light weight rotary-wing uav for large … aniqah mohd azhar.pdf2.0 application uav in mapping –...
TRANSCRIPT
Asia Geospatial Forum, 24-26 September 2013, Kuala Lumpur, Malaysia
LIGHT WEIGHT ROTARY-WING UAV FOR LARGE SCALE MAPPING
APPLICATIONS
1Norhadija Darwin,
1Nurul Farhah Abdul Hamid,
2Wani Sofia Udin
1Noor Aniqah Binti Mohd
Azhar & 3Anuar Ahmad
1Department of Geoinformation,
Faculty of Geoinformation & Real Estate
Universiti Teknologi Malaysia,
81310 Skudai, Johor, MALAYSIA
2Institute for Science and Technology Geospatial (INSTEG),
Faculty of Geoinformation & Real Estate
Universiti Teknologi Malaysia,
81310 Skudai, Johor, MALAYSIA
3Institute for Science and Technology Geospatial (INSTEG),
Faculty of Geoinformation & Real Estate
Universiti Teknologi Malaysia,
81310 Skudai, Johor, MALAYSIA
ABSTRACT
This paper aims to demonstrate the potential use of unmanned aerial vehicle (UAV) system attached with
calibrated high resolution digital based on a simulation model. In this study, a strip of aerial images was
captured using a calibrated high resolution compact digital camera known as Canon Power Shot SX230
HS and it has 12 megapixel image resolutions. The digital camera was calibrated in the laboratory and
field. For laboratory calibration, a 3D test field in form of calibration plate was used. The dimension of
the calibration plate is 0.4m x 0.4m and consists of 36 grid targets at different heights. For field
calibration, a 3D test field has been constructed which comprise of 81 target points at different heights
and located on a flat ground with dimension of 9m x 9m. The light weight UAV can be used in various
applications such as coastal, archeological and meandering. The UAV equipped with an autopilot system
and automatic method known as autonomous flying, can be utilized for rapid and low cost data
acquisition. In this study, the UAV system has been employed to acquire aerial images of a simulation
model at low altitude. From the aerial images, photogrammetric image processing method is completed to
produce mapping outputs such a digital terrain model (DTM), contour line and orthophoto. In term of the
accuracy, of measurement, a milimeter-level is reached by ground control point (GCP) and check point
(CP) using conventional ground surveying method (i.e total station). It will anticipate that the UAV will
be used for surveying and guideline with good accuracy. Finally, the UAV has shown great potential and
produce accurate results or products using high resolution camera calibration.
Keywords: UAV; high resolution; environmental survey
Asia Geospatial Forum, 24-26 September 2013, Kuala Lumpur, Malaysia
1.0 Introduction
Basically, there are several methods in geoinformation that can be used to map the environmental sites
such as aerial photogrammetry, remote sensing, LIDAR (Light Detection and Ranging), GPS (Global
Positioning system), TLS (Terrestrial Laser Scanning) and total station. The geoinformation technology
could also be used in environmental survey and could assist the developers for societal impact in the
developing country. The remote sensing and aerial photogrammetry is widely used for mapping
environmental sites. For remote sensing, with the existing of high resolution satellite imagery such as
Ikonos, QuickBird and WorldView 2, it can be used for environmental survey. On the other hand, the
development of remote sensing technology where the satellites can capture high-resolution imagery with
the capability of producing stereo imagery using IKONOS satellite images (Li et al., 2004).
However, there are some limitations or draw back for these methods. The problem related to this
technology is the difficulties of possessing clear image of the study area. According to Biesemans et al.,
(2005) and Everaercts (2008), the limitation of satellites and manned aircraft are flight costs, slow and
weather-dependent data collection, limited availability, limited flying time, low ground resolution. In
aerial photogrammetry, the aircraft can be flown under the cloud and imagery can be obtained much
easier than satellite imagery.
