1. Fusion of Multi-Frequency
in GPR
Submitted by: Muddasar Hussain
Submitted to: Sir Dr. Muhammad Younis Khan
Course code: Geop 705
National Centre of Excellence in Geology
University of Peshawar
2. Content
• Introduction
• Early Detection of Delamination
• Effect of Antenna vs Frequency
• Reinforced Concrete
• Imaging Objects
• Advantages of Multi-frequency
• Conclusions
• References
1
3. • Ground-penetrating radar (GPR) is a geophysical method using high-frequency
electromagnetic waves to detect and locate the subsurface objects and
interfaces.
• It can get high resolution but shallow propagation depth with higher centre
frequency, while low resolution but deeper depth with lower centre frequency.
• It’s hard to simultaneously take into account the detection depth and resolution
due to the only one centre frequency of a GPR antenna generally.
• Multi-frequency GPR data fusion is proposed here to try to solve this problem.
2
Introduction
4. Cont..
• Firstly, registration of spatial information of GPR forward simulation data of
three main frequencies is established.
• Then, data fusion process is carried out respectively by using four different
algorithms: time-domain fusion without weight, time-domain fusion with
weight, frequency-domain fusion without weight and frequency-domain
fusion with weight.
• The fusion effect is qualitatively and quantitatively evaluated and
compared with the two factors: information entropy and four-point mean
gradient.
3
5. Cont..
• The results of multi-frequency data fusion show that the fuse profile not only
retains the high resolution capability of the high-frequency antenna in the
shallow part, but also reflects the advantage of the large propagation depth of
the low-frequency antenna, and realizes the complementary advantages of the
high and low frequencies antenna.
• Also, the real field GPR data is handled by the proposed data fusion methods, and
it definitely makes the GPR image clearer and more subsurface information
included.
4
6. • For any kind of work involving
excavation of the ground, the
exact knowledge of the location
of underground structures is a
key element for proper
execution of the work. In this
perspective, the use of GPR is
an excellent tool to obtain all
the information needed to
perform the actions provided
for in an effective way, safe and
without damage to the
infrastruc
5
7. • There is an unfortunate perception held by many engineers that there is little
difference between one radar and any other. If, therefore, one radar fails to
detect a particular feature or problem, it is assumed that this is because the
technique does not work rather than because the wrong radar was used.
• The reality is that frequency and wavelength are entirely relevant and that some
radars, in order to be more “user-friendly” allow the operator restricted access to
the very survey parameters that might make the survey successful.
6
Early Detection of
Delamination
8. Cont..
• Alternatively the access is couched in terms such as “shallow” or
“wet” which, while helpful to the layman, are effectively
undefined and may not comply fully with the requirements of
the investigation. The general rule of thumb for the smallest
target which can be detected is 0.1 wavelength
7
9. Effect of
Antenna vs
Frequency
• The effect of using antennas in a range of frequencies is shown in table
above.
Antenna
Frequency
Wavelength, λ, in Dry Soil
400MHz 25cm
500MHz 20cm
1GHz 10cm
1.5GHz 6.7cm
2.5GHz 4cm
4GHz 2.5cm
6GHz 1.67cm
8
10. Cont..
• The table demonstrates the improvement in target definition derived
directly from increasing frequency. The higher the frequency, the better the
probability of detecting the beginnings of delamination before a sizeable
void has been created.
• Given that the difference in potential target detection is also reflected in
the accuracy of measurement (typically 0.25 of a wavelength), it is
surprising how often mid-range radars are used in preference to high
frequency ones.
9
11. Figure 2: Bridge deck investigation showing delamination at more than 1 level. Data collected using a
4GHz antenna of a 3 multichannel radar.
10
12. Cont..
• The data shows layers in the tarmac with underlying reinforced
concrete (as evidenced by the reinforcement bars). Where
delamination has taken place, the signal strength (amplitude)
increases significantly indicating the presence of air. This forms the
basis for mapping delamination throughout the structure.
