Advantages of multi-region kriging over bi-region techniques for computed tomography-scan segmentation
M. Azhar A , X. Chang A , J. Debes A , P. Delmas
A G , C. Duwig B , N. Dal Ferro C , T. Gee A , J. Marquez D , F. Morari C , K. Müller E , T. Mukunoki F , I. Piccoli C and A. Gastelum Strozzi D
A Department of Computer Science, The University of Auckland, Auckland, New Zealand.
B Université Grenoble Alpes, CNRS, IRD, Grenoble INP, IGE, F-38000 Grenoble, France.
C Department of Agronomy, Food, Natural Resources, Animals and Environment, Agripolis, 10 University of Padova, Viale Dell’Università 16, 35020 Legnaro, Italy.
D Centro de Ciencias Aplicadas y Desarrollo Tecnológico, Universidad Nacional Autónoma de México, Mexico, D.F., México.
E The New Zealand Institute for Plant and Food Research Limited (PFR), Production Footprints, Bisley Road, Hamilton 3214, New Zealand.
F Faculty of Advanced Science and Technology, Kumamoto University, Kumamoto 860-8555, Japan.
G Corresponding author. Email: p.delmas@auckland.ac.nz
Soil Research 57(6) 521-534 https://doi.org/10.1071/SR18294
Submitted: 30 September 2018 Accepted: 27 June 2019 Published: 29 August 2019
Abstract
Quantifying the structure of soil is essential for developing effective soil management for farming and environmental conservation efforts. One approach to quantify soil structure is to scan intact soil cores by X-ray computed tomography (CT), which allows using computer vision algorithms to identify internal components within the soil. One commonly used approach is the colour-based segmentation of CT-scan soil images into two regions – matter and void – for the purpose of determining the soil porosity. A key problem with this approach is that soil CT images tend to be rather complicated, and thus this type of bi-region segmentation is a non-trivial problem, with algorithms following this type of bi-region approach typically performing unreliability across a variety of image sets. In this work, a technique is proposed that identifies an optimal number of regions present in the soil, rather than just two. It is claimed that this more sophisticated representation of soil structure leads to a more accurate representation than traditional bi-region segmentation; however, it is reducible to a bi-region segmentation yielding the required estimation of porosity with more accuracy and robustness than traditional methods. It is also proposed that segmentation is performed using a multi-region kriging algorithm, which establishes relationships between distance and regions that allows the segmentation to overcome many of the artefacts and noise issues associated with CT scanning. Our experiments focused on layer-by-layer segmentation and results demonstrated that the proposed approach produced segmentations consistent across a variety of scanned cores and were visually more correct than current state-of-the-art bi-region techniques.
References
Barnes R (2003). Variogram tutorial. Available at http://www.goldensoftware.com/variogramTutorial.pdf [verified 9 June 2016].Barrett JF, Keat N (2004) Artifacts in CT: recognition and avoidance. Radiographics 24, 1679–1691.
| Artifacts in CT: recognition and avoidance.Crossref | GoogleScholarGoogle Scholar | 15537976PubMed |
Baveye PC, Laba M, Otten W, Bouckaert L, Dello Sterpaio P, Goswami RR, Grinev D, Houston A, Hu Y, Liu J, Mooney S, Pajor R, Sleutel S, Tarquis A, Wang W, Wei Q, Sezgin M (2010) Observer-dependent variability of the thresholding step in the quantitative analysis of soil images and X-ray microtomography data. Geoderma 157, 51–63.
| Observer-dependent variability of the thresholding step in the quantitative analysis of soil images and X-ray microtomography data.Crossref | GoogleScholarGoogle Scholar |
Bezdek JC (1980) A convergence theorem for the fuzzy ISODATA clustering algorithms. IEEE Transactions on Pattern Analysis and Machine Intelligence PAMI-2, 1–8.
| A convergence theorem for the fuzzy ISODATA clustering algorithms.Crossref | GoogleScholarGoogle Scholar |
Blimes JA (1998) A gentle tutorial of the EM algorithm and its application to parameter estimation for Gaussian mixture and hidden Markov models. International Computer Science Institute 4, 126–140.
Bohling G (2005). Kriging. Kansas Geological Survey, University of Kansas Technical Report.