There are few ways to produce map by photogrammetric technique which are by Analog Photogrammetry
(from about 1900 to 1960) and Analytical Photogrammetry (1960 until present). However, the presence of
Digital photogrammetry in the photogrammetric industry has revolutionized the industry. Nowadays,
most countries in the world have produced their topographic map using aerial photogrammetry. Recently,
digital photogrammetry has embraced UAV technology known as UAV photogrammetry. According to
Eisenbeiss (2009), UAV photogrammetry can be understood as a new photogrammetric measurement
tool. UAV photogrammetry opens various new applications in the close range domain, combining aerial
and terrestrial photogrammetry, and also introduces low-cost alternatives to the classical manned aerial
photogrammetry. UAVs have been under development since the beginning of flight. For the first
development of UAV, it was introduced for military purpose only. According to Pardesi (2005), during
World War I, the US decided to make a contribution in the novel area of the flying bomb. The most
important development for unmanned aviation during the interwar years was radio control.
UAV system has been used to produce digital map and orthophoto of UTM Johor Bahru (Anuar Ahmad,
2011; Anuar Ahmad & Wan Aziz Wan Mohd Akib, 2010; Anuar Ahmad, 2009a, 2009b). In the study
carried out, fixed wing UAV was used to acquire the digital aerial photograph at low altitude of
approximately 300m. The output of the study showed that the digital map was produced at large scale and
accurate. Therefore, UAV system has expanded data capture opportunities for photogrammetry
techniques. Usually, the UAV system uses the concepts of close range photogrammetry (CRP). In CRP,
the photography is acquired where the object-to-camera distance is less than 300m (Cooper and Robson,
1996; Wolf and Dewitt, 2000). Moreover, Baoping et al., (2008) stated that numerous UAV had been
developed by organization or individual worldwide including a complete set of UAV which used high
quality fibers as material for plane model. The development of this technology is very beneficial for
monitoring purpose for limited time and budget. It is supported that UAV has been practiced in many
applications such as farming, surveillance, road maintenance, recording and documentation of cultural
heritage (Bryson and Sukkarieh, 2009).
In this study, two main hardwares are used which comprise of light weight rotary-wing UAV and high
resolution digital camera. Low altitude UAV is preferable in this study because it focuses on simulation
model which covered small area only. The compact digital camera provides small format images. Figure
1 show examples of UAV known as Hexacopter and compact digital camera used in this study.
Asia Geospatial Forum, 24-26 September 2013, Kuala Lumpur, Malaysia
In this study Canon Power Shot SX3 digital camera has been used in acquiring simulation model images.
This digital camera has 14x optical zoom lens and 2.0” LCD screen. Table 1 depicts the compact digital
camera specification.
Table 1: Canon PowerShot XS230 HS digital camera specifications
Specification
Maximum Resolution 4000 x 3000 pixels
Effective pixels 12.10 megapixels
Lens 14.00x zoom, f3.1-5.9, 28-392mm (35mm equivalent)
LCD size 3”
Sensor size 1/2..3”, 460K dots/None
Sensor type CCD
Dimensions 4.2 x 2.4 x 1.3 in. (106 x 62 x 33 mm)
Weight (Body) 218g includes batteries
Shutter 15-1/3200
ISO 100-3200
Memory type SD/SDHC
File formats JPEG (conforms to Exif 2.2), conforms to DCF2.0, DPOF,
PRINT Image Matching III, AVI (Motion JPEG), with
WAV (PCM), mono
In this study, micro UAV known as Hexacopter (Figure 1) has been used in acquiring images of the
simulation model. The specification of the rotary wing used in this study is shown in Table 2.
Table 2: Hexacopter Specification
Specification
Weight 1.2kg
Rotor 6 rotor
Endurance Up to 36 minutes
Payload 1kg
GPS on board Yes
Special function Automatically return to home location (1st point)
Stabilizer Inbuilt stabilizer to deal with wind correction
Capture data Using software to reached waypoints
Flight control Manual and autonomous
Camera stand Flexible camera holder
a b
Figure 1: (a) Hexacopter; (b) Digital Camera
Asia Geospatial Forum, 24-26 September 2013, Kuala Lumpur, Malaysia
2.0 Application UAV in mapping – An Experimental
The simulation was developed with three models which comprises of simulation of coastal area,
archaeological site and meandering flume using high calibrated digital camera. In this study, a
rotary wing UAV was employed to acquire aerial images of a simulation model of environmental
area. The dimension of this simulated model is 2.4m x 7.2m for coastal area and 2.4m x 3.5m for
archaeological site. The dimension of meandering flume is 12.0 × 3.0m and the channel width is
0.5m. This laboratory flow channel was attempted to replicate physical structures such as
meandering streams found in the real world. The following section discusses the experimental
conducted in this study.
