20
January 2014
PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING
As for the current civilian applications for the UAS, they
are many and scattered over a wide spectrum of applications,
such as:
• Archeological site mapping and documentation
• Rapid response situations, such as wild fires or hurricanes
• Precision agriculture
• Vineyard mapping and management
• Rangeland monitoring
• Traffic monitoring
• Dam inspection
• Volcanic activity monitoring
When UAS has been used for most of the previously men-
tioned applications, the resulting products did not have a
mapping-grade accuracy (that is, accuracy according to a
certain mapping accuracy standard) due to the following
reasons:
• Miniaturized auxiliary systems such as GPS and IMU lack
survey-grade accuracy; however, more accurate miniatur-
ized auxiliary systems products are emerging, which will
eliminate or diminish this effect.
• The same reason as above applies for the imaging sen-
sors. Cameras used onboard these UAS are either video
cameras or non-metric imaging sensors that lack stability
and known calibration parameters. Again, the market is
witnessing new miniaturized but ruggedized and stable
imaging sensors.
• Lack of interest or the absence of need for metric products.
That is true for most of the above-mentioned applications,
as many of the applications focus on thematic and spectral
interpretation.
In order to generate mapping-grade products, the imaging
system must be certified as a metric system with an imaging
sensor having the following characteristics:
• Known interior calibration parameters (i.e., calibrated):
-- Focal length
-- Principal point
-- Lens distortion
-- CCD array
• Maintains the internal geometric relationship of (i.e., stays
calibrated)
• Accurate GPS signal recording
One of the challenges facing users of the UAS is the handling
of the imagery produced by UAS in a standard photogram-
metric workstation and software; the footprint of this imagery
is very small, resulting in a large number of frames. The
large number of frames, coupled with the lack of accurate
GPS and IMU and the unstable platform and its effect on
the resulting imagery, makes it difficult to process such data
on a standard photogrammetric workstation optimized for a
smaller number of frames and a standard 30% side lap and
60% forward lap. Such shortcomings open the door for the
technical entrepreneurs, as we are witnessing a few new
processing software packages that are optimized for the UAS
imagery. Some of those software packages departed from the
conventional photogrammetric mathematical model and uti-
lized computer vision-based and image-matching techniques
to deal with the UAS data. Among those software packages
(without endorsing any of them) are the following:
• Agisoft of St. Petersburg, Russia (
/)
• EnsoMOSAIC of Finland (
index.html)
• Pix4UAV 3D of Switzerland
/)
• Menci APS of Italy (
/)
Most of the above packages support GPS/IMU-based exterior
orientation and camera self-calibration. When looking for a
software package to handle imagery from a UAS with poten-
tial map quality products, one needs to make sure that the
software has most, if not all, of the following capabilities:
• Import thousands of images in TIFF and/or jpeg format
• Import and use ABGPS and IMU data
• Perform camera self-calibration
• Deal with ABGPS and IMU shift and drift issues
• Orthorectification
• Fully automated workflow, such as aerial triangulation
and ortho production
• Generate DTM/DSM automatically
• Generate fully automated mosaic
• Distributed batch processing
• Seamline editing
• Automatic radiometric correction and color balancing
• Data viewer and editor
“Most of those small UAS
are optimized with flight
control and management
systems to make the flight
almost autonomous with
the user responsible only
for initiating the flight.”
“In order to generate
mapping-grade products,
the imaging system must
be certified as a metric
system.”
