Landmine Models

Ready-to-use geometry and material files for PMA-1, PMN and TS-50 landmines, and a metal can as a false target. Each model is supplied at 1 mm and 2 mm cubic mesh spacing.

Files to use

Use the HDF5 geometry and JSON materials in this folder. Both resolutions of a model share one material file; for example:

LandmineModels/
    PMN_1x1x1.h5
    PMN_2x2x2.h5
    PMN_materials.json
    examples/free_space.py
    legacy/                 # Original HDF5 and text files for older readers.

The same naming applies to PMA, TS50 and can. Keep the material file beside the geometry files. The JSON lists named materials and their electromagnetic properties; both resolutions use those same properties.

The simulation mesh must match the selected file’s spacing. These are the actual stored dimensions, in millimetres:

Table 34 Geometry extents (x × y × z)

Model	1 mm geometry	2 mm geometry
PMA	140 × 64 × 33	140 × 64 × 32
PMN	116 × 156 × 49	116 × 156 × 48
TS50	90 × 90 × 44	90 × 90 × 44
can	76 × 76 × 108	76 × 76 × 108

Run the example

From the repository root, in your gprMax Python environment:

python -m gprMax.toolboxes.LandmineModels.examples.free_space --model PMN --resolution-mm 2

This imports the target, adds a theoretical source and receiver above it, and runs a short CPU simulation. The output is landmine_results/PMN_2mm/target.h5. Choose another target with --model PMA, --model TS50 or --model can, and use --resolution-mm 1 for 1 mm spacing. Use --directory results/new_run to choose a fresh output folder.

Edit the example to set your background, antenna, target position and recording time. The default 3 ns run checks model import and execution; it is not a calibrated GPR measurement. The example shows how to build the model, execute it, and read the receiver’s electric field from the output file.

Add a target to your model

In a saved Python script launched from the repository root:

from pathlib import Path
import gprMax

geometry = Path("gprMax/toolboxes/LandmineModels/PMN_2x2x2.h5").resolve()
# Your scene must already define a 2 mm mesh, a sufficiently large domain,
# and the background geometry. Add the target after the background.
target = gprMax.GeometryObjectsRead(
    p1=(0.032, 0.032, 0.032),  # Lower corner of the imported array, in metres.
    geofile=geometry,
    material_database="PMN_materials",  # JSON filename without .json.
    averaging=False,
)
# scene.add(target)

In a hash-command input file, use the equivalent command below, with the geometry and JSON files beside your input file:

#geometry_objects_read: 0.032 0.032 0.032 PMN_2x2x2.h5 PMN_materials

Build the soil or other background before importing the target. Transparent voxels (-1 in the geometry array) leave that background unchanged; explicitly stored air voxels insert air. Changing the simulation mesh does not resample the stored target.

Reference and licence

Author/Contact: Iraklis Giannakis (iraklis.giannakis@abdn.ac.uk), University of Aberdeen, UK

License: Creative Commons Attribution-ShareAlike 4.0 International License

Attribution/cite: Giannakis, I., Giannopoulos, A., Warren, C. (2016). A Realistic FDTD Numerical Modeling Framework of Ground Penetrating Radar for Landmine Detection. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 9(1), 37–51. https://doi.org/10.1109/JSTARS.2015.2468597

The paper describes the models and how their dielectric properties were fitted to laboratory measurements of scattered fields in free space. The same licence and attribution apply to the current files and the unchanged originals in legacy/.

Fig. 64 FDTD geometry mesh showing the PMA-1 landmine model.

Fig. 65 FDTD geometry mesh showing the PMN landmine model.