1243
THE ERROR ANALYSIS AND CORRECTION METHOD RESEARCH OF THE
ATTITUDE DATA FOR THE UAV REMOTE SENSING IMAGES
Hongying Zhao a> *, Yuanchen Qi b
a Institute of Remote Sensing & Geographic Information System, Peking University, Beijing 100871, China -
zhaohy@pku.edu.cn
b College of Surveying and Geographical Science, Liaoning Technical University, Fuxin 123000, China -
qiyuanchen649@ 163 .com
KEY WORDS: GPS/INS, Navigation, Photogrammetry, Calibration, Error, IMU
ABSTRACT:
As an important way to obtain the high-resolution remote sensing images, the unmanned aerial vehicle (UAV) aviation remote
sensing plays a more and more important role in the area of photogrammetry and remote sensing application. The correction of the
UAV remote sensing images without the ground control points needs high-resolution attitude data. The original attitude data can be
obtained from the UAV airborne GPS/INS. The attitude data error analysis and correction method is proposed, and the
corresponding error calibration mode is established in this paper. The images of the UAV aerial photography is used to verify the
effectiveness and feasibility of this model, and the experimental results show that the UAV aerial remote sensing images can realize
high-precision correction without the ground control points.
1. INTRODUCTION
With the rapid development of civil remote sensing recently,
the demand for remote sensing image is increasing, especially
large-scale and high-resolution remote sensing images. At
present, the main information acquisition platforms of the world
are still satellite and manned plane. The shortcomings of the
satellite remote sensing platform are high price and long return
cycle, so when high-resolution images are got, the update speed
of the images is slow. While the manned-plane remote sensing
system is mainly limited by rising-landing and safety condition,
it is unable to satisfy the users’ requests about safety. As a new
and effective type of earth observation system, the UAV remote
sensing information platform is proposed to obtain remote
sensing images at present. Its main characteristics are as follows:
1. Without considering human factors by auto-control flight; 2.
The flight mode is flexible and have long flight time; 3. High
flight trajectory precision; 4. Flying under the clouds. These
characteristics enable the UAV aviation remote sensing
information platform to become an effective supplement way to
the satellite and manned-plane remote sensing
UAV can realize auto-control flight as an aviation remote
sensing platform, and the position and attitude data of the
airplane can be obtained directly through the airborne GPS/INS
integrated navigation system. But if we regard this group of
posture data as the corrected exterior orientation elements for
the images, it must carry on coordinate transformation (Naci
Yastikli, 2005) and systematic error compensation in order to
convert the aircraft attitude data obtained from inertial
navigation system to the high-precision exterior orientation
elements for the images. This systematic error mainly refers to
the spatial displacements and the deviation angle error. Though
the spatial displacement between the GPS antenna phase centre
and the camera projection center is easy to get, the deviation
angle error can’t be determined directly through the traditional
method. This error seriously affects the correction precision of
the UAV remote sensing images.
At present, in aerial photogrammetry assisted by the manned-
plane airborne GPS/INS, two calibration methods are generally
used to correct the deviation angle error. One way is that the
ground control points is firstly laid out in the survey area, then
each photo's exterior orientation elements directly obtained
from GPS/INS as weighted observation value participate in the
photogrammetry block adjustment. So we can obtain higher
precision exterior orientation elements for the images; another
method is that the region which includes two (or more) flight
strips in or near the survey area as calibration field is firstly
selected, then the exterior orientation element of each aerial
photo is exactly determined through the method of aerial
triangulation. After that, matrix operation is carried out using
the exterior orientation elements computed above and
observation values including deviation angle error obtained
from the GPS/INS system, then the deviation angle error can be
got. This way is called direct orientation method. The first way
needs to lay out ground control points in the whole operation
range, which not only restricts operation range, but also
decreases work efficiency and enlarges the cost of operation.
Based on the second way, according to the UAV’s operation
standard and corresponding instrument indexes, in this paper
the attitude data correction mathematical model of the UAV
aerial remote sensing image is established and the right
deviation angle error is calculated. So the original aerial attitude
data obtained from the GPS/INS system can be transformed to
the images’ exterior orientation elements which are needed for
direct georeferencing in aerial photogrammetry.
* Corresponding author.