Segmentation Of Pores Research Articles

Interfacial characteristics of cement mortars containing aggregate derived from industrial slag waste

ABSTRACT In this study, porosity assessment of cement mortars containing Precious Slag (PS) Ball – an industrial waste produced by rapid cooling of slag generated from steel making process by Slag Atomizing Technology – is done by Back-Scattered Electron (BSE) image analysis. PS incorporated mortars were fabricated and subsequently evaluated for compressive strength. BSE images were analyzed by pore segmentation method. Binarization on the images using Otsu’s thresholding method was done followed by the porosity determination. As the interfacial attributes were the primary concern, the porosity was determined from the aggregate surface to the 100 μm width. A total of 24 sections were analyzed and the mean values were plotted. Although the PS Ball incorporated mortar exhibited 10% lower strength at 28-day age, it showed stronger and less porous aggregate- matrix interface between (from particle surface up to 35 μm width) as compared to that of normal mortar containing sand, which is justifiable by the chemical composition of PS Ball attributed to the potential chemical binding. Further, the porosity decreased as the distance from aggregate surface increased. While the use of PS Ball is environmentally friendly, its enhanced interfacial properties with some strength loss of resulting cement mortars encourages its use in sustainable construction materials.

Image-based porosity classification in Al-alloys by laser metal deposition using random forests

Additive manufacturing (AM) technologies enable complex, high-value components to be printed, with potential applications in the automotive, aerospace, and biomedical sectors. Porosity in AM processes for metals is a recurrent problem which can lead to adverse effects such as crack initiation and ultimately to parts’ early-life failure. There are several pore classifications described in the literature, which are focused on traditional manufacturing processes. The current lack of information makes it difficult to accurately identify and classify pores in AM-made parts. The present work describes a proposal based on image processing and machine learning, specifically random forests, to classify porosity automatically in metallographic images. The proposed method is divided into 3 stages. (1) Preprocessing stage: image denoising, smoothing, and unblurring to highlight the areas with pores. (2) Feature extraction stage: segmentation of pores and the morphological/geometrical features that describe the porosity. (3) Intelligent classifier stage: definition, training, testing, and validation of the random forest classifier. Our proposal has an accurate balance between the calculation of the feature importance as well as the number to use, the adequate number of trees to grow per forest, and the correct selection of the size of the database. The proposed method achieves an accuracy of 94.41% and out-of-bag error less than 5%. These results guarantee high precision in the porosity classification task. Our approach has the potential to be used in the porosity analysis of any metallic additively manufactured component.

An Automatic Method for Stomatal Pore Detection and Measurement in Microscope Images of Plant Leaf Based on a Convolutional Neural Network Model

Stomata are microscopic pores on the plant epidermis that regulate the water content and CO2 levels in leaves. Thus, they play an important role in plant growth and development. Currently, most of the common methods for the measurement of pore anatomy parameters involve manual measurement or semi-automatic analysis technology, which makes it difficult to achieve high-throughput and automated processing. This paper presents a method for the automatic segmentation and parameter calculation of stomatal pores in microscope images of plant leaves based on deep convolutional neural networks. The proposed method uses a type of convolutional neural network model (Mask R-CNN (region-based convolutional neural network)) to obtain the contour coordinates of the pore regions in microscope images of leaves. The anatomy parameters of pores are then obtained by ellipse fitting technology, and the quantitative analysis of pore parameters is implemented. Stomatal microscope image datasets for black poplar leaves were obtained using a large depth-of-field microscope observation system, the VHX-2000, from Keyence Corporation. The images used in the training, validation, and test sets were taken randomly from the datasets (562, 188, and 188 images, respectively). After 10-fold cross validation, the 188 test images were found to contain an average of 2278 pores (pore widths smaller than 0.34 μm (1.65 pixels) were considered to be closed stomata), and an average of 2201 pores were detected by our network with a detection accuracy of 96.6%, and the intersection of union (IoU) of the pores was 0.82. The segmentation results of 2201 stomatal pores of black poplar leaves showed that the average measurement accuracies of the (a) pore length, (b) pore width, (c) area, (d) eccentricity, and (e) degree of stomatal opening, with a ratio of width-to-maximum length of a stomatal pore, were (a) 94.66%, (b) 93.54%, (c) 90.73%, (d) 99.09%, and (e) 92.95%, respectively. The proposed stomatal pore detection and measurement method based on the Mask R-CNN can automatically measure the anatomy parameters of pores in plants, thus helping researchers to obtain accurate stomatal pore information for leaves in an efficient and simple way.