Bonilla FA (1997). Assessment of pesticide exposure levels in soils and groundwater at the regional scale, MSc University of California, Davis.
Carr JR, de Miranda FP (1998) The semivariogram in comparison to the co-occurrence matrix for classification of image texture. IEEE Transactions on Geoscience and Remote Sensing 36, 1945–1952.
| The semivariogram in comparison to the co-occurrence matrix for classification of image texture.Crossref | GoogleScholarGoogle Scholar |
Clark, I. (1979). ‘Practical geostatistics.’Applied Science Publishers, London.
Deutsch CV (1996) Correcting for negative weights in ordinary kriging. Computers & Geosciences 22, 765–773.
| Correcting for negative weights in ordinary kriging.Crossref | GoogleScholarGoogle Scholar |
Dunn JC (1973) A fuzzy relative of the ISODATA process and its use in detecting compact well-separated clusters. Journal of Cybernetics 3, 32–57.
| A fuzzy relative of the ISODATA process and its use in detecting compact well-separated clusters.Crossref | GoogleScholarGoogle Scholar |
Glasbey CA, Horgan GW, Darbyshire JF (1991) Image analysis and three-dimensional modelling of pores in soil aggregates. Journal of Soil Science 42, 479–486.
| Image analysis and three-dimensional modelling of pores in soil aggregates.Crossref | GoogleScholarGoogle Scholar |
Gonzalez RC, Woods RE (2008). ‘Digital Image Processing.’ (Prentice Hall, Upper Saddle River, New Jersey)
Gupta MR, Chen Y (2011). ‘Theory and use of the EM algorithm.’ (Now Publishers Inc, Hanover, United States.)
Haralick RM, Shanmugam K, Dinstein I (1973) Textural features for image classification. IEEE Transactions on Systems, Man, and Cybernetics SMC-3, 610–621.
| Textural features for image classification.Crossref | GoogleScholarGoogle Scholar |
Havlin JL, Beaton JD, Tisdale SL, Nelson WL (2005). ‘Soil fertility and fertilizers: an introduction to nutrient management.’ (Pearson Prentice Hall, Upper Saddle River, NJ, USA.
Helliwell JR, Sturrock CJ, Grayling KM, Tracy SR, Flavel RJ, Young IM, Whalley WR, Mooney SJ (2013) Applications of X-ray computed tomography for examining biophysical interactions and structural development in soil systems: a review. European Journal of Soil Science 64, 279–297.
| Applications of X-ray computed tomography for examining biophysical interactions and structural development in soil systems: a review.Crossref | GoogleScholarGoogle Scholar |
Houston AN, Otten W, Baveye PC, Hapca S (2013) Adaptive-window indicator kriging: A thresholding method for computed tomography images of porous media. Computers & Geosciences 54, 239–248.
| Adaptive-window indicator kriging: A thresholding method for computed tomography images of porous media.Crossref | GoogleScholarGoogle Scholar |
Hudson N (2015). ‘Soil conservation: fully revised and updated.’ 3rd Edn, Scientific Publisher, India.
Iassonov P, Gebrenegus T, Tuller M (2009) Segmentation of X-ray computed tomography images of porous materials: A crucial step for characterization and quantitative analysis of pore structures. Water Resources Research 45,
| Segmentation of X-ray computed tomography images of porous materials: A crucial step for characterization and quantitative analysis of pore structures.Crossref | GoogleScholarGoogle Scholar |
Lin E, Alessio A (2009) What are the basic concepts of temporal, contrast, and spatial resolution in cardiac CT? Journal of Cardiovascular Computed Tomography 3, 403–408.
| What are the basic concepts of temporal, contrast, and spatial resolution in cardiac CT?Crossref | GoogleScholarGoogle Scholar | 19717355PubMed |
Mukunoki T, Miyata Y, Mikami K, Shiota E (2016) X-ray CT analysis of pore structure in sand. Solid Earth 7, 929–942.
| X-ray CT analysis of pore structure in sand.Crossref | GoogleScholarGoogle Scholar |
Nimmo JR (2004). Porosity and pore size distribution. In in Hillel, D., ed. Encyclopedia of Soils in the Environment: London, Elsevier, v. 3, p. 295–303.
