggregate and mortar. The training data is used to model a planar decision boundary using the logistic regression method. For this, the radial distance from the centre of the image, gray value, and gray value of the filtered embossed image features are considered. The accuracy of the model to quantify the voids is validated with the commercial software. The machine learning model based on logistic regression method exhibits very good accuracy () in detecting the aggregate.
Use of machine learning based technique to X-ray microtomographic images of concrete for phase segmentation at meso-scale
Saha, Sanat Kumar (author) / Pradhan, Subhasis (author) / Barai, Sudhirkumar V. (author)
2020-03-12
Article (Journal)
Electronic Resource
English
Analysis of X-ray microtomographic images to measure work of fracture in concrete
| Tema Archive | 1997
Optimal removal of topological artefacts in microtomographic images of porous materials
| British Library Online Contents | 2011