Imagery in Action

Exercise
Use ArcGIS Dashboards to share
imagery results
Section 6 Exercise 1
April 18, 2022
 Imagery in Action

Use ArcGIS Dashboards to share imagery results

Time to complete
30 minutes

Introduction
Imagery can be used in many ArcGIS applications. In ArcGIS Online, you can add imagery
layers to web maps to provide context and guide various workflows. Results derived from
imagery analysis can also be shared to and used in various apps. In a previous section, you
used a deep learning package to detect buildings. While you conducted a simple visual
review of detected building footprints in ArcGIS Image for ArcGIS Online, there are other
methods that you can use. One method is to review the data in ArcGIS Dashboards. A
dashboard is a powerful tool that allows you to configure various elements—such as maps,
lists, charts, gauges, indicators, and tables—in a way that enables quick and efficient data
review.

Exercise scenario
As a GIS analyst for the city of Athens, Texas, you will review the results of a deep learning
building footprint detection in a dashboard that has been created for you. The dashboard
contains a web map with the building footprints, as well as some contextual elements to help
you understand the results. In this exercise, you will review the building footprints in the
dashboard and rely on your understanding of imagery in ArcGIS to propose recommendations
on the next steps to detect all the building footprints.
Note: This exercise provides a simplified starting point for what is often a much more in-depth
workflow. The data and parameters used in this exercise are not intended to be a
comprehensive list of what might be required for this type of analysis.

Step 1: View the dashboard

In this step, you will view the dashboard that includes the detected building footprints.

a In a web browser, browse to www.arcgis.com and sign in using your MOOC credentials.

b At the top of the page, to the left of your profile icon, click the Search button , and
then type building dashboard owner:esritrainingsvc and press Enter.

Copyright © 2022 Esri. All rights reserved. 1

 Imagery in Action

c In the left pane under Filters, turn off the option to Only Search In Imagery In Action, if
necessary.

d In the search results, point to the Review Of Building Footprint Results Detected Through
Deep Learning dashboard thumbnail and click View Dashboard.

Note: The figures listed in your dashboard may be slightly different from the preceding
graphic, depending on screen size.
You can also click the name of the dashboard to view its item page.
The dashboard displays the NAIP imagery layer that was used to extract building footprints. It
also includes the feature layer containing the extracted building footprints. The feature layer
serves as the source of the data displayed in the indicator elements on the left side of the
dashboard.

Step 2: Explore the dashboard

In this step, you will explore the dashboard to consider how well the deep learning package
detected the buildings. The web map in the dashboard can be interactively modified during
the review process.

a In the upper-right corner of the map, click the Legend button to open the legend.

Copyright © 2022 Esri. All rights reserved. 2

 Imagery in Action

1. Which layer types have been added to the web map?

__________________________________________________________________________________

__________________________________________________________________________________

The map element usually contains the data that will be the focus of the rest of the elements in
the dashboard. Therefore, it is important to review the data included in the map element to
understand the data that is populating the elements.

b Next to the Legend button, click the Layers button to show the list of available layers.

c To the left of Natural Color Imagery, click the Visibility button to make the layer
invisible.

Copyright © 2022 Esri. All rights reserved. 3

 Imagery in Action

The Color Infrared Imagery layer is now visible for you to use when reviewing the building
footprints. This is the same NAIP layer as the Natural Color Imagery layer, but it is displayed
with a different raster band combination.

d Zoom in near the lake in the middle of the map.

2. What happened to the elements on the left side of the dashboard?

Copyright © 2022 Esri. All rights reserved. 4

 Imagery in Action

__________________________________________________________________________________

__________________________________________________________________________________

The elements in the dashboard have been designed to react to spatial extent changes in the
map and update accordingly.

e Make the Natural Color Imagery layer visible again, and then open the legend.

3. Which attribute is being used to symbolize the detected buildings?

__________________________________________________________________________________

__________________________________________________________________________________

The colors of the detected buildings layer reveal some information about the confidence
value reported during the deep learning detection.

f Zoom in to the map and click one of the building polygons.

In the pop-up window, you will see three pop-up pages associated with the layers in the map.
For each building footprint, you will see the pixel values of the two imagery layers in the web
map. By clicking through the pop-up pages associated with the imagery layers, you can see
the pixel values at the location selected in the map.
In addition to the pixel values, the pop-up reports the building number, confidence value, and

Copyright © 2022 Esri. All rights reserved. 5

 Imagery in Action

the calculated tax value.

Note: The tax value was generated by a simple calculation of the area within each polygon,
multiplied by a value to simulate the price per square foot. These values are fictitious and
should not be used other than to demonstrate the capabilities of the dashboard.

g In the upper-right corner of the pop-up, click the arrows to review the information on each
page.

h Close the pop-up and legend.

On the left side of the dashboard beside the elements, there is a sliding panel, as specified in
the following graphic.

i Click the arrow in the minimized panel to expand the Selectors panel.

j At the top of the Selectors panel, drag the slider bar to the right.