Combined mechanical and 3D-microstructural analysis of strain-hardening cement-based composites (SHCC) by in-situ X-ray microtomography

The paper presents the results of a series of combined mechanical and in-situ morphological investigations on highstrength strain-hardening cement-based composites (SHCC). Tension and compression experiments were performed in a CT scanner employing a dedicated mechanical testing rig. The in-situ microtomographic scans enabled correlating the measured specimen response with relevant microstructural features and fracture processes. The microstructural segmentation of SHCC was performed in the framework of Deep Learning and it targeted an accurate segmentation of pores, fibers and aggregates. Besides their accurate volumetric representation, these phases were quantified in terms of content, size and orientation. The fracture processes were monitored at different loading stages and Digital Volume Correlation (DVC) was employed to spatially map the strains and cracks in the specimens loaded in compression. The DVC analysis highlighted the effect of loading conditions, specimen geometry and material heterogeneity at the mesolevel on the strain distribution and fracture localization.

Dynamic capillary pressure analysis of tight sandstone based on digital rock model

In recent studies, dynamic capillary pressure has shown significant impacts on the flow behaviors in porous media under transient flow condition. However, the effect of dynamic capillary pressure effect on tight sandstone is still not very clear. Since lattice Boltzmann method (LBM) is a very promising and widely used method in analyzing flow behaviors, therefore, a two-phase D3Q27 LBM model is adopted in this paper to simulate the flow behaviors and analyze the dynamic capillary pressure effect in tight sandstone. Moreover, a new pore segmentation method for tight sandstone base on U-net deep learning model is implemented in this study to improve the pore boundary qualities of pore space, which is crucial for two-phase LBM simulation of tight sandstone. A total of 3800 3D sub-volume data sets extracted from computed tomography data of 19 tight sandstone samples are selected as ground truth data to train the network and segment the pore space afterward. The simulation results based on the segmented digital rock model, show that nonwetting phase fluid prefer the path with lower dynamic capillary pressure in the seepage process before breaking through the porous model. Furthermore, the increase of injection rate causes the saturation changes more quickly, injection rate also shows apparent positive correlation relationship with capillary pressure, which implies that dynamic capillary pressure effect also exists in tight sandstone, and LBM based two-phase flow simulation could be used to quantitatively analyze such effect in tight sandstone. Cited as : Cao, Y., Tang, M., Zhang, Q., Tang, J., Lu, S. Dynamic capillary pressure analysis of tight sandstone based on digital rock model. Capillarity, 2020, 3(2): 28-35, doi: 10.46690/capi.2020.02.02

A hybrid method to segment the pores and throats of micromodels

Micromodels made of silicon wafer are widely used in microfluidic system or oil displacement visual experiment. In this study, we develop a technique to partition the pores and throats of 2D porous micromodels, based on which the porous media structure could be described quantitatively. First, two series of medial axis are obtained based on the watershed algorithm and thinning algorithm, respectively. The thinning medial axis can be divided into four parts, i.e., node, endpoint, trunk, and branch, while the watershed medial axis does not have an endpoint. Then, the segmentation of pores and throats is initially carried out based on the watershed medial axis, and the interface of pore and throat is identified using the wavelet denoising method. Subsequently, the dead ends are further segmented, induced by the endpoints of thinning medial axis. Finally, we verify the method using the regular models and describe a random network micromodel quantitatively. The hybrid method not only accurately guarantees the medial axis located in the middle center but also keeps all the structure information of pore dead ends.

