**3.3. Data processing and a new amplitude-balancing approximation for transparent 3D half bird's-eye view of the GPR data set**

The GPR data, gathered within and on the northeast side of the tomb, were processed using REFLEXW software. After sequencing the profiles as discussed at the Anitkabir site, the starttime correction was applied. De-wow and background removals were applied. The secondorder band-pass Butterworth filter was then applied to the whole data set, to eliminate unwanted frequency noise. A simple linear gain function was applied as discussed in section 2.3. Velocity analysis indicated that the average velocity of electromagnetic wave propagation was 0.11m/ns. Finally, Kirchhoff migration was applied to the data.

**Figure 18. (a)** GPR data measurement on the northeast side of the Zeynel Bey tomb.

structures. During the control of the interactive depth slices, it is also important to control the colour of the amplitude range of the remains. It is known that the maximum amplitude ranges on an amplitude–colour scale represent subsurface features. However, the amplitude decreas‐ es with increasing recording time even if an appropriate time gain is applied to the profile

Transparent 2d/3d Half Bird's-Eye View of Ground Penetrating Radar Data Set in Archaeology and Cultural Heritage

http://dx.doi.org/10.5772/54998

129

It is also known that if the colours of maximum amplitude ranges are dominant at the beginning of the recording time or at depths very near to the surface, then the resolution of the slices can not be sufficiently high to differentiate buried structures very near to the surface, especially for small study areas as seen on Figure 19. Our third approximations after rearranged opacity function and viewing angle into the 3D volume, related to achieving amplitude balance according to the recording time or the depth (Figure 20). The aim of the amplitude balancing was to protect the colour range of the anomaly representing the target remains according to the time or depth range of the 3D sub-volumes; this was achieved by increasing or decreasing the maximum amplitude values of the amplitude–colour scale for the special depth slice or for a depth range. Therefore, the full data volume should be divided into

To balance the amplitude–colour scale in a depth range, firstly the changing colour range of the remains was controlled. Then, the amplitude range of the remains was determined in the 3D sub-volume and the maximum amplitude values of the colour range were reduced or increased in the depth range. The balancing of the amplitude–colour scale according to the amplitude range of the buried wall remnants produced better resolution and showed the remains with the same colour ranges on the slices or their 3D sub-volumes with increasing time or depth axis. The balancing procedure was also a time gain, which was a non-uniform stair function, weighting the electromagnetic wave field according to the time axis (Figure

20). The approximation was important to obtain a 3D representation of the volume.

**Figure 20. (a)** Re-scaling maximum amplitudes according to selected time or depth range of 3D GPR sub-volumes, and assigning the same colour range; **(b)** assigning a new opaque range according to the re-scaled amplitude–colour

data.

range.

sub-data volumes in time or depth.

**Figure 19.** Processed radargrams of the profiles along (a) the east–west direction and (b) the south–north direction inside the Zeynel Bey tomb.

The remains of a buried archaeological wall and foundational infrastructures can be defined on depth slices with location, and shapes according to depth. However, depth slices may not explain the subsurface if the area is small and complex, as in the Zeynel Bey tomb. Therefore, it is necessary to check the most meaningful depth slices and profiles to define the subsurface structures. During the control of the interactive depth slices, it is also important to control the colour of the amplitude range of the remains. It is known that the maximum amplitude ranges on an amplitude–colour scale represent subsurface features. However, the amplitude decreas‐ es with increasing recording time even if an appropriate time gain is applied to the profile data.

It is also known that if the colours of maximum amplitude ranges are dominant at the beginning of the recording time or at depths very near to the surface, then the resolution of the slices can not be sufficiently high to differentiate buried structures very near to the surface, especially for small study areas as seen on Figure 19. Our third approximations after rearranged opacity function and viewing angle into the 3D volume, related to achieving amplitude balance according to the recording time or the depth (Figure 20). The aim of the amplitude balancing was to protect the colour range of the anomaly representing the target remains according to the time or depth range of the 3D sub-volumes; this was achieved by increasing or decreasing the maximum amplitude values of the amplitude–colour scale for the special depth slice or for a depth range. Therefore, the full data volume should be divided into sub-data volumes in time or depth.