2.1 Model of Coastal Area
Digital images were acquired using the Canon PowerShot SX3 HS digital camera. It has wide angle lens
for acquiring of aerial photography at the height of 3 meter. The scale of simulation model is 1:500 for
dimensions 2.4m x 7.2m. There are 37 ground control points (GCPs) and 7 check points (CPs). Figure 1
depicts the simulation model of coastal area.
Figure 2: Simulation Model of Coastal Area
2.2 Model of Archaeological Site
The dimension of the archaeological site simulation model is 2.4m x 3.5m. It consists of sand, some
broken porcelains, and some artificial bones and 35 GCPs and 6 CPs for ground control point.
Asia Geospatial Forum, 24-26 September 2013, Kuala Lumpur, Malaysia
Figure 3: Simulation Model of Archeological Site
2.3 Model of Riverbed Topography
A laboratory flow channel or meandering flume is located at the Universiti Teknologi Malaysia.
The dimension of this flume is 12.0 × 3.0m and the channel width is 0.5m. Photogrammetric
control targets were established on the flood plain and inside the channel bed. Ninety (90) GCPs
were registered as a full control (XYZ) and 31 check points (CPs) were established evenly in the
channel bed. The targets must be placed on the meandering flume and maintained until image
acquisition is completed. The distribution of the GCPs is flexible and they need to be seen on a
pair of photograph (i.e stereopair) at known location. Figure 3 represents where the control
points were placed throughout the area of interest. These targets were 10 mm in diameter and of
red and black design
Figure 4: Simulation Model of Riverbed Topography
Asia Geospatial Forum, 24-26 September 2013, Kuala Lumpur, Malaysia
3.0 Methodology
Figure 5: Flowchart of Research Methodology
Asia Geospatial Forum, 24-26 September 2013, Kuala Lumpur, Malaysia
3.1 Flight Planning
Since the development UAV is utilizing autonomous flight including two main stations known as ground
control station (GCS) and pilot station (PS). Generally, flight planning shows a flight map which consists
of waypoints on a topographic map showing the starting and ending points of each line. Flight planning
encompasses calculation of boundary of the study area, number of strips required, pixel size, photo scale,
flying height and percentage of end lap and side lap. The aerial photographs should overlap at least 60
percent and side lap at least 30 percent. Before any observation starts, first need is to properly install the
instruments. There are few important components of the system in autonomous fly that should be
installed as shown in Figure 6.
Figure 6: Instrument Installation Diagram
3.2 Data Acquisition
The aerial images acquisition is carried out using the hexacopter. In this experiment, for the
hexacopter was equipped with a Canon PowerShot SX 230 HS digital camera in acquiring the simulation
model images. The digital camera was used to acquire photographs of the simulation model at a constant
distance of 3 meter from the model. The hexacopter uses the autonomous flight control system and
controlled by two operators where one act as pilot on ground and the other in charge of monitoring flight
mission at ground station. The establishment of ground control point (GCP) and check point (CP) was
performed after the acquisition of aerial image. The 3D coordinates of these GCPs and CPs were
determined using total station. Figure 7 represents how the light weight rotary wing UAV and Ground
Control Station were carried out.
Asia Geospatial Forum, 24-26 September 2013, Kuala Lumpur, Malaysia
3.3 Digital Camera Calibration
Digital camera calibration becomes essential to achieve the precision of the measurement task.
However, when talk about non-metric digital camera, there are consideration must be aware the internal
especially geometry camera instability. Camera parameters usually could be recovered through camera
calibration process which comprised of focal length (c), principal point offset (xp, yp) which represent the
coordinates of the center of the image, radial lens distortion (k1, k2, k3), tangential lens distortion (p1, p2)
and others. Nowadays, many camera calibration approaches are presented. However, in this study the
automatic self-calibration bundle adjustment are adopted.