Stokes‐Brinkman Flow Simulation Based on 3‐D μ‐CT Images of Porous Rock Using Grayscale Pore Voxel Permeability

AbstractThe direct flow simulation using high‐resolution micro‐computed tomographic (μ‐CT) images of porous rock can be used to help understand the flow characteristics at the pore‐scale and to estimate fluid properties; however, segmentation of pore space in grayscale 3‐D μ‐CT images, a necessary step in this process, is challenging because of issues related to the image resolution and pore‐filling matrix in the gray‐level images. We present a novel process for determining the voxel porosity and permeability of the gray‐level regime and evaluating the bulk permeability using the Brinkman force lattice Boltzmann method. In this study, μ‐CT images of Berea sandstone are acquired with two spatial resolutions. After the pore size distribution curve is experimentally obtained, the “apparent pore” voxels and “gray pore” voxels are determined based on the designated gray‐level values in the gray‐level (CT number) histogram, the cumulative and fractional pore volumes, and the linear relationship between CT number and individual voxel porosity. The results show that the boundary between apparent and gray pores in terms of size determines the volumetric fraction and specific surface area. Additionally, the permeability computed by considering the gray pore regime based on the proposed method is more similar to the experimentally measured value than the results of other segmentation methods because the gray pore domain is fully incorporated into the flow domain while preserving the pore connectivity. The proposed sequence of pore segmentation presents a method for handling gray‐level pore space without compromising pore connectivity.

Influence of bonded mortar of recycled concrete aggregates on interfacial characteristics – Porosity assessment based on pore segmentation from backscattered electron image analysis

Abstract In this study, the porosity assessment of bonded mortar of recycled aggregates (RA) is done by backscattered electron (BSE) image analysis, for the first time. The resulting effect on concrete mechanical properties was determined as well. Different mixes of recycled aggregate concrete (RAC) were formulated. The concrete specimens were cast to determine the compressive strength at various ages. Later, after 28 – day age, the hardened concrete specimens were used to extract samples for BSE image analysis. The BSE analysis was done by using the pore segmentation method previously developed to ascertain the microstructural features of hydrated cement pastes. From the image analysis, plots of segmented area versus grey value and brightness histograms at different threshold levels were obtained. The BSE image analysis was used to determine the pore area and total area in each frame for porosity assessment. The results indicated that RA affect the resulting mechanical strength of hardened concrete due to the lower specific gravity and higher water absorption as compared to the natural aggregates (NA). Recycled coarse aggregate (RCA) has about 28% higher bonded mortar content (BMC) with porosity of bonded mortar (BM) almost double as that of recycled fine aggregate (RFA). This leads to inferior interface between aggregate and cement paste thereby affecting the overall concrete properties. It is suggested to use pre-treatment methods for surface modification of RA for efficient use in eco-friendly and sustainable construction.

Semi-automated procedure of digitalization and study of rock thin section porosity applying optical image analysis tools

A detailed description of a semi-automated process of digitalization, visualization and quantification of rock thin sections by applying Optical Image Analysis (OIA) is presented. The main goal of this process is to guarantee the reproducibility of the OIA method on images acquired by a high-resolution scanner (HRS, Cannon V850 Pro) from thin sections, emulating an optical microscope. The method consists of the following sequence of operations and functions: i) image acquisition under polarized conditions by image analysis programs (SilverFast SE Plus 8I), ii) spatial correction of images to ensure correct spatial correlation and registration by using a Geographic Information System software (QGIS 2.14.0-Essen), iii) pore identification and pore size and shape parameter quantification (ImageJ 1.50i), iv) storage of the high resolution polarized images. This semi-automated process grants effective pore segmentation based on images of different cross-polarized light conditions (90°/0°; 135°/45°) and plane-polarized light conditions (90°/-; 0°/-) of the same petrographic scene. The optical resolution of the used images (6400 dots per inch) enables the visualization of pores with diameters ranging from 3.97 μm to 2000 μm. The proposed method was applied to and tested by the quantification of optical pores of two sedimentary rocks before and after their exposure to supercritical CO2 (t: 970 h). Through this process, rock pore evolution after interaction with supercritical CO2 can be correctly assessed. The applied OIA system and, in particular, the sets of obtained high resolution polarized digital images provided a basis for the development of a remote visualization system. This system makes it possible to virtually represent thin sections similarly to a simple virtual microscope. The described workflow offers a fast, effective and reproducible methodology that could complement petrographic studies.

