Introduction
There are many methods to obtain
aerial imagery of an area whether it is via satellite or aircraft. The
development of this new technology in the 20th century has provided
the world with reliable high resolution aerial imagery of the earth. Although
these are the principal tools being used today to collect images, they are expensive
to build and run. This week’s exercise entails for the class to develop a model
that will map part of the University of Wisconsin-Eau Claire campus using a
balloon filled with helium. Each of the six groups is responsible for
georeferencing and mosaicking a particular section of campus. This report will
cover what the class did last week to prepare for this week’s exercise, a
breakdown of the methods employed during the creation and launch of the
balloon, the resulting aerial image from group 3, and the obstacles faced
during the activity.
Methodology
Balloon mapping requires a
significant amount of preparation and experimentation before the official
launch of the device that will collect the data for the final aerial image of
school grounds. The study is split up into two activities that include the
preparation and testing of the balloon mapping tools in week one and the actual
launch in week two, which will collect the data for producing the complete
campus image. This project is laid out in the following methods: prepare and
experiment with the balloon mapping tools to conduct two trial launches in week
one, modify rig designs, measure out 400 feet of rope, blow balloon up with a proper
amount of helium, and determine a clear flight path.
In last week’s exercise, the class
had the opportunity to prepare and experiment for this week’s official launch
of the helium balloon over the UWEC campus. There were two practice launches in
this activity, one for taking aerial images of campus and the other for
recording video images. This allowed for the creation of various rig designs,
which attach to the balloon and contain the camera inside. Since wind is a big
factor at 400 feet in the air, the rig had to be designed to withstand strong
gusts and ensure the camera could take quality images of the ground below. Stacy,
Brandon, and I had the responsibility of creating and designing the rig for the
two trial launches as well as the final one. Figure 1 displays the rig design
used in the first experiment launch of the balloon. Another objective of week
one’s exercise was to get familiar with the flight paths of the balloon over
certain areas of the campus. The flight path is important because the best
quality images are taken directly below camera and the class wanted to capture
in detail recent aerial imagery showing the transformation of UWEC particularly
in the campus mall and the new Davies center. One obstacle with the flight path
is staying clear of branches that get in the way of the individual on the
ground driving the device and the rope attached to the balloon 400 feet above.
Figure 1: Rig Design in the First Balloon Launch
After
learning from last week’s trial launches, the class is more familiar with using
the balloon mapping equipment and ready to make any necessary improvements to the
rig and any other tactics that will be employed in the final launch of the
device. Since the rig from last week’s launch experienced too much movement and
snapped off, the next step is to modify the rig in this week’s launch and
minimize motion as much as possible from the wind. Instead of going with the
Styrofoam container that held the camera in last week’s launch, the new rig
will be much simpler and consist of construction paper cut out in the shape of
a triangle attached to a plastic rod (Stacy’s idea), firm string, the digital
camera, zip ties that attaches the camera to the rod, electrical tape around
the camera for more stability, and a rubber band wrapped around the camera
button to take multiple continuous pictures throughout the duration of the
launch. Figure 2 shows the design of the new and improved rig for the final
launch.
Figure 2: New and Improved Rig for the Final Balloon Launch
The
third method is to measure out 400 feet of rope that will be the height of the
balloon above campus when it is launched. In order to keep track of how much
rope it let out once the balloon is ready to be launched in the field, the 400
feet of rope is marked at every 50 foot segment. The class was able to do this
in one of the hallways long enough in the Phillips science building on campus
shown in Figure 3.
Figure 3: Measuring 400 Feet of Rope
The fourth method is blowing up the balloon with
enough helium to ensure that it will reach the goal of 400 feet above campus. A
helium tank and a balloon specialized in reaching higher altitudes are used in
this step shown in Figure 4. Once the balloon inflates to a diameter of about 5
feet, it is ready for launching.