In this study, two camera calibration methods are used. The first method is a lab calibration which
comprise of 3D calibration plate with a dimension of 0.4 x 0.4 meter. A bar scale of length 553 mm is
used too. A second camera calibration is performed on the field where the 3D test field was used with a
dimension of 9 x 9 meter. This test field comprised of 81 wooden pegs located into the ground as
illustrated as Figure 8. The size of each wooden peg is 2 x 3 inches. All the wooden pegs are at different
height and the 3D coordinates of these wooden pegs were determined using close traverse. Three scale
Aircraft accordance
to waypoint &
within 5 seconds
captured the images
Ground Control
Station (GCS)
Aircraft is flying
Control by
Connect both
Radio Controller
used by Pilot Landing
Set Coming
Home (CH)
Communication
Set by Flight
Planning
Net
Figure 7: Cropcam UAV and Ground Control Station
Asia Geospatial Forum, 24-26 September 2013, Kuala Lumpur, Malaysia
bars were used where the length of 332 mm, 582 mm, and 1178 mm. Calibration site was built near the
Faculty Geoinformation and Real Estate (FGRE), Universiti Teknologi Malaysia.
Figure 8: Three dimensional test field 9 x 9 meter.
3.3.1 Laboratory Calibration
Three digital cameras was setup at different configurations (convergent, generic network and
stereo) and different heights at 80cm, 100cm and 120cm. All camera configurations, images of calibration
plate were taken at each position in landscape position of 0 degree and in portrait position of 90 degrees.
The light rays from the camera station are pointing towards center of calibration plate. For convergent and
stereo configuration, eight images were taken from four stations around the calibration plate while for
generic network configuration sixteen images were taken from eight stations around the calibration plate
at different height per dataset as illustrated as Figure 9.
After images of the calibration plate were acquired, these images were downloaded into a
computer for data processing and analyzed using Australis software. As standard procedure of camera
calibration, the results comprised of eight camera calibration parameters which include focal length (c),
principal point (xp, yp), radial distortions (k1, k2, k3) and tangential distortions (p1, p2).
Asia Geospatial Forum, 24-26 September 2013, Kuala Lumpur, Malaysia
Figure 9: Image acquisition from eight camera station.
3.3.2 Field Calibration
Field camera calibration site is located near the Faculty of Geoinformation and Real Estate
(FGRE), Universiti Teknologi Malaysia (UTM) as shown as Figure 10. For the field calibration, a test
field area with dimension of 9 x 9 m and 81 target points was established. The 3D coordinates of each
target of the test field was determined based on close traverse around the test field using total station.
Image acquisition is divided into two parts based on convergent and stereo configuration. For the
convergent case, the UAV was flown at the height of approximately 5 m while for the flying height is
stereo 20 m.
Figure 10: Field calibration site (red box) located near FGRE building.
9 m
9 m
Asia Geospatial Forum, 24-26 September 2013, Kuala Lumpur, Malaysia
The UAV was flown manually, due to the test field is near to the building. A total of 32 images
for camera configuration in convergent and 28 images of the stereo camera configuration were acquired.
Eight images per camera configuration were chosen for image processing. The field calibration process
was performed similar to laboratory calibration using Australis software.
3.4 Preliminary Results and Analysis for Digital Camera Calibration
In this section, the results of measurement for different camera configuration setup, camera
elevation and different calibration methods which are laboratory calibration and field calibration using
Canon PowerShot SX230 HS are briefly discussed. After the image processing, camera calibration
parameters were obtained from the camera calibration software which utilizes self-calibration bundle
adjustment. The results for the different camera setup, different camera elevation and different methods
are tabulated in the following sections.
3.4.1 Camera Configuration Setup versus Camera Elevation
Table 3, 4 and 5 show the mean and standard deviation for the camera calibration parameters for
the three camera configurations and three camera elevations respectively. The results of camera
calibration which utilized camera configuration setup at the position of 80 cm height (Table 3) showed
that the lowest standard deviation for focal length is ±0.00567mm achieved by the generic network
configuration. While the lowest and best standard deviation of xp and yp, is ±0.005612mm and
±0.005497mm respectively for generic network configuration.