Quantitative analysis on porosity of reactive powder concrete based on automated analysis of back-scattered-electron images

In this study an automated analysis of back-scattered-electron images (SEM-BSE) for segmentation of the porosity of reactive powder concrete is presented, making use of Matlab. The development of the proposed threshold method is based on a classification of the images according to two different brightness levels, which correspond to low and high grey value thresholds. The intersection point of the tangent lines in the area segmented versus grey value curve is used as the upper threshold to determine threshold 1 and threshold 2 and the resulting pore volume is validated with mercury intrusion porosimetry. Nevertheless, a difference between porosities obtained by SEM-BSE and MIP can be expected, because the first method can measure all pore types, while the second will only measure open pores. The result shows that the use of the proposed threshold method achieves a good correlation with MIP for pore segmentation in the RPC microstructure.

Analysis of proppant particles porosity based on microCT image processing

The proppant is a granular material with a typical size of 0.2 to 1.2 mm. It is used to prevent the closure of fractures in the reservoir created by hydraulic fracturing procedure, which is actively used in the oil and gas industry. Some types of proppant are manufactured from porous technical ceramics. Presence of internal voids can dramatically decrease the proppant grain mechanical strength and consequently proppant pack conductivity under natural stress. For a detailed study of proppant particles’ internal porosity structure and its relation to the pack’s strength, we applied X-ray microtomography (microCT), which allows to observe this structure non-destructively.In the work, we presented our approaches for digital analysis of reconstructed 3D microCT images for studying the internal voids and the homogeneity of their distribution inside the proppant. We used an automatic thresholding for primary segmentation of pores and particles. We apply 3D marker-controlled watershed to separate individual proppant particles. We propose features for characterization of radial and layered porosity distribution for each particle and homogeneity evaluation. The correctness of our method was tested on synthetic models. Current results indicate probable dependence of proppant strength properties on its internal porosity, but not on the homogeneity of porosity distribution.

BMD-based assessment of local porosity in human femoral cortical bone.

Cortical pores are determinants of the elastic properties and of the ultimate strength of bone tissue. An increase of the overall cortical porosity (Ct.Po) as well as the local coalescence of large pores cause an impairment of the mechanical competence of bone. Therefore, Ct.Po represents a relevant target for identifying patients with high fracture risk. However, given their small size, the in vivo imaging of cortical pores remains challenging. The advent of modern high-resolution peripheral quantitative computed tomography (HR-pQCT) triggered new methods for the clinical assessment of Ct.Po at the peripheral skeleton, either by pore segmentation or by exploiting local bone mineral density (BMD). In this work, we compared BMD-based Ct.Po estimates with high-resolution reference values measured by scanning acoustic microscopy. A calibration rule to estimate local Ct.Po from BMD as assessed by HR-pQCT was derived experimentally. Within areas of interest smaller than 0.5 mm2, our model was able to estimate the local Ct.Po with an error of 3.4%. The incorporation of the BMD inhomogeneity and of one parameter from the BMD distribution of the entire scan volume led to a relative reduction of the estimate error of 30%, if compared to an estimate based on the average BMD. When applied to the assessment of Ct.Po within entire cortical bone cross-sections, the proposed BMD-based method had better accuracy than measurements performed with a conventional threshold-based approach.