To balance the amplitude–colour scale in a depth range, firstly the changing colour range of the remains was controlled. Then, the amplitude range of the remains was determined in the 3D sub-volume and the maximum amplitude values of the colour range were reduced or increased in the depth range. The balancing of the amplitude–colour scale according to the amplitude range of the buried wall remnants produced better resolution and showed the remains with the same colour ranges on the slices or their 3D sub-volumes with increasing time or depth axis. The balancing procedure was also a time gain, which was a non-uniform stair function, weighting the electromagnetic wave field according to the time axis (Figure 20). The approximation was important to obtain a 3D representation of the volume.

**Figure 19.** Processed radargrams of the profiles along (a) the east–west direction and (b) the south–north direction

128 Imaging and Radioanalytical Techniques in Interdisciplinary Research - Fundamentals and Cutting Edge Applications

The remains of a buried archaeological wall and foundational infrastructures can be defined on depth slices with location, and shapes according to depth. However, depth slices may not explain the subsurface if the area is small and complex, as in the Zeynel Bey tomb. Therefore, it is necessary to check the most meaningful depth slices and profiles to define the subsurface

inside the Zeynel Bey tomb.

**Figure 20. (a)** Re-scaling maximum amplitudes according to selected time or depth range of 3D GPR sub-volumes, and assigning the same colour range; **(b)** assigning a new opaque range according to the re-scaled amplitude–colour range.

**Figure 21.** Transparent 3D half bird's-eye views of the GPR data set aligned in the south–north direction inside the Zeynel Bey Tomb between **(a)** 0 and 20 cm, **(b)** 20 and 40 cm, **(c)** 40 and 60 cm, **(d)** 60 and 80 cm, and **(e)** 80 and 100 cm depth ranges; and **(f)** infrastructures (orange colour) of the base and a cemetery (yellow colour) in the base wall inside the tomb.

**Figure 22.** Transparent 3D half bird's-eye views of the GPR data set aligned east–west inside the Zeynel Bey tomb at

Transparent 2d/3d Half Bird's-Eye View of Ground Penetrating Radar Data Set in Archaeology and Cultural Heritage

http://dx.doi.org/10.5772/54998

131

Practically, it was only possible to highlight the amplitude–colour ranges representing the remains, which were previously established as maximum amplitude–colour ranges of the 3D sub-data volumes. It was therefore straightforward to construct an opacity function by appointing opacity coefficients of one (maximum opacity) for maximum negative and

depths between (a) 90 and 120 cm, (b) 120 and 150 cm, and (c) 90 and 150 cm.

Transparent 2d/3d Half Bird's-Eye View of Ground Penetrating Radar Data Set in Archaeology and Cultural Heritage http://dx.doi.org/10.5772/54998 131

**Figure 22.** Transparent 3D half bird's-eye views of the GPR data set aligned east–west inside the Zeynel Bey tomb at depths between (a) 90 and 120 cm, (b) 120 and 150 cm, and (c) 90 and 150 cm.

Practically, it was only possible to highlight the amplitude–colour ranges representing the remains, which were previously established as maximum amplitude–colour ranges of the 3D sub-data volumes. It was therefore straightforward to construct an opacity function by appointing opacity coefficients of one (maximum opacity) for maximum negative and

**Figure 21.** Transparent 3D half bird's-eye views of the GPR data set aligned in the south–north direction inside the Zeynel Bey Tomb between **(a)** 0 and 20 cm, **(b)** 20 and 40 cm, **(c)** 40 and 60 cm, **(d)** 60 and 80 cm, and **(e)** 80 and 100 cm depth ranges; and **(f)** infrastructures (orange colour) of the base and a cemetery (yellow colour) in the base wall

130 Imaging and Radioanalytical Techniques in Interdisciplinary Research - Fundamentals and Cutting Edge Applications

inside the tomb.

maximum positive amplitude–colour ranges and zero (transparent) for unwanted amplitude– colour ranges on the amplitude–colour scale. A transparent 3D half bird's-eye view was obtained only by eliminating unwanted amplitude range to reveal subsurface remnants and features. However, depth range depended on the maximum amplitude volume. In order to construct a large depth range, the maximum amplitude–colour range would need to be restricted more than the normal range to produce a meaningful image. Figure 21 shows transparent 3D sub-data volumes with half bird's-eye view between 0 and 20cm, 20 and 40 cm, 60 and 80 cm, and 80 and 100cm depth ranges from the GPR data set aligned in the south– north direction inside the tomb. Wall structures and a cemetery with stair could be clearly seen. In addition, Figure 22 shows transparent 3D sub-data volumes between 90–120 cm. 120–150 cm and 90–150cm depth ranges along the east–west direction inside the tomb. The visualiza‐ tion results of the east–west side data set supported those of the south–north. However, the results of the south–west side perfectly imaged the foundational infrastructure and buried cemetery.

