Create a hosted imagery layer and extract features from imagery
To identify infrastructure vulnerable to natural hazards such as landslides, you must first know the infrastructure locations. After having obtained aerial imagery representing a portion of Grenada, you'll create an online imagery layer hosted in the ArcGIS Online cloud. You'll then use deep learning analysis capabilities in ArcGIS Online to automatically extract the building footprints from the imagery layer.
These capabilities are part of ArcGIS Image for ArcGIS Online, a complete software as a service (SaaS) offering for hosting, analyzing, and streaming imagery and raster collections. With the Creator user type and an add-on license for ArcGIS Image for ArcGIS Online, you can manage imagery collections, stream tiled and dynamic image services, and perform advanced analysis, such as deep learning, in the cloud.
Download data and create a tiled imagery layer
First, you'll download a compressed file containing the aerial imagery so that you have it on your local computer.
- Download the Grenada_TIFF_files.zip file.
- Locate the downloaded Grenada_TIFF_files.zip file on your computer.
Note:
Most web browsers download files to your computer's Downloads folder by default.
- Right-click the Grenada_TIFF_files.zip file and unzip it to a location on your computer, for example, your C drive.
- Open the extracted Grenada_TIFF_files folder to inspect it.
The folder contains 16 TIFF images along with their auxiliary files. Together they represent the extent of Grenada you want to analyze.
Now that you have downloaded the imagery to your computer, you are ready to start the actual workflow.
Create online tiled imagery layers
Since you want to perform this analysis workflow in the cloud, you will upload the 16 images to ArcGIS Online, gathering them into a single hosted imagery layer.
- Sign in to your ArcGIS organizational account.
Note:
If you don't have an organizational account, see options for software access.
- On the top ribbon, click Content.
- Click New item.
- In the New item window, click Imagery layer.
- On the Step 1 – Get started tab, ensure that Tiled Imagery Layer is checked.
The Tiled Imagery Layer type is optimized for distributed processing and analysis in the cloud. It will work well for your analysis.
Note:
To learn more about the various types of online imagery layers supported by ArcGIS Online, see Your Guide to Sharing Imagery & Raster Data.
- Click Next. On the Step 2 – Configure layer tab, for Choose a layer configuration based on your imagery, choose One Mosaicked Image.
This option allows you to gather all 16 images into a single layer that will cover your entire area of interest.
- Click Next. On the Step 3 – Define imagery tab, for Choose the raster type that best describes your imagery, leave the parameter set to Raster Dataset.
- For Select input imagery, click Browse.
- In the Open window, browse to the downloaded Grenada_TIFF_files folder. Press Ctrl+A to select all 64 files listed and click Open.
The image files begin to upload. You can monitor the progress in the Upload status column.
- When all files show as 100 percent uploaded, click Next.
- In Step 4 – Set item details, for Title, type Grenada_aerial_imagery followed by the initials for your name.
Note:
As you create any imagery or feature class layer in ArcGIS Online, you must ensure its name is unique across your organization. One way to do this is to add your initials at the end of every layer you create. For instance, if your name is Jane Smith, the layer name would be Grenada_aerial_imagery_JS.
- Enter the remaining information about the images:
- For Tags, type Grenada.
- For Summary, type Aerial imagery of the island of Grenada.
- For Save in folder, accept the default location or choose a folder of your choice in your ArcGIS Online account.
- Click Create.
The process to create your hosted imagery layer starts. After a few minutes, the item details page for the new layer appears. Starting from 16 TIFF images, you created a single tiled imagery layer hosted in the ArcGIS Online cloud. You will now open the layer in a web map and inspect it visually.
- On the item details page of the imagery layer, click Open in Map Viewer.
Note:
If the button reads Open in Map Viewer Classic, click the drop-down arrow, and choose Open in Map Viewer.
ArcGIS Online offers two map viewers for viewing, using, and creating maps. For more information on the map viewers available and which to use, see this FAQ.
This tutorial uses Map Viewer. The image layer appears in a new map.
- Zoom in and pan around the map to inspect the building locations more closely.
There are hundreds of buildings in this part of Grenada. You could manually trace each building and store the footprints as features in a feature layer, but this would be tedious and time consuming. Instead, you will use deep learning to extract the building footprints automatically.
