Docstoc

Automatic Selection of GPR Wavelets in Soil Moisture Determination

Document Sample
Automatic Selection of GPR Wavelets in Soil Moisture Determination Powered By Docstoc
					        Automatic Selection of GPR Wavelets
      in Soil Moisture Determination Algorithm
                           Allen Lizarraga, Adolfo Torres
                            Advisor: Dr. Hamed Parsiani
                                   200


Ground-Based soil moisture         180

                                             1
determination has been more        160
                                                     2
                                                               3
actively researched in recent      140


years thanks to the non-           120
                                                                                   v

invasive nature of Ground-         100
                                                                   3
Penetrating Radar (GPR). In         80
                                                           2
                                                 1
the previous work, wavelets         60


were selected by visual             40


inspection.                              0           100                   200   300   400    500




                                   200


In this work, wavelet extraction   180


is successfully automated in       160
                                                           2           3
order to reduce user               140
                                             1

interaction. The C-language        120

algorithm (AQWE) functions         100                                     3
by detecting the boundaries in                   1
                                                               2
                                   80

the scan and extracting the
                                   60

wavelet based on several
                                   40

temporal criteria.                       0           100                   200   300    400    500
           Automatic Selection of GPR Wavelets
         in Soil Moisture Determination Algorithm

                                                   18.00000

                                                   16.00000

                                                   14.00000
Comparison of moisture




                             Moisture Percentage
                                                   12.00000

percentages for theta                              10.00000                                                        Theta Probe
                                                                                                                   Hand-cut
probe, hand-cut and                                 8.00000

                                                    6.00000
automatically extracted                             4.00000

results for different soil                          2.00000

samples clearly depict                              0.00000
                                                              1   2   3   4     5    6     7     8   9   10   11

accuracy improvement                                                          GPR Image Number


using the algorithm over                           18.00000

visual extraction.                                 16.00000

                                                   14.00000
                             Moisture Percentage




                                                   12.00000

                                                   10.00000                                                        Theta Pobe
                                                    8.00000                                                        AQWE

                                                    6.00000

                                                    4.00000

                                                    2.00000

                                                    0.00000
                                                              1   2   3   4     5     6    7     8   9   10   11
                                                                              GPR Image Number

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:14
posted:7/27/2011
language:English
pages:2