red bar
University of Arkansas


Center for Advanced Spatial Technologies
red bar
     
Home

Research


Outreach


Geospatial Solutions/Spatial Data Distribution

Education & Training

Teaching & Research Facilities

Collaborators

Corporate Partners

Highlights

Publications

RGIS

ArkansasView

EAST Project


CAST Support

Inside: CAST

Contact

Site Map

Printer Friendly Page

.

..Center for Advanced Spatial Technologies

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Home > Streamlined Archaeo-Geophysical Data Processing and Integration for DoD Field Use > Technical Description > Data Processing Details

Technical Objectives | Data Processing Details | Processing Steps

The chart below illustrates the labor intensive character of data fusion using multiple COTS softwares. It documents the person hours per surveyed hectare required to process typical GPR data to a level matching other geophysical data layers for optimized information extraction. It should be emphasized that this estimate assumes that an experienced analyst is processing the data. Note that the time needed to process one hectare of GPR data greatly exceeds that needed for other geophysical data types, and this underscores the need for bringing GPR processing up to par with other geophysical processing routines. Many of the GPR routines are simple, yet are not streamlined in COTS software. GPR data processing is necessarily more time consuming owing to its three-dimensional nature and high data volume, but it can be greatly improved by the this effort.

TASK

I. Preprocessing GPR profiles (per ha)

HOURS (best, worst)
 
 

1. Edit marks

2. Distance normalization

3. Frequency high and low pass filtering

4. Gaining

5. Repeat frequency high and low pass filtering

6. Position correction

7. Background removal

8. Profile alignment

9. Velocity calculation

0

2

2

2

2

8

2

0

8

6

6

6

6

6

20

6

18

10

II. Enhancing GPR profiles (per ha)    
 

10. Stacking

11. Frequency filtering

12. Selective gaining

13. Hilbert transform

14. Migration

2

2

4

2

8

6

6

8

6

12

III. Slicing (per ha)    
 

15. Assemble 3D file/cube for each grid

16. Slice

2

2

6

12

IV. Image Preprocessing (per 1 ha slice or image):    
 

1. Mosaicing

2. Destriping

3. Destaggering

4. Despiking replace with mean, median, mode

5. Edge-matching by mean and variance

6. Desloping

1

1

1

0

1

1

4

2

4

1

2

2

V. Image enhancement (per 1 ha slice or image)    
 

7. High pass filter

8. Low pass filter

9. Interpolation

10. Sharpening

0

0

0

0

1

1

1

1

VI. Feature Extraction and fusion    
 

11. Image registration

12. Fusion

1

15

4

40

GPR Totals 65 197

The total for Tasks I-VI of 65-197 hours is for the processing of each GPR slice data set. Tasks IV-VI must also be applied to each of the five to ten GPR slices and to the other five data modalities. The combined totals for the five geophysical sensors would be the sum of the GPR processing (I-III), plus five to ten times steps IV-VI (depending on how many GPR slice maps to be processed). For example, the time needed to process one hectare each of magnetometry, electrical resistance, induced electrical conductivity, magnetic susceptibility, and six GPR slice maps would include 53 - 140 person hours for steps I-III plus 220 – 630 person hours for steps IV-VI, for a total of 273 – 770 person hours.

Note: These estimates also do not include the time spent importing and exporting from one software package to another, and the learning process involved for each.

 

 

Bottom image of campus, globe and map.

Center for Advanced Spatial Technologies
University of Arkansas
Ozark Hall, Room 12 Fayetteville AR 72701
Phone: (479)575-6159 | Fax: (479)575-5218 | Email: info@cast.uark.edu