For the rest of camera calibration parameters, the standard deviations are very small and close to
one another. For the case of stereo camera configuration for every camera elevation, the photogrammetric
calibration software failed to process the image due to weak geometry which means the results depend on
the configuration position of the camera and the angle between the cameras. The smaller the angle, the
less will be the accuracy of the result. On the other hand, for aerial photogrammetry normally height-base
ratio A/B, is employed. Based on this configuration, the higher accuracy could be achieved when the
intersection angle is near 90˚ and also other constraints must be considered.
Table 3. Camera calibration parameters for camera configuration setup at 80 cm height.
Camera
Calibration
Parameters
Camera Configuration Setup at 80cm Elevation
Convergent
(Mean) Std. Dev.
Generic
Network
(Mean)
Std. Dev. Stereo
(Mean) Std. Dev.
c (mm) 5.105 ± 0.015185 5.099660 0.005668 Failed Failed
xp (mm) -0.040 ± 0.008704 -0.040620 0.005612 Failed Failed
yp (mm) -0.019 ± 0.011944 -0.023320 0.005497 Failed Failed
k1 0.002 ± 0.000310 0.001795 0.000264 Failed Failed
k2 0.000 ± 0.000224 -0.000179 0.000390 Failed Failed
Asia Geospatial Forum, 24-26 September 2013, Kuala Lumpur, Malaysia
k3 0.000 ± 0.000031 0.000030 0.000248 Failed Failed
p1 0.001 ± 0.000115 0.000500 0.000041 Failed Failed
p2 0.001 ± 0.000151 0.000615 0.000065 Failed Failed
Table 4. Camera calibration parameters for camera configuration setup at 100 cm height.
Camera
Calibration
Parameters
Camera Configuration Setup at 100cm Elevation
Convergent
(Mean) Std. Dev.
Generic
Network
(Mean)
Std. Dev. Stereo
(Mean) Std. Dev.
c (mm) 5.096 ± 0.006751 5.099 ± 0.006838 Failed Failed
xp (mm) -0.055 ± 0.005864 -0.044 ± 0.003905 Failed Failed
yp (mm) -0.026 ± 0.006116 -0.022 ± 0.004072 Failed Failed
k1 0.002 ± 0.001103 0.002 ± 0.000264 Failed Failed
k2 0.000 ± 0.001157 0.000 ± 0.000175 Failed Failed
k3 0.000 ± 0.000350 0.000 ± 0.000029 Failed Failed
p1 0.000 ± 0.000065 0.001 ± 0.000052 Failed Failed
p2 0.001 ± 0.000095 0.001 ± 0.000041 Failed Failed
The results of camera calibration which utilizes camera configuration setup at the position of
100cm height are shown in Table 4. In this table, the lowest standard deviation for focal length is
±0.006751mm achieved by the convergent configuration. While for the principal point offset coordinates
xp and yp, is ±0.003905mm and ±0.004072mm respectively for generic network configuration. For the
remaining results of camera calibration, the differences in standard deviation are small.
In Table 5 shows the results of camera calibration which utilizes camera configuration setup at
the position of 120cm height. The performance of generic network configuration is still better than other
camera configuration where the standard deviation for focal length is ±0.006694mm. While, for the
principal point offset coordinates xp and yp is ±0.011368mm and ±0.004021mm respectively which are
better than convergent configuration. Once again the camera calibration results indicate that generic
network configuration is the most efficient camera configuration for camera calibration.
Table 5. Camera calibration parameters for camera configuration setup at 120 cm height.
Camera
Calibration
Parameters
Camera Configuration Setup at 120 cm Elevation
Convergent
(Mean) Std. Dev.
Generic
Network
(Mean)
Std. Dev. Stereo
(Mean)
Std.