Extract building footprints using deep learning
Deep learning models can classify or detect features in imagery effectively. Building and training your own model or fine-tuning an existing pretrained model is an advanced task. The most difficult and time-consuming aspect of using deep learning is to create a series of training samples to teach a model to recognize the specific type of information that you are interested in.
Alternatively, you can use an existing model that was trained for you. ArcGIS Living Atlas of the World provides a growing library of such pretrained deep learning models. By using these models, you can get started right away with using artificial intelligence to extract information and gain insights from your imagery. Next, you will use a pretrained model from ArcGIS Living Atlas to detect building footprints from your imagery layer.
- On the Settings (light) toolbar, click Analysis.
- In the Analysis pane, click Tools.
- In the Tools pane, click the Use deep learning section, and click the Detect Objects Using Deep Learning tool.
The Detect Objects Using Deep Learning tool pane appears.
- Under Input layer, set the following parameter values:
- For Input imagery layer or feature layer, click Layer and choose Grenada_aerial_imagery.
- For Processing mode, ensure that Process as mosaicked image is selected.
Tip:
You can learn more about each parameter by pointing to the i button next to it or by reviewing the Detect Objects Using Deep Learning documentation.
- Under Model settings, for Model for object detection, click Select model.
- In the Select item window, click My content and choose Living Atlas.
A list of pretrained deep learning models managed by ArcGIS Living Atlas appears.
- In the search box, type Building Footprint Extraction.
- In the result list, select Building Footprint Extraction - USA and click Confirm.
Once you select the deep learning model, the model arguments load automatically in the tool pane.
- Under Model arguments, for Threshold, type 0.6.
The objects detected will only be added to the output dataset if the confidence level is equal to or greater than the threshold value. The optimal threshold value can be found by trial and error.
- Under Result layer, for Result layer name, type Grenada_buildings followed by the initials for your name.
The tool is ready to run. Anytime you perform analysis in ArcGIS Online, there is a cost in credits for Esri to use its resources to process your data. You can determine how many credits a tool will consume before you run it. Once you run the analysis, the credits will be deducted from your organization's available credits.
- Click Estimate credits.
After a few moments, the number of credits it will cost to run the tool on your data appears: 10.63.
Note:
The cost of running a tool in ArcGIS Image for ArcGIS Online is based on the complexity of the analysis and the number of pixels to be processed. You can reduce the cost by running the tool on a smaller extent: on the map, zoom in to an area of interest, then on the tool pane, expand the Environment settings section, and for Processing extent, choose Display extent.
- At the bottom of the tool pane, click the Run button.
After a few moments, a pop-up announces that the process has been submitted. The tool may take 10 to 15 minutes to run.
- In the pop-up, click View status.
The History tab appears, indicating that the process is currently running.
Tip:
You can also access the History tab by clicking the History button on the The Detect Objects Using Deep Learning tool pane.
When the process is complete, the status message updates to say so.
Note:
You can click the status message to display more information about the process.
The result layer, Grenada buildings - ObjectsDetected, appears listed in the Layers pane.
It also appears on the map. It is a feature layer in which each polygon represents a building.
- Zoom in to the map and inspect the results of the analysis.
- In the Layers pane, next to the Grenada buildings - ObjectsDetected layer, turn the Visibility button on and off.
On the map, you can compare the buildings in the imagery and the buildings that were detected using deep learning.
Imagery only (left) and detected buildings (right). You can observe that the model was successful in detecting nearly all the buildings in the imagery.
So far in this workflow, you created an online imagery layer made of 16 individual images mosaicked together. You then used deep learning analysis capabilities in ArcGIS Online and a pretrained model from ArcGIS Living Atlas to automatically extract the building footprints from that layer.
Now that you have the buildings as a feature layer, you can use it for many types of operations. In the rest of this tutorial, you will use it to better understand how potential landslides could affect building structures in the area.
Perform landslide susceptibility analysis
Now that you know the building locations, you must identify the areas in Grenada that are susceptible to landslides. For this analysis, you'll use four raster layers, and apply to them several raster functions gathered into a single raster function template. Finally, you'll compare the landslide susceptibility result layer to the extracted buildings layer to identify the structures that are most at risk.