11
13. • Figure 3; A good example of delamination
detection using a 6GHz antenna is shown. The
investigation was carried out over a tiled floor
which was suspected of having a number of
inadequately grouted tiles. Repeat survey lines
were taken across the floor with readings every
10mm along lines at 300mm intervals to a depth
of 5ns (equivalent to a depth of 250mm in dry
conditions) and the 2-dimensional data formed
12
14. • There are two critical factors, both dependent on wavelength: the ability to
penetrate between adjacent reinforcement bars (rebars) and the depth to which
inspection can be made below the rebars.
• . Since metal reflects radio waves, the wavelength requires to be shorter than the
spacing of the rebar mesh for the signal to penetrate through the reinforcement.
• The maximum depth of penetration is also a function of the wavelength but the
500mm or 400mm depth reach of a very high frequency radar is not normally a
limitation.
13
Reinforced Concrete
15. • Figure; illustrates data from an investigation to determine the
position, depth and, if possible size, of rebars within concrete
slabs forming the floors and ceilings of a modular built block of
flats in Islington, London (Utsi et al, 2004).
14
16. Figure 6: Bridgedeck information
(composite layers, delamination,
reinforcement) from a 4GHz GPR
survey
Figure 7: Bridgedeck Information
(composite layers, delamination,
reinforcement and jointing) from
a 6GHz GPR Survey
15Bridgedeck Examination
17. Figure 8: Horizontal view of the tomb showing
a near circular object at the western end.
Figure 9: Horizontal view of the tomb showing the
reduction of the larger circular object to a narrower
circular object at the western end.
16Imaging Objects
18. Cont..
• The 3-dimensional data set shows the presence of a circular object
whose diameter changes with depth, decreasing and then increasing
again towards the base but which is centered on the same point.
• illustrate two levels of that profile, essentially cutting through one
part of the upper large anomaly and the part below with a much
smaller diameter.
17
19. • being able to measure more accurately,
• being able to detect problems such as delamination at an early stage,
• being able to image objects with greater clarity.
18
Advantages of Multi-frequency
20. • A basic understanding of the effect of selecting one frequency of antenna
over another can be very beneficial to the geophysicist/engineer faced with
detecting delamination, investigating concrete structures or other built
structures, including heritage buildings.
• The use of a very high frequency radar has the potential to give much
earlier notice of problems developing in the built structure, be it bridge,
road, or building, than can a lower frequency GPR. This is simply due to the
shorter wavelengths enabling earlier detection and finer resolution.
19
Conclusion
21. • Bungey, J (2004) “Concrete” in Daniels, D. (2004). Ground Penetrating Radar, 2nd edition, IEE,
Stevenage, UK, pp402-404.
• Calia, A., Colangiuli, D., Leucci, G., Matera, L., Lettieri, M., Persico, R., and Sileo, M. (2012).
• “Non-destructive and Laboratory Diagnostic Study on the mosaic of the crypt of St Nicholas” in
Yongsheng, L. and Xiongyao, X. (eds) the Proceedings of the 14th International Conference on
Ground Penetrating Radar, 4-8 June, Shanghai, China, Vol 2, pp583-588
• Daniels, D. (2004). “Ground Penetrating Radar”, 2nd edition, IEE, Stevenage, UK, pp381-422.
• Giannopoulos, A. GPRMAX: Finite-Difference Time-Domain Simulator for Ground Penetrating
Radar at http://www.gprmax.org.
20References
22. Cont..
• Leucci, G. (2002). “Ground-Penetrating Radar Survey to Map the Location of Buried Structures
under Two Churches in Archaeological Prospection Vol 9, pp 217-228.
• Martinaud, M., Frappa, M. and Chapoulie, R. (2004). “GPR Signals for the understanding of the
shape and filling of man-made underground masonry” in Slob, E., Yarovoy, A. and Rhebergen, J.
(eds) Proceedings of the Tenth International Conference on Ground Penetrating Radar, Vol II,
pp439 – 442.
• Utsi, E. (2006) “Improving Definition: GPR Investigations at Westminster Abbey” in Daniels, J.J.
and Chen, C (eds) GPR 2006, Proceedings of the 11th International Conference on Ground
Penetrating Radar, June 19-22, Columbus, Ohio, USA.
21