Dev.
c (mm) 5.107 ± 0.011803 5.105 ± 0.006694 Failed Failed
xp (mm) -0.046 ± 0.016438 -0.046 ± 0.011368 Failed Failed
yp (mm) -0.008 ± 0.007599 -0.009 ± 0.004021 Failed Failed
k1 0.002 ± 0.000430 0.002 ± 0.000180 Failed Failed
k2 0.000 ± 0.000280 0.000 ± 0.000182 Failed Failed
Asia Geospatial Forum, 24-26 September 2013, Kuala Lumpur, Malaysia
k3 0.000 ± 0.000113 0.000 ± 0.000051 Failed Failed
p1 0.001 ± 0.000141 0.001 ± 0.000111 Failed Failed
p2 0.000 ± 0.000112 0.000 ± 0.000051 Failed Failed
3.5.2 Laboratory Calibration versus Field Calibration
Table 6 shows the camera calibration parameters and standard deviation for laboratory calibration
and field calibration. The results of the laboratory calibration and field calibration showed slight
difference of standard deviation for both methods. Similarly, for radial lens distortion and tangential lens
distortion the standard deviation for both methods showed slight difference. Based on the results of both
methods, the field camera calibration method is reliable and significant in calibrating non-metric digital
camera.
Table 6. Camera calibration parameters for laboratory calibration and field calibration
Camera
Calibration
Parameter
Two Camera Calibration Methods
Lab Calibration Standard
Deviation Field Calibration
Standard
Deviation
c (mm) 5.095 ± 6.894e-003 5.116 ± 1.066e-002
xp (mm) -0.035 ± 5.997e-003 -0.028 ± 9.176e-003
yp (mm) -0.020 ± 5.470e-003 -0.026 ± 1.317e-002
k1 0.002 ± 4.666e-004 0.001 ± 9.189e-005
k2 -0.001 ± 4.164e-004 0.000 ± 1.850e-005
k3 0.000 ± 1.040e-004 0.000 ± 1.145e-006
p1 0.000 ± 9.747e-005 0.001 ± 9.641e-005
p2 0.001 ± 1.295e-004 0.001 ± 8.528e-005
In photogrammetric application especially for close range photogrammetry, both convergent and
generic network configurations are widely used. In general, it is found that the standard deviation of focal
length improve well as the height increases. For the other camera calibration parameters, the standard
deviations are very small, minimum and close to zero value. For generic network configuration, it
produces better result compared to convergent configuration with reference to the standard deviation of
focal length as shown in Table 5. The results also showed that as the height of the camera increases the
standard deviation decreases as shown in Table 3, 4 and 5. For stereo configuration, the result showed that
this configuration are not suitable for camera calibration.
In this study, it is clearly shown that the field calibration has the advantage that the images were
taken under similar conditions to the images taken using UAV. That has proved be very efficient and
provides accurate results for the purpose of camera calibration. Finally, the field calibration can be
employed for obtaining good measurement and results.
Asia Geospatial Forum, 24-26 September 2013, Kuala Lumpur, Malaysia
4.0 Image Processing and Results
All images were processed using digital photogrammetric software. The process comprises of interior
orientation, which requires the camera calibration parameter (Table 5) and exterior orientation which
require the registration of GCPs and auto generation of tie points. The flow chart provides clear picture on
how the processing was performed (Figure 11).
Figure 11: Process of performing Aerial Triangulation
Each pair of photographs has 60 percent overlapped and eight (8) photographs for coastal area
and two (2) photographs for archeological site. Based on the image processing, there are 267
points, which consists of 37 GCPs and 230 tie points for simulation of coastal area. While for
simulation model of archeological site, there are 57 points which consists of 35 GCPs and 22 tie
points. After performing Aerial Triangulation (AT), the footprint of the AT can be displayed. The
foot prints of the digital photographs showing the location and names of all points (i.e control
points, check points and tie points) that participated in the adjustment. The distribution of GCP
and CP for the digital camera in riverbed topography can be viewed in Figure 12, Figure 13 and
Figure 14 respectively.
Figure 12: Footprint for simulation of coastal
area
Figure 13: Footprint for simulation of
archaeological site
Asia Geospatial Forum, 24-26 September 2013, Kuala Lumpur, Malaysia
Figure 14: Footprint for Simulation of Riverbed topography
There were two main results were produced i.e. DTM and orthophoto. The generated digital
orthophoto for the simulated model for coastal area, archaeological site and meandering flume
are shown in Figure 15, 16 and 17 respectively. The quality of orthophoto depends on the high
resolution camera calibration and quality of GCPs produced.