Figure 4: Inflating the Balloon with Helium
The
fifth method is determining a clear flight path to drive the balloon around
campus. After experimenting on the areas that were possible to go with the
balloon last week, the class followed a similar path around the campus mall and
the new Davies center. There were also new flight paths taken around campus,
which was permitted by a day with light winds and a significant portion of the
school grounds being cleared from thick trees. The other flight paths traced
around the Phillips science building and then across to the footbridge by the
Haas building. Aerial imagery of upper campus was also collected by the device.
Once the launch has been completed,
images from the camera can be uploaded to the computer and used for the sixth
method, which is georeferencing the ones that are in the section assigned to my
group (group 3) in ArcMap. When images are georeferenced, they are aligned to a
map coordinate system because they usually do not contain information about the
area that they capture from a camera. ArcMap allows the user to bring in jpegs
as a raster, which can be manipulated using the georeferencing tool bar. The
images will be georeferenced to another raster data set that has existing
spatial data and a desired map coordinate system. This reference image of the
UW Eau Claire vicinity is provided to us by the instructor. A minimum of 9
control points, which are ground points with known x,y coordinates that link
locations on the image with locations in the spatially referenced image, are to
be added to each image linking it to the reference photo.
The last method is to mosaic the
newly georeferenced images to a new raster shown in Figure 5. This will combine
the images into one whole image that covers the desired section assigned to
group 3. In ArcMap, the Mosaic to New
Raster tool is used to output a new raster image with all of the other images
connected together. It is important to note that in order to run this mosaic
tool, the input raster images need to have the same number of bands and bit
depth. Since all of the images from the camera have the same characteristics
aforementioned to run the tool without error, there is no issue.
Figure 5: Mosaicked Image for Group 3
Discussion
Although balloon mapping provides an
easy and inexpensive way to collect aerial imagery, there were a few obstacles
encountered during the activities. The major issue was the design of the rig
and how it behaved in the wind. Since the study depends on the collection of
images, it is imperative that the camera lens faces in a downward direction and
not swing at angles. The rigs of the first two trial launches were made out of Styrofoam
and shaped as a cube, which proved to be too much mass that was easily affected
by even delicate wind gusts. It also did not help the fact that the winds were
at sustained speeds of 15 miles per hour that first day the balloon rigs were
launched. As a result, the photos captured from the camera did not turn out as
clear or taken at direct angles as images from the second week when the winds
were light and a new rig was launched. If the two trial launches were not
conducted a week prior, the images for this week’s balloon launch would have
been more distorted and lesser quality.
Another obstacle during the balloon
mapping exercise was the tree branches getting in the way of the flight path.
Although the class planned a clear flight path a week prior, the instructor and
the students driving the device (me and Zac) decided to go to other parts of
campus to capture images. Once the area around campus mall and the new Davies
center was covered, new flight paths were added spontaneously including parts
of upper campus and the Hass building across the Chippewa River.
Finally, the last problem occurred
when using the Mosaic to New Raster
tool in ArcMap when combining the images. The output image would appear black
or not show up at all. To fix the output when it appears black, one can go into
properties of the feature class and go into the Symbology tab. Make sure the
RGB Composite is highlighted in the upper left of the Layer Properties screen
and where it says Type in the Stretch section click the down arrow and select
none. This should change the coloring of the image and it should appear as
normal once applied.
The new design of the rig in the
second week helped tremendously capture quality images. It can be windy at
higher altitudes of over 400 feet especially near the Chippewa River, but with
the rig being light weight and having a wind breaking wing attached to it
created for more stability. Different designs of rigs should be researched
before launching to the sky.
Conclusion
The balloon mapping exercise helped
improve my ability to collect aerial imagery by using simple tools and how to
conduct post processing of the images in ArcMap. After preparing and testing
out the various rig designs, the final launch proved to be successful at
capturing high resolution images. I have taken a lot out of this experience by
learning how to obtain aerial imagery over a particular parcel of land without
using high tech equipment.
No comments:
Post a Comment