Open the web map and explore the analysis layers
To analyze landslide susceptibility, you'll use as input four raster layers. Each one represents a major factor in landslide risk assessment:
- Soil type—Areas with specific types of clay in the soil are at higher risk of landslides.
- Elevation—Areas with steeper slopes are at higher risk.
- Distance from rivers—Areas closer to rivers are at higher risk.
- Land use—Areas with roads and buildings, and areas that are artificially vegetated are at higher risk; forested areas are at lower risk.
You'll now explore the four raster layers that were gathered for you in a shared web map. First, you'll open the map.
- In a new tab of your web browser, open the item details page for the Grenada landslide analysis web map.
- On the item details page, if necessary, sign in with your ArcGIS organizational account.
- Click Open in Map Viewer.
Note:
If the button reads Open in Map Viewer Classic, click the drop-down arrow, and choose Open in Map Viewer.
The web map appears, showing the island of Grenada. No layers other than the topographic basemap are currently displaying because they are turned off.
This map is hosted in another ArcGIS Online organization and doesn't belong to you. Before you continue the analysis, you will save a copy of the web map to your own account.
- On the Contents (dark) toolbar, click Save and open and choose Save as.
- In the Save map window, accept the default values and click Save.
The map you see is now your own copy saved in your ArcGIS Online account. You'll now examine the four raster layers that represent important landslide risk factors.
- On the Contents toolbar, click Layers.
- In the Layers pane, next to the Land use layer, turn the Visibility button on.
The layer appears on the map.
- On the Contents toolbar, click Legend.
The legend for the Land use layer appears.
Tip:
The Legend pane displays only the legend information for the layers that are currently visible on the map.
- Examine the Land use layer on the map along with its legend in the Contents pane. Zoom in and pan to better understand the information contained in the layer.
- When you are done with your examination, in the Contents pane, click Layers.
- Turn off the visibility for Land use and turn on the visibility for Distance to rivers.
- Examine the Distance to rivers layer. Similarly, examine the Elevation and Soil types layers, consulting their legends as necessary.
Note:
The Distance to rivers values are expressed in meters and show distances to the nearest river.
The Elevation values are also expressed in meters and indicate the elevation above sea level.
Four raster layers: (A) Land use, (B) Distance to rivers, (C) Elevation, (D) Soil types.
Next, you'll use these layers as input to your landslide susceptibility analysis.
Create a susceptibility layer using a raster function template
You will perform the susceptibility analysis using the four raster layers as input. You will apply to them several raster functions gathered (or chained) together into a single raster function template (RFT). You'll use a preexisting RFT that was shared in ArcGIS Online. First, you'll find the RFT and examine its content.
- On the Settings toolbar, click Analysis.
- In the Analysis pane, click Raster Functions.
- At the bottom of the Raster Functions pane, click Open Raster Function Template.
- In the Browse raster function templates window, click My content and choose ArcGIS Online.
You will search for your RFT of interest in ArcGIS Online.
- In the search box, type Landslide Susceptibility Grenada owner:Learn_ArcGIS. In the list of results, for Landslide Susceptibility Grenada, click More details.
A side pane appears displaying more information about the RFT.
- At the bottom of the side pane, click View full item details.
The item page for the RFT appears in a new tab of your web browser.
- On the item details page, if necessary, sign in with your ArcGIS organizational account.
To review the content of the RFT, you'll open it in the raster function editor.
- Click Open in Raster Function Editor.
After a few moments, in the Raster Function Editor window, the RFT appears displaying all the elements it contains chained together.
The four green elements in the RFT represent the four raster inputs that you'll need to provide when running the RFT. Each yellow element represents a raster function. The process goes as follows:
- First, some rasters are preprocessed, for instance, the Elevation raster is transformed into a slope raster in which each cell identifies the steepness of the terrain at its specific location (Slope function).
- Each raster is then processed so that the original value of each cell is transformed into a 1-to-10 value, with 10 representing the highest landslide risk and 1 representing the lowest (the Remap or Calculator function).
- The four result layers are then combined (Weighted