This research presents the potential of using a 1-m resolution
LIDAR-based DEM in orthorectifying a single panchromatic Ikonos
image. First, the research investigates the use of the affine and
projective 2-D transformation models to rectify the panchromatic
Ikonos image. Results showed about 5m RMSE in the horizontal
direction using nine to 28 GCPs. Moreover, the results showed a
relationship between the elevations of the checkpoints and the
errors in their horizontal position. In order to overcome the effects
of the elevation differences in the rectification process, three 3-D
to 2-D transformation models were employed to orthorectify the
Ikonos image. The first two models are 1st order relief-corrected
and 2nd order relief-corrected transformation models. The third
model is the DLT transformation model. The three 3-D to 2-D
transformation models showed that less than 2m horizontal
RMSE could be achieved for the checkpoints. In addition, the
results showed that increasing the model complexity does not
reduce the checkpoints RMSE significantly. These results suggest
that the required planimetric accuracy for 1:5000 topographic
maps could be achieved using a 1-m resolution LIDAR-based DEM
and a single panchromatic Ikonos image. Future work will focus
on generating 1:5000 topographic maps using the proposed
technique. In addition, the assessment of using LIDAR-based DEM
to orthorectify multispectral Ikonos imagery will be investigated.
The results also suggest exploring the use of LIDAR-based DEM to
orthorectify other high-resolution satellite imagery such as that
from the QuickBird satellite.