The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part B4. Beijing 2008
1342
3. WORK PROCESS
3.1 Exact center of the wall located
immediately to guarantee the quality. This work includes GPS
and IMU data preprocess to get trajectory, laser and image
coverage checking, image definition checking etc.
Because the corridor is narrow we have to locate the wall’s
exact center so to cover every part of the wall. We use 1:50000
relief map and the historical photography data to get the basic
location of the wall. At the same time we use GPS to get the
comers’ coordinates so as to locate the wall exactly.
3.2 Airborne laser system
In order to get enough laser points of the wall we design three
lines for LiDAR system that are one just above the wall, each
one for both sides they are 75m to 120m from the wall.
Followings are the parameters:
■ Altitude above ground = 300m(average elevation is
about 35 m)
■ Airplane speed = 150 km/h
■ Scanner angle = 45 degree
■ Laser frequence = 100,000hz
■ Focal length = 50mm
The point spacing is 10cm to 30cm. The pixel size on ground is
about 7cm.
There are two synchronized ground GPS receiver when flying,
which place at the control points of south-east comer of the wall
and south to Big-Stone bridge respectively. For the definition of
photo we choose clear day and check the picture quality when
work. For the covered west parts we acquire the data when sun
angle is wide enough and shadow is most small.
Figure 2: Laser data of the wall
3.3 Oblique picture taking
Figure 3: DEM of the wall
Figure 4: DOM (weng cheng)
3.6 DEM and DOM making
Firstly we classify laser data by LiDAR software and eliminate
the points which are not belong to surface. Then we make DEM
through surface laser points. The wall points are classify to
surface.
On the base of classified laser data we micro rectify each
pictures and mosaic them automatically then produce
orthophotos. According to requirement we do some mix work to
pictures’ color to get high quality DOM.
In order to get the side texture of the wall we design two lines
which are 350m inside and outside of the center of the wall. For
the west parts they are covered by mass trees we add one more
line which is 225m to the west of the wall.
Based on DEM and DOM data we can achieve 3D browse,
distance, acreage and volume measurement, section cutting and
making.
3.4 Terrestrial laser scan
The parts of wall which are covered by tress LiDAR and
oblique camera cannot their ideal textures. Therefore we use
terrestrial laser scanner to get them.
3.5 Data checking
After data acquiring we will do some preprocess work