Entry Name: "SMU-Team JKY-MC3"
VAST Challenge 2017
Mini-Challenge3
Team
Members:
Dr. Kam Tin Seong, Singapore
Management University, tskam@smu.edu.sg
Kishan
Bharadwaj Shridhar, Singapore Management University, kishanbs.2016@mitb.smu.edu.sg
Ong Guan Jie Jason, Singapore Management University, jason.ong.2016@mitb.smu.edu.sg
Zhang Yanrong, Singapore Management University, yrzhang.2016@mitb.smu.edu.sg ,
PRIMARY
Student Team: YES
Tools Used:
- Tableau
- QGIS
Approximately how many hours were spent working on this submission
in total?
200
May we post your submission in the Visual Analytics Benchmark
Repository after VAST Challenge 2017 is complete? YES
Video
Tableau Workbook
Questions
1 – Boonsong Lake
resides within the preserve and has a length of about 3000 feet (see the Boonsong Lake image file). The image of Boonsong Lake is oriented north-south and is an RGB
image (not six channels as in the supplied satellite data). Using
the Boonsong Lake image as your guide, analyze and report on the scale and orientation of the
supplied six-channel satellite images. How much area is covered by a
pixel in these images? Please limit your answer to 3 images and 500
words.
The left image is the RGB image of Boonsong Lake.
As we can see in the graph, the vertical length of Boonsong Lake
is 3,000 feet. After using NDWI as the measure to identify the feature of Water
Region, we focus on the coordinates involving Boonsong Lake
only, which is shown on the right. The range of Y coordinate in the filtered
graph is from 474 to 504. So the length of
30 pixels (504 minus 474) are 3,000 feet in the real world. Then the length of
each pixel is 100 feet (3000 divided by 30) and the area of each pixel in the
real world is 10,000 sqaure feet(100 ft × 100 ft). Besides,
there are 287 pixels so the area of Boonsong Lake
is 2,870,000 sqf which is around 27
hectares.
2 –Identify features you can discern in the Preserve area as
captured in the imagery. Focus on image features that you are reasonably
confident that you can identify (e.g., a town full of houses may be identified
with a high confidence level). Please limit your answer to 6 images and 500
words.
For feature identification.,
we use three different methods to determine our features. The first one using
is by QGIS. We use false color (B5, B4, B2 as RGB color) then it will differentiate some features in the
image. As shown below, all kinds of features have been labeled
by observation of the image after using false color
processing.
Frequently used false-color images are:
• B4, B3, B2 --> RGB: can
be useful in seeing changes in plant health.
• B5, B4, B2 --> RGB: can
be used to show floods or newly burned land.
• B1, B5, B6 --> RGB: can
differentiate between snow, ice and clouds
The second method is
determined by some measurements such as NDVI, NDWI and so on. We use Tableau do
filter all the pixels with the value in the target range for each measurement.
For example, we use NDVI to
determine plant region, NDWI to determine water region, BSI to determine soil
region and paths, etc.
The third method is by using Tableau clustering by analyzing the all 6 bands and cluster them into 10 possible
clusters, then we can see from the graph to determine that which region belongs
to which cluster.
In conclusion, we cluster all the features into six main
categories. They are plant region, water region, roads, farmland, construction
and other region. The graph below showed
both the image processed by QGIS and Tableau.
3 – There are most likely many features in the
images that you cannot identify without additional information about the
geography, human activity, and so on. Mitch is interested in changes
that are occurring that may provide him with clues to the problems with the
Pipit bird. Identify features that change over time in these images,
using all channels of the images. Changes may be obvious or subtle,
but try not to be distracted by easily explained phenomena like cloud
cover. Please limit your answer to 6 images and 750 words.
This graph board showed the analysis with each measure by
choosing the target measure in the dropdown list on the top-right. Graph on the
left is the choropleth map of the satellite image. Line chart on the left shows
the two factors deriving the measure and their change by month. Chart on the
bottom is the histogram for certain measure.
For example, we choose NDVI as our measure. The higher NDVI
value is, the healthier or denser the plant is. And the factor1 is B4 Factor2
is B3 according to the formula of NDVI. From the line chart, there is obvious
difference between two factors from June to September. The result would be
significant if we use month among them.
From the multiple dropdown list, we compare the performance of
NDVI on September from 2014 to 2015. There is obvious shift towards left from
2014 to 2015, which means the plant became less dense and the preserve is less
healthy for some certain reasons. And it might be the possible cause of the
decreasing of the Pipit Birds. It deserves a deeper research for that.
Because the habitat for Pipit might be grassland rather than
tall trees. Certain areas with relatively low NDVI could be the key factors.
Although the overall NDVI can be improved, further research is required to focus
more on the target areas that can be the habitat of Pipit
By using NDSI, we can
compare the snow coverage between 2014 and 2016. And the area graph shifted
left from Dec 2014 to Dec 2016, which showed that the snow coverage became
smaller and that the temperature became higher during this period. In addition,
by using NDMI, we can see that the area graph shifted right, which meant the
moisture became higher from Jun 2015 to Jun 2016. And the area graph came can
show us this kind of change more direct.
We use the difference of NDVI between 2015-06 and 2016-06 to
find the place which plant become less healthy. In addition, we use the
difference of NBR between 2015-06 and 2016-06 to find the burned area.
In conclusion, the plant
health became less healthy and we knew its region, temperature and moisture
became higher. These changes might be caused by human activities or the
manufacturing progress or other factors. However, more imagery information is
required to study into this further.