Methodology for Optical Imaging of NAPL 3D Distribution in Transparent Porous Media

被引:18
作者
Kashuk, Sina [1 ,2 ]
Mercurio, Sophia R. [2 ]
Iskander, Magued [2 ]
机构
[1] CUNY, New York, NY 10017 USA
[2] NYU, Civil & Urban Engn Dept, Brooklyn, NY 11201 USA
来源
GEOTECHNICAL TESTING JOURNAL | 2015年 / 38卷 / 05期
基金
美国国家科学基金会;
关键词
bench-scale study; optical image processing; transparent soil; color dye tracer; colorimetry; napl volume mapping; 3D reconstruction; discrete tomography; back projection; 3D model; LIGHT TRANSMISSION; COLOR SPACE; TRANSPORT; FLOW; VISUALIZATION; MODEL; RECONSTRUCTION; SEGMENTATION; POLYOMINOES; OIL;
D O I
10.1520/GTJ20140153
中图分类号
P5 [地质学];
学科分类号
0709 ; 081803 ;
摘要
Three-dimensional mapping of non-aqueous phase liquid (NAPL) distribution within saturated porous media is an important issue in bench-scale geo-environmental studies. In this paper, a methodology for optical imaging of 3D multiphase liquid distribution is presented. The natural aquifer is simulated using a transparent soil surrogate that represents the macroscopic behavior of natural sand. To achieve transparency transparent fused quartz grains were saturated with a matched refractive index mineral oil solution that represents the natural aquifer and injected with a green-dyed sucrose solution to simulate dense NAPL contamination. The spatial volume was first estimated and then reconstructed using orthogonal images acquired at the model boundaries at various stages of contamination and remediation. Color space analysis was employed to segment the NAPL zone and transform pixel information into integrated concentration values using a previously published calibration model. The chromatic components C-R and a of YCBCR and Lab, respectively, were combined to render the spatial concentration profile. A novel iterative reconstruction algorithm named 3D carving was used to resolve three 2D projections into a 3D model. The results show that the proposed methodology provides better efficacy for NAPL zone reconstruction in comparison with conventional image analysis routines. The technology presented in this paper is a sustainable, fast, accurate, non-intrusive, and inexpensive method for spatial mapping of contamination zones using bench-scale transparent soil models.
引用
收藏
页码:603 / 619
页数:17
相关论文
共 68 条
  • [1] Fluorescence imaging of tracer distributions in soil profiles
    Aeby, P
    Schultze, U
    Braichotte, D
    Bundt, M
    Moser-Boroumand, F
    Wydler, H
    Fluhler, H
    [J]. ENVIRONMENTAL SCIENCE & TECHNOLOGY, 2001, 35 (04) : 753 - 760
  • [2] [Anonymous], 1986, COLORIMETRY
  • [3] [Anonymous], 2018, Digital Image Processing
  • [4] [Anonymous], 1995, 7095 ITU EBU
  • [5] Balazs P., 2008, INT J SHAPE MODELING, V14, P189
  • [6] Fluorescent dye imaging of the volume sampled by single well forced-gradient tracer tests evaluated in a laboratory-scale aquifer physical model
    Barns, Gareth L.
    Wilson, Ryan D.
    Thornton, Steven F.
    [J]. JOURNAL OF CONTAMINANT HYDROLOGY, 2012, 128 (1-4) : 58 - 70
  • [7] Enhanced biodegradation by hydraulic heterogeneities in petroleum hydrocarbon plumes
    Bauer, Robert D.
    Rolle, Massimo
    Bauer, Sebastian
    Eberhardt, Christina
    Grathwohl, Peter
    Kolditz, Olaf
    Meckenstock, Rainer U.
    Griebler, Christian
    [J]. JOURNAL OF CONTAMINANT HYDROLOGY, 2009, 105 (1-2) : 56 - 68
  • [8] Bear J., 2010, MODELING GROUNDWATER
  • [9] Color spaces and image segmentation
    Busin, Laurent
    Vandenbroucke, Nicolas
    Macaire, Ludovic
    [J]. ADVANCES IN IMAGING AND ELECTRON PHYSICS, VOL 151, 2008, 151 : 65 - 168
  • [10] Reconstructing hv-convex polyominoes from orthogonal projections
    Chrobak, M
    Dürr, C
    [J]. INFORMATION PROCESSING LETTERS, 1999, 69 (06) : 283 - 289