Characterization of Paper Coating Structure Using FIB and FE-SEM. 1. New Method for Image Analysis

We have developed a new method for characterizing paper coating structure using FIB milling technique and FE-SEM. A new image analysis method combining smart blur and locally adaptive thresholding technique was applied. Compared to conventional method, this method preserved the edge of target and gave better automatic image segmentation of pores. We applied this method for the measurement of the pore structure in pigment coating layers. Staining with OsO4 and resin-embedding treatments were used for the sample preparation of coated papers for image analysis. A control sample without any treatment was also used in analysis. BSE images of osmium stained and resin-embedded coated paper gave the clearest image segmentation of coating components. The results showed that three sample preparation methods gave similar z-directional porosities. Lower porosity and smaller pore size was obtained from the top surface layer with the osmium stained and resin embedded samples.

Pore Space Quantification of Sedimentary Rocks before-after Supercritical CO2 Interaction by Optical Image Analysis

Optical image analysis (OIA) allows a quantitative characterization of pore space measured through thin sections. Together with petrographic studies, by using optical microscopy (OpM), OIA could provide a reproducible pore characterization of sedimentary rocks in applications related to the geological storage of CO2. This research is focused on the application of the OIA technique to measure porosity on thin section samples of sandstones from Spain before and after CO2 injection to supercritical (SC) conditions in autoclave (P ≈ 7.5 MPa and T ≈ 35°C). The aim of this article is to show preliminary measures by OIA using images acquired with high-resolution scanner (HRS). Mineral images were acquired from thin sections before and after SC-CO2 injection. The optical measures of porosity on the thin sections of sandstones showed an effective pore segmentation considering different images in cross-polarized light conditions (90°/0°; 120°/30°) and plane-polarized light conditions (90°/-) of the same petrographic scene. Mineralogical studies by OpM and OIA-HRS have allowed an exhaustive characterization of the samples studied and a preliminary approximation of the CO2-rock interactions. This study shows a fast, effective and reproducible methodology that allowed the researchers to obtain qualitative information about changes in the pore network (optical porosity) distribution on the thin section that could be applied to similar experimental injection tests.

3D Monte Carlo simulation of backscattered electron signal variation across pore-solid boundaries in cement-based materials

Three-dimensional (3D) Monte Carlo simulation was used to study the variation of backscattered electron (BSE) signal across pore-solid boundaries in cement-based materials in order to enhance quantitative analysis of pore structure. The effects of pore size, depth and boundary inclination angle were investigated. It is found that pores down to 1nm can generate sufficient contrast to be detected. Visibility improves with larger pore size, smaller beam probe size and lower acceleration voltage. However, pixels in shallow pores or near pore boundaries display higher grey values (brightness) than expected due to sampling sub-surface or neighbouring solid material. Thus, cement-based materials may appear less porous or the pores appear smaller than they actually are in BSE images. Simulated BSE images were used to test the accuracy of the Overflow pore segmentation method. Results show the method is generally valid and gives low errors for pores that are 1μm and greater.

Multi-Scale Correlative Tomography of a Li-Ion Battery Composite Cathode.

Focused ion beam/scanning electron microscopy tomography (FIB/SEMt) and synchrotron X-ray tomography (Xt) are used to investigate the same lithium manganese oxide composite cathode at the same specific spot. This correlative approach allows the investigation of three central issues in the tomographic analysis of composite battery electrodes: (i) Validation of state-of-the-art binary active material (AM) segmentation: Although threshold segmentation by standard algorithms leads to very good segmentation results, limited Xt resolution results in an AM underestimation of 6 vol% and severe overestimation of AM connectivity. (ii) Carbon binder domain (CBD) segmentation in Xt data: While threshold segmentation cannot be applied for this purpose, a suitable classification method is introduced. Based on correlative tomography, it allows for reliable ternary segmentation of Xt data into the pore space, CBD, and AM. (iii) Pore space analysis in the micrometer regime: This segmentation technique is applied to an Xt reconstruction with several hundred microns edge length, thus validating the segmentation of pores within the micrometer regime for the first time. The analyzed cathode volume exhibits a bimodal pore size distribution in the ranges between 0–1 μm and 1–12 μm. These ranges can be attributed to different pore formation mechanisms.