Figure 15: Orthophoto for Simulation of Coastal Area
Figure 16: Orthphoto for Simulation of Archeological Site
Asia Geospatial Forum, 24-26 September 2013, Kuala Lumpur, Malaysia
Figure 17: Orthophoto for Simulation of Riverbed topography
5.0 Analysis and Discussion
Nowadays with the development of digital camera, analysis can be carried out for the small format digital
camera. A small format photograph from digital camera has the potential to be used in aerial
photogrammetry and analysis can be carried out for the product of aerial photogrammetry such as
orthophoto, DTM, contour line and digital map.
For point analysis, the RMSE was below one (1) meter that indicates the orthophoto has sub-meter
accuracy. The smaller the RMSE, the better orthophoto could be produced. It can be concluded that the
higher the GCPs was, the better the RMSE. Table 7 shows the comparison of check points between
coordinates from ground survey (i.e. total station) and coordinates obtained from ERDAS Imagine for
coastal area, where the calculated RMSE is ± 0.004, ± 0.006 and ± 0.002 meter (<1 meter) for coordinate
x, y and z respectively.
Table 7: Comparison check points for coastal area
Check
Points
Total Station Erdas Imagine Software Differences
X Y Z X Y Z ΔX ΔY ΔZ
1039 10013.262 9993.894 20.155 10013.284 9993.909 20.130 0.022 0.015 -0.025
1048 10013.017 9995.136 20.108 10012.995 9995.121 20.133 -0.023 -0.015 0.025
1050 10011.993 9994.318 20.183 10012.016 9994.338 20.169 0.023 0.020 -0.014
1051 10011.835 9994.314 20.157 10011.847 9994.329 20.137 0.012 0.015 -0.020
1055 10011.188 9995.444 19.972 10011.174 9995.453 20.001 -0.014 0.009 0.029
1026 10010.346 9994.063 20.086 10010.330 9994.044 20.061 -0.016 -0.019 -0.025
1061 10009.942 9996.762 19.847 10009.965 9996.777 19.862 0.023 0.015 0.015
RMSE ±0.004 ±0.006 ±0.002
Asia Geospatial Forum, 24-26 September 2013, Kuala Lumpur, Malaysia
From Table 8 it shows that the RMSE of differences in coordinates from the image processing software
and total station were less than 1.0 for archeological site which indicate the good results were achieved in
this simulation model.
Table 8: Comparison of check points for Archeological Site
Check
Points
Erdas Imagine Total station Diff. in Coordinates
X(m)
Y(m)
Z(m)
X(m)
Y(m)
Z(m)
∆X(m)
∆Y(m)
∆Z(m)
C01 10010.687 9998.507 20.275 10010.783 9998.575 20.332 -0.096 -0.068 -0.057
C02 10009.475 9999.649 20.190 10009.559 9999.707 20.278 -0.084 -0.058 -0.088
C03 10009.719 9997.949 20.197 10009.782 9997.970 20.262 -0.063 -0.021 -0.065
C04 10010.130 9998.232 20.246 10010.188 9998.319 20.319 -0.058 -0.087 -0.073
C05 10008.896 9999.169 20.154 10008.953 9999.266 20.213 -0.057 -0.097 -0.059
C06 10009.937 9999.477 20.199 10009.962 9999.533 20.314 -0.025 -0.056 -0.115
RMSE 0.171 0.173 0.204
The accuracy of orthophoto planimetry and vertical of meandering flume is shown in Table 9. In
planimetry accuracy, a sub-meter ±0.049m and ±0.025m were obtained for X and Y coordinates
respectively. Meanwhile the RMSE for Z coordinates is ±0.091. For average RMSE, ±0.055m
was obtained by averaging the planimetry and vertical RMSE of small format digital imagery
orthophoto of riverbed.