Copyright © 2022 Esri. All rights reserved. 6

 Imagery in Action

4. How does the map change?

__________________________________________________________________________________

__________________________________________________________________________________

The buildings in this feature layer were extracted by setting a 70 percent minimum confidence
threshold. With the selector, you can dynamically reveal which buildings were detected within
a certain confidence range above the 70 percent threshold.

k Adjust the slider to different values and see how the footprints change, zooming out in
the map, if necessary.

l When you are finished, in the Selector panel, click Reset, and then close the panel.
Hint: Click the arrow in the middle of the right edge of the sliding panel.
In this step, you used a dashboard to review characteristics about the detected building
footprints.

Step 3: Review building footprints

You will now more closely review the building footprints throughout the city.

a Zoom and pan throughout the map to review the building footprints in each of the three
confidence classes.

Copyright © 2022 Esri. All rights reserved. 7

 Imagery in Action

b Turn the visibility of the All Buildings Detected layer and the Natural Color Imagery layer
on and off in order to get a better look at the buildings.
Take note of any characteristics about the buildings which were detected, and those which
were not detected. Themes could include the following:

• Are missing buildings covered by other features?

• Which roof color is missed more often?
• Which roof color is always detected?

5. What are your observations about the missing buildings?

__________________________________________________________________________________

__________________________________________________________________________________

Due to a variety of factors, some buildings may not have been detected.
6. What are your observations about the buildings classified with a confidence value
between 70 percent and 80 percent?

__________________________________________________________________________________

__________________________________________________________________________________

__________________________________________________________________________________

7. What are your observations about the buildings classified with a confidence value
between 80 percent and 90 percent?

__________________________________________________________________________________

__________________________________________________________________________________

The differences that you notice here may not be as obvious, but they are still important to
consider. Gathering details about what is causing the lower confidence values will allow you to
configure the deep learning algorithm more effectively.
8. What are your observations about the buildings classified with a confidence value
between 90 percent and 100 percent?

Copyright © 2022 Esri. All rights reserved. 8

 Imagery in Action

__________________________________________________________________________________

__________________________________________________________________________________

The building footprints detected with the deep learning package represent a good start to
the process of mapping all the buildings in the city. Buildings have been missed, but many
have been found.

c If you would like to continue working with the dashboard, continue to the stretch goal;
otherwise, close the web browser.

Stretch goal
If you would like to further review the detected building footprints, you can perform the
following tasks.
1. Identify the next step in refining the building footprint detection.
If this was a real-world project, additional steps would be necessary.

• Based on your experience with imagery in ArcGIS over the course of this MOOC, what
next steps would you take to identify more buildings?

Use the Lesson Forum to post your questions and answers. Be sure to include the #stretch
hashtag in the posting title.
2. Add additional elements to the dashboard to help review the data more effectively.
To add additional elements to the dashboard, first make a copy of the dashboard:

1. View the dashboard in a web browser.

2. In the web browser address bar, click between https://<your
organization>.maps.arcgis.com/apps/dashboards/ and the number/letter sequence, and
then type new#id= without any spaces before or after it.

3. Press Enter.
4. On the Create New Dashboard page, for Title, add your initials to the existing title.
5. At the bottom of the page, click Create Dashboard.

Copyright © 2022 Esri. All rights reserved. 9

 Imagery in Action

You will now be able to edit the new copy of the dashboard. You can add elements or alter
the way that the current ones are configured.

• Modify the dashboard in a way that will aid your review.

For more information about ArcGIS Dashboards, see ArcGIS Dashboards Help: Create a
dashboard.
Use the Lesson Forum to post your questions. Be sure to include the #stretch hashtag in the
posting title.

Conclusion
Thank you for participating in this MOOC! There are numerous opportunities to use various
ArcGIS apps and tools to visualize and analyze imagery, including those which you learned
about in this MOOC. These exercises are an introduction to working with imagery in ArcGIS
and should help you better understand how to use ArcGIS to gain insight into and answer
questions about your data.

Copyright © 2022 Esri. All rights reserved. 10

 Imagery in Action

Answers to Exercise Questions

1. Which layer types have been added to the web map?

A feature layer containing the building footprints and two imagery layers have been
added.

2. What happened to the elements on the left side of the dashboard?

The values all changed in each of the elements. Note that the values change based on
how far that you zoom in.

3. Which attribute is being used to symbolize the detected buildings?

The Confidence attribute is being used to symbolize the building footprints.

4. How does the map change?

Only the building footprints with a higher confidence value remain in the map.

5. What are your observations about the missing buildings?

Some buildings appear to be partially covered by trees. Others have varied roof
forms in one building.

6. What are your observations about the buildings classified with a confidence value between
70 percent and 80 percent?
Some buildings appear to be less obscured by trees as the missing buildings, or have
pixel values spectrally similar to the roads in the imagery layer.

7. What are your observations about the buildings classified with a confidence value between
80 percent and 90 percent?
Some buildings appear to be in irregular shapes but have the desired spectral
similarities as the other buildings.

8. What are your observations about the buildings classified with a confidence value between
90 percent and 100 percent?
Some of these buildings are clearly visible in the imagery layer and are easily
recognizable.

Copyright © 2022 Esri. All rights reserved. 11