The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part Bl. Beijing 2008 
an area of about 15 km 2 which includes highways, railways, 
transformer substations, high voltage wires, water systems and 
residential areas. The mapping scale is 1:1000, with a designed 
flight altitude of 236 m and GSD 10cm . 
Considering the aircraft flying capacity as well as emergency 
disposal, the test adopted a relatively secure flight strategy, i.e., 
autonomous flight under both program control and remote 
control within the visual range. Aircraft is always within the 
operator’s visual range while carrying out the flying mission 
autonomously. Once any abnormal condition occurs, the 
operator can conveniently switch to remote-controlled mode 
and take over the flight of the aircraft. In order to do this, we 
divide the experimental area into 6x4 horizon grids. As 
shown in Figure 1 (a), one flight is made within one grid and it 
is made one by one, making sure that the aircraft is always with 
a visual range of 450m. 
Figure 1.Visual Grids and Route Planning 
After the division of visual grids, route planning is made within 
the grid, i.e., to determine the horizontal location of each 
exposure station and select the flight route of the plane. The 
horizontal position of each exposure station can be calculated 
by the picture format, longitudinal overlap (60% ± 5% ) and 
lateral overlap (30% ± 5% ). To minimize the wind impact on 
posture as well as to ensure consistency of image rotation angle, 
we made the equidirectional flying principles of shooting from 
west to east and that head and end of two adjacent routes is 
connected by the transition route, as is shown in Figure 1 (b). 
In each flight, the ground station carries out real-time 
monitoring of flight condition and flight path of the aircraft, and 
the pre-monitoring signals of digital camera is also transmitted 
via wireless link to the ground stations. At the pre-set exposure 
station, the haeundae remote control sends out attitude control 
data and camera control data to complete haeundae leveling 
adjustment, the camera’s exposure metering and shooting tasks. 
After the flight mission is completed, the aircraft returns to the 
vicinity of its take-off point automatically and switches to mode 
of manual control to implement landing. 
the largest rotation angle being 5°, the bending of the route 
being less than 1%, and the biggest aircraft height difference 
being 1 m. Figure 2 (a) shows the thumbnails of 384 original 
images in the surveyed area. 
The directional mapping adopts overall field control, under 
which a total number of 300 control points are evenly laid out. 
For accuracy measurement, 130 points are evenly selected as 
check points, using aerial triangulation to minor control. 
Precision statistical result shows as follows: Mx = 0A2m , 
My = 0.\0m , Mz = 0.15m , which meets 1:1000 topographic 
map surveying requirement. The subsequent photographic 
measuring produced and acquired 80 ortho images and vector 
diagrams. Figure 2 (b) shows the wall map of 1:5000 ortho 
image in the coverage. 
(a)Index image of the 
coverage 
(b)Orthoimage wall 
map 
Figure 2. Index image of the coverage and orthoimage wall map 
3. AERIAL PHOTOGRAMMETRY EXPERIMENT ON 
UNMANNED FIXED-WING AIRCRAFT 
3.1 Unmanned Fixed-wing Aircraft and Payload 
A number of aerial photogrammetry tests have been done on 
civilian unmanned fixed-wing aircraft. The document (CHEN 
Hongbin,2001) introduces an aerial remote sensing system 
based on a miniature autopilot plane, which is able to carry out 
fixed-point and fixed height shooting. The image resolution 
reaches decimeter level and it’s suitable for remote sensing 
surveys in unpopulated areas. The document (SUN Jie,2003) 
introduces a UAVRS-II UAV, the system of which takes plane 
array CCD digital cameras and single axes stable platform as 
airborne remote sensing equipment to acquire decimeter- 
resolution remote sensing images that meet the 1:10000 scale 
mapping application requirement. The document (YAN Lei, 
2004.) reports a UAV remote sensing systems which possesses 
man-machine performance as well as the functions of automatic 
taking off and landing. With a 30-hour biggest cruise duration, 
100 kg payloads, the aircraft-borne remote sensing equipment 
may choose SAR, infrared, visible light CCD imaging 
equipment, but as is stated frankly in the paper, this UAV needs 
to further improve its cost performance to become sustainable 
management of remote sensing platform. 
2.3 Test Results 
It took the flying crew of seven members six days of aerial 
photographing work to acquire a total of 384 images, each one 
being clear with uniform color without any aerial loopholes, 
and longitudinal overlap as well as lateral overlap are 100% 
qualified. All the indexes are in line with regulated 
requirements, with the largest photograph tilt angle being 2°, 
This paper holds the belief that there is much room for 
improving the cost performance of the fixed-wing unmanned 
aircraft aerial photogrammetry system. The technical 
approaches concerned are to combine with professional 
application, enabling the system’s performance to meet the 
requirements in the professional field in order to avoid the 
resources waste cost by the simplified civil use of high- 
performance military aircraft. For this reason, aiming at the