Table 9: Comparison check points for riverbed topography
Aerial
Triangulation
RMSE (m)
X(m) Y(m) Z(m)
31 Check Points ±0.049 ±0.025 ±0.091
Mean ±0.055m
6. Conclusion
In conclusion, this study found that using the high resolution digital camera for environmental survey
application can be used where it showed that the digital camera must be calibrated for obtaining accurate
measurement or results. The best method of calibration depends on the type of applications. Practically,
for most applications the digital camera is calibrated on site, hence, laboratory and field calibration is the
reliable and useful method of calibration and could be employed for obtaining accurate measurement. It is
also proven that the light weight rotary-wing UAV was successfully used for capturing the images of the
simulated model for mapping applications. From these experiment it have shown that this UAV-system is
easy to handle, can cover small areas, able to fly within the predicted flight path, can fly closer to the
object and thus the resolution can be increased dramatically. Moreover, it offers great advantages in
inaccessible and dangerous areas which involved archaeological site recording for documentation
purposed and also for the monitoring of coastal erosion. Recently, this UAV-system have shown the
capability in photogrammetry especially for close range applications with the low-cost budget for rapid
data acquisition.
Asia Geospatial Forum, 24-26 September 2013, Kuala Lumpur, Malaysia
Acknowledgement
Faculty of Geoinformation & Real Estate, Universiti Teknologi Malaysia (UTM) is greatly
acknowledged. The authors also would like to thank the Sustainability Research Alliance, Universiti
Teknologi Malaysia for providing the fund to enable this study is carried out.
References
Anuar Ahmad & Wan Aziz Wan Mohd Akib, 2010. Photogrammetric capabilities of high resolution
digital camera and unmanned aerial vehicle for mapping. MRSS 6th International Remote Sensing &
GIS Conference & Exhibition. (MRSS 2010), 28-29 april 2010, PWTC, Kuala Lumpur, Malaysia
(Invited Paper).
Anuar Ahmad, 2009a, 2009b Anuar Ahmad, 2009a. Mapping using small format digital imagery and
unmanned aerial vehicle platform. South East Asia & Survey Congress (SEASC 2009), 4-6 August,
2009, Bali, Indonesia.
Anuar Ahmad, 2009b. Aerial mapping using small format digital camera and unmanned aerial vehicle.
Map Asia 2009, 18-20 August 2009, Suntec Singapore Internaional Convention & Exhibition Centre,
Singapore.
Anuar Ahmad., (2011). Digital mapping using low altitude UAV. Pertanika Journal of Science &
Technology, vol 19 (S) October 2011 pp 51-58.
Baoping et al., (2008) Baoping, L., Xinpu, S., Ahiyu, X., Chengwen, E. and Bing, L., (2008). Actualize of
Low Altitude Large Scale Aerophotography and Geodesic base on Fixed-wing Unamanned Aerial
Vehicle Platform. The International Archives of the Photogrammetry, Remote Sensing and Spatial
Information Sciences. Vol. XXXVII. Part B1. Beijing, China.
Bryson and Sukkarieh, 2009 Bryson., M., and Sukkarieh., S., (2009). Architecture for Cooperative
Airbone Simulataneous Localization and Mapping, Journal of Intelligent Robort System (2009) 55,
pp. 267-297.
Cooper, M.A.R., and Robson, S., (1996). Theory of close range photogrammetry. In Atkinson, K.B (Ed.).
Close Range Photogrammetry and Machine Vision. (pp.9-51).
Eisenbeiß H., (2009). UAV Photogrammetry, Dissertation for the degree of Doctor of Sciences, Germany:
University of Technology Dresden.
Pardesi (2005), Pardesi, M., S., (2005). Unmanned Aerial Vehicle/Unmanned Combat Aerial Vehicles,
Likely Missions and Challenges for the Policy-Relevant Future. Air and Space Power Journal.
Tahar, K. N., Ahmad, A., Wan Mohd Akib, W. A. A., and Udin W. S., (2011). Unmanned Aerial Vehicle
Technology For Large Scale Mapping, ISG & ISPRS 2011, Sept. 27-29, 2011 – Shah Alam,
Malaysia.
Wolf and Dewitt, 2000 Wolf, P.R. and Dewitt, B.A, (2000). Element of Photogrammetry with
Applications in GIS. McGraw-Hill, 608pp.