Although the maximum diameter inside the tomb was 4 m, it would be possible for users to envisage buried archaeological remains when viewed using the transparent 3D image. Figure 23 shows the excavation results inside the tomb. The excavation findings perfectly matched the transparent 3D half bird's-eye view. The visualization results also precisely indicated all

Transparent 2d/3d Half Bird's-Eye View of Ground Penetrating Radar Data Set in Archaeology and Cultural Heritage

http://dx.doi.org/10.5772/54998

133

**Figure 24.** Transparent 3D half bird's-eye views of the GPR data set on the northeast side of the Zeynel Bey tomb, and pictured buried wall remains in depth between **(a)** 50 and 100 cm, **(b)** 75 and 150 cm; **(c)** 75 and 150 cm but with

Half bird's-eye views of the transparent 3D depth volume ranges of the GPR data were also produced for the northeast side of the tomb (Figure 24). The wall structures were seen exactly in the corresponding transparent 3D imaging. The results showed that the buried walls were very near the surface and very complex. These imaging results remain to be confirmed through direct observation, as excavation has not yet taken place at these sites. The results could be confirmed by comparing the excavations inside of the tomb and their

different viewing angles of the x, y and z axes.

3D visualization results.

details of the foundational wall remains and the cemetery with stairs.

**Figure 23. (a)** The North side of the Zeynel Bey tomb, **(b)** the excavation on (a) and a cemetery, (c) the walls of the cemetery and stairway.

Although the maximum diameter inside the tomb was 4 m, it would be possible for users to envisage buried archaeological remains when viewed using the transparent 3D image. Figure 23 shows the excavation results inside the tomb. The excavation findings perfectly matched the transparent 3D half bird's-eye view. The visualization results also precisely indicated all details of the foundational wall remains and the cemetery with stairs.

maximum positive amplitude–colour ranges and zero (transparent) for unwanted amplitude– colour ranges on the amplitude–colour scale. A transparent 3D half bird's-eye view was obtained only by eliminating unwanted amplitude range to reveal subsurface remnants and features. However, depth range depended on the maximum amplitude volume. In order to construct a large depth range, the maximum amplitude–colour range would need to be restricted more than the normal range to produce a meaningful image. Figure 21 shows transparent 3D sub-data volumes with half bird's-eye view between 0 and 20cm, 20 and 40 cm, 60 and 80 cm, and 80 and 100cm depth ranges from the GPR data set aligned in the south– north direction inside the tomb. Wall structures and a cemetery with stair could be clearly seen. In addition, Figure 22 shows transparent 3D sub-data volumes between 90–120 cm. 120–150 cm and 90–150cm depth ranges along the east–west direction inside the tomb. The visualiza‐ tion results of the east–west side data set supported those of the south–north. However, the results of the south–west side perfectly imaged the foundational infrastructure and buried

132 Imaging and Radioanalytical Techniques in Interdisciplinary Research - Fundamentals and Cutting Edge Applications

**Figure 23. (a)** The North side of the Zeynel Bey tomb, **(b)** the excavation on (a) and a cemetery, (c) the walls of the

cemetery.

cemetery and stairway.

**Figure 24.** Transparent 3D half bird's-eye views of the GPR data set on the northeast side of the Zeynel Bey tomb, and pictured buried wall remains in depth between **(a)** 50 and 100 cm, **(b)** 75 and 150 cm; **(c)** 75 and 150 cm but with different viewing angles of the x, y and z axes.

Half bird's-eye views of the transparent 3D depth volume ranges of the GPR data were also produced for the northeast side of the tomb (Figure 24). The wall structures were seen exactly in the corresponding transparent 3D imaging. The results showed that the buried walls were very near the surface and very complex. These imaging results remain to be confirmed through direct observation, as excavation has not yet taken place at these sites. The results could be confirmed by comparing the excavations inside of the tomb and their 3D visualization results.
