TY - JOUR
T1 - Comparison of 2D and 3D dynamic image analysis for characterization of natural sands
AU - Li, Linzhu
AU - Iskander, Magued
N1 - Publisher Copyright:
© 2021 Elsevier B.V.
PY - 2021/9/5
Y1 - 2021/9/5
N2 - The efficacy of Dynamic Image Analysis (DIA) for evaluating particle size and shape parameters was explored using three natural sands, having varying particle morphologies. Two-dimensional (2D) captures binary images of the particles as they free fall in the imaging frame. Although 2D DIA is practical for statistical size and shape analysis, there is a prevailing perception that it fails to fully quantify particle granulometry. In the past few years, 3D DIA has been introduced and has gained acceptance in the pharmaceutical industry. In 3D DIA the system tracks a particle as it falls through the imaging frame and captures gray-scale images from 8 to 12 perspectives of the same particle, and the results are analyzed using average values of these 2D images, which are believed to verge on true particle morphology. Although 2D and 3D devices employ similar methodology they differ in resolution, frame rate, lighting systems, and algorithms. In this work we compare the performance of 2D and 3D DIA. Particle size distributions were expressed using EQPC and a variety of Feret diameters, while particle shape descriptors including Aspect Ratio, Sphericity, Convexity and Roundness were compared for both systems. It is shown that 3D DIA requires a smaller number of sand particles to achieve mean particle shape values. Particle size characterization is generally independent of the machines and algorithms used in this study; however, 3D DIA provides maximum and minimum particle axes which are closer to the real sand particle sizes. Image-based particle shape characterization is more sensitive to the technology employed; it largely depends on image quality, particle angularity, and a hierarchy of shape descriptors; thus, at this time shape analysis for engineering applications must be carried out with similar machines and algorithms. In particular, the image resolutions captured by the available 2D and 3D DIA apparatus are 4 μm and 15 μm per pixel, respectively. At this time, the higher resolution and shorter exposure time of 2D DIA permits particle shape analysis down to particle sizes having D50 on the order of 40 μm, while 3D DIA is limited to D50 larger than 150 μm. These thresholds will certainly change as technology allows finer resolutions and shorter exposure times. Finally, the importance of image processing cannot be overlooked, and manufacturers and researchers are encouraged to open-source their algorithms in order to establish confidence in them.
AB - The efficacy of Dynamic Image Analysis (DIA) for evaluating particle size and shape parameters was explored using three natural sands, having varying particle morphologies. Two-dimensional (2D) captures binary images of the particles as they free fall in the imaging frame. Although 2D DIA is practical for statistical size and shape analysis, there is a prevailing perception that it fails to fully quantify particle granulometry. In the past few years, 3D DIA has been introduced and has gained acceptance in the pharmaceutical industry. In 3D DIA the system tracks a particle as it falls through the imaging frame and captures gray-scale images from 8 to 12 perspectives of the same particle, and the results are analyzed using average values of these 2D images, which are believed to verge on true particle morphology. Although 2D and 3D devices employ similar methodology they differ in resolution, frame rate, lighting systems, and algorithms. In this work we compare the performance of 2D and 3D DIA. Particle size distributions were expressed using EQPC and a variety of Feret diameters, while particle shape descriptors including Aspect Ratio, Sphericity, Convexity and Roundness were compared for both systems. It is shown that 3D DIA requires a smaller number of sand particles to achieve mean particle shape values. Particle size characterization is generally independent of the machines and algorithms used in this study; however, 3D DIA provides maximum and minimum particle axes which are closer to the real sand particle sizes. Image-based particle shape characterization is more sensitive to the technology employed; it largely depends on image quality, particle angularity, and a hierarchy of shape descriptors; thus, at this time shape analysis for engineering applications must be carried out with similar machines and algorithms. In particular, the image resolutions captured by the available 2D and 3D DIA apparatus are 4 μm and 15 μm per pixel, respectively. At this time, the higher resolution and shorter exposure time of 2D DIA permits particle shape analysis down to particle sizes having D50 on the order of 40 μm, while 3D DIA is limited to D50 larger than 150 μm. These thresholds will certainly change as technology allows finer resolutions and shorter exposure times. Finally, the importance of image processing cannot be overlooked, and manufacturers and researchers are encouraged to open-source their algorithms in order to establish confidence in them.
KW - Elongation index
KW - Flatness index
KW - Granulometry
KW - Johnson coefficients
KW - Zingg and Krumbein chart
UR - http://www.scopus.com/inward/record.url?scp=85106501604&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85106501604&partnerID=8YFLogxK
U2 - 10.1016/j.enggeo.2021.106052
DO - 10.1016/j.enggeo.2021.106052
M3 - Article
AN - SCOPUS:85106501604
SN - 0013-7952
VL - 290
JO - Engineering Geology
JF - Engineering Geology
M1 - 106052
ER -