6A-4-4
Figure 2. Full resolution image detail of a control point.
A set of 20 images with effective coverage of the region,
which contained survey control points, was identified.
These images were organized into a block consisting of
four flight lines, each containing five images. Two of the
flights were oriented north to south and were nearly
coincident. Similarly, the two south to north flight lines
were also nearly coincident. A total of 17 control points
were identified and measured within the block. The block
formation required the measurement of 51 additional tie-
points. The bundle adjustment resulted in positioning
accuracies averaging 1.23, 1.24, and 0.43 meters in X, Y,
and Z, respectively. The orientation accuracies average
0.079, 0.075, and 0.022 degrees in omega, phi, and kappa,
respectively. The relatively modest horizontal positioning
accuracy is due to the limitations in accurately identifying
the control points.
The EarthData LIDAR system has a scan FOV of 75
degrees. At 65 m/s airspeed, the along track spacing is
about 8 m, while the cross track spacing is roughly 6 m at
the maximum scan rate of 7.5 Hz and a 2500 m AGL
flying altitude. The illuminated footprint is 0.6 m. Typical
accuracies on the ground are 0.25-0.35 m in cross track,
0.2-0.25 m in along track and 0.15-0.25 m in height error.
The distribution of the LIDAR spots is shown in Figure 3.
Using the hierarchical warped image-based surface
reconstruction technique, DEMs were generated at a 5 m
grid - this was the closest to the average spacing of the
irregular LIDAR data. There was no editing of the
automatically extracted surfaces, although the densely
built residential areas caused substantial difficulty in the
processing. For performance examinations, DEMs were
extracted from different stereo pairs over the test area, and
the results showed only slight deviations. The
photogrammetrically derived DEM from the 4K by 4K
imagery is shown in Figure 4, followed by the LIDAR-
produced surface in Figure 5. Figures 6 and 7 depict the
ground truth, topographic surface and extracted objects,
measured by skilled operators on analytical instruments.
Figure 3. The distribution of the LIDAR observations
over the Hagerstown, MD test area, NE quadrant.
Figure 4. Photogrammetrically-derived DEM from the 4K
by 4K imagery.
Figure 5. LIDAR DEM observations over the
Hagerstown, MD test area.