Oh W, Lindquist B (1999) Image thresholding by indicator kriging. IEEE Transactions on Pattern Analysis and Machine Intelligence 21, 590–602.
| Image thresholding by indicator kriging.Crossref | GoogleScholarGoogle Scholar |
Otsu N (1979) A threshold selection method from gray-level histograms. IEEE Transactions on Systems, Man, and Cybernetics 9, 62–66.
| A threshold selection method from gray-level histograms.Crossref | GoogleScholarGoogle Scholar |
Peth S, Horn R, Beckmann F, Donath T, Fischer J, Smucker AJM (2008) Three-dimensional quantification of intra-aggregate pore-space features using synchrotron-radiation-based microtomography. Soil Science Society of America Journal 72, 897–907.
| Three-dimensional quantification of intra-aggregate pore-space features using synchrotron-radiation-based microtomography.Crossref | GoogleScholarGoogle Scholar |
Prado B, Duwig C, Hidalgo C, Gómez D, Yee H, Prat C, Esteves M, Etchevers JD (2007) Characterization, functioning and classification of two volcanic soil profiles under different land uses in Central Mexico. Geoderma 139, 300–313.
| Characterization, functioning and classification of two volcanic soil profiles under different land uses in Central Mexico.Crossref | GoogleScholarGoogle Scholar |
Prodanović M, Lindquist WB, Seright RS (2006) Porous structure and fluid partitioning in polyethylene cores from 3D X-ray microtomographic imaging. Journal of Colloid and Interface Science 298, 282–297.
| Porous structure and fluid partitioning in polyethylene cores from 3D X-ray microtomographic imaging.Crossref | GoogleScholarGoogle Scholar | 16364351PubMed |
Ridler TW, Calvard S (1978) Picture thresholding using an iterative selection method. IEEE Transactions on Systems, Man, and Cybernetics 8, 630–632.
| Picture thresholding using an iterative selection method.Crossref | GoogleScholarGoogle Scholar |
Thompson LM (1952) Soils and soil fertility. Soil Science 74, 264
| Soils and soil fertility.Crossref | GoogleScholarGoogle Scholar |
Tibshirani R, Walther G, Hastie T (2001) Estimating the number of clusters in a data set via the gap statistic. Journal of the Royal Statistical Society. Series B, Statistical Methodology 63, 411–423.
| Estimating the number of clusters in a data set via the gap statistic.Crossref | GoogleScholarGoogle Scholar |
Tomasi C, Manduchi R (1998). Bilateral filtering for gray and color images. In Sixth International Conference on Computer Vision (IEEE Cat. No.98CH36271), 7 January 1998, Bombay, India.
Velasco, F. R. (1979). Thresholding using the isodata clustering algorithm. Maryland University, College Park. Computer Science Center, Technical Report Nbr. 198004.s
Voroney RP, Heck RJ (2015). The soil habitat. In ‘Soil microbiology, ecology and biochemistry’. pp. 15–39. Elsevier Academic Press, 30 Corporate Drive, Suite 400, Burlington, MA 01803, USA.
Wang R (2016). Unsupervised classification - clustering. Available at http://fourier.eng.hmc.edu/e161/lectures/classification/node12.html [verified 10 June 2016].
Wang W, Kravchenko AN, Smucker AJM, Rivers ML (2011) Comparison of image segmentation methods in simulated 2D and 3D microtomographic images of soil aggregates. Geoderma 162, 231–241.
| Comparison of image segmentation methods in simulated 2D and 3D microtomographic images of soil aggregates.Crossref | GoogleScholarGoogle Scholar |
Weng Y-T, Wang C-C, Chiang C-C, Tsai H, Song Y-F, Huang S-T, Liang B (2018) In situ evidence of mineral physical protection and carbon stabilization revealed by nanoscale 3-D tomography. Biogeosciences 15, 3133–3142.
| In situ evidence of mineral physical protection and carbon stabilization revealed by nanoscale 3-D tomography.Crossref | GoogleScholarGoogle Scholar |
Zhang Y, Brady M, Smith S (2001) Segmentation of brain MR images through a hidden Markov random field model and the expectation-maximization algorithm. IEEE Transactions on Medical Imaging 20, 45–57.
| Segmentation of brain MR images through a hidden Markov random field model and the expectation-maximization algorithm.Crossref | GoogleScholarGoogle Scholar | 11293691PubMed |