Analysis of irradiated U-7wt%Mo dispersion fuel microstructures using automated image processing

The High Performance Research Reactor Fuel Development (HPPRFD) program is responsible for developing low enriched uranium (LEU) fuel substitutes for high performance reactors fueled with highly enriched uranium (HEU) that have not yet been converted to LEU. The uranium-molybdenum (U-Mo) fuel system was selected for this effort. In this study, fission gas pore segmentation was performed on U-7wt%Mo dispersion fuel samples at three separate fission densities using an automated image processing interface developed in MATLAB. Pore size distributions were attained that showed both expected and unexpected fission gas behavior. In general, it proved challenging to identify any dominant trends when comparing fission bubble data across samples from different fuel plates due to varying compositions and fabrication techniques. The results exhibited fair agreement with the fission density vs. porosity correlation developed by the Russian reactor conversion program.

Durability of mortar with calcined marl as supplementary cementing material

Calcined marl – ordinary ‘blue' clay containing calcium carbonate – is an effective pozzolan in cementitious materials in terms of strength if calcined at the correct temperature. Marl is unsuitable for the clay product industry, and might thus be the high-volume available alternative pozzolan globally that the cement and concrete industry is looking for in order to reduce its carbon footprint. Mortars are made where calcined marl replaces cement by volume at levels of 0, 20, 35, 50 and 65 vol.%. Excellent strengths at both 1 and 28 d relative to reference are obtained for up to 50 vol.% replacement, and the strength continues to increase up to 1 year. The capillary porosity volume increases along with the electrical resistivity, indicating a pore refinement or segmentation, as well as perhaps reduced ionic strength in the pore-water. These are properties that might reduce the propagation rate of rebar corrosion once initiated by carbonation, as carbonation rate increased with cement replacement. The chloride ingress is significantly reduced when cement is replaced by calcined marl. Sulfate resistance is also tested at both 5 and 20°C, but no significant changes are observed after 9 months' exposure.

Pore REconstruction and Segmentation (PORES) method for improved porosity quantification of nanoporous materials.

Electron tomography is currently a versatile tool to investigate the connection between the structure and properties of nanomaterials. However, a quantitative interpretation of electron tomography results is still far from straightforward. Especially accurate quantification of pore-space is hampered by artifacts introduced in all steps of the processing chain, i.e., acquisition, reconstruction, segmentation and quantification. Furthermore, most common approaches require subjective manual user input. In this paper, the PORES algorithm "POre REconstruction and Segmentation" is introduced; it is a tailor-made, integral approach, for the reconstruction, segmentation, and quantification of porous nanomaterials. The PORES processing chain starts by calculating a reconstruction with a nanoporous-specific reconstruction algorithm: the Simultaneous Update of Pore Pixels by iterative REconstruction and Simple Segmentation algorithm (SUPPRESS). It classifies the interior region to the pores during reconstruction, while reconstructing the remaining region by reducing the error with respect to the acquired electron microscopy data. The SUPPRESS reconstruction can be directly plugged into the remaining processing chain of the PORES algorithm, resulting in accurate individual pore quantification and full sample pore statistics. The proposed approach was extensively validated on both simulated and experimental data, indicating its ability to generate accurate statistics of nanoporous materials.

Understanding sintering characteristics of ZnO nanoparticles by FIB-SEM three-dimensional analysis

Abstract In this study, we utilize three-dimensional (3D) reconstruction by focused ion beam (FIB) cutting and SEM imaging to understand the evolution of pore volume, pore-solid interfacial area, pore shape, pore connectivity, and pore number during the two-step sintering of ZnO nanoparticles. After the first-step sintering, the density is 75% and all the pores are connected. During the second-step sintering, the decrease of pore volume leads to the segmentation of pores and formation of closed pores. The shape of closed pores is irregular. The pore number first increases with the formation of closed pores, and then decreases due to the disappearance of small pores. The pore-solid interfacial area keeps decreasing during sintering. FIB-SEM 3D reconstruction offers an opportunity to directly and quantitatively observe the pore evolution and understand the sintering process at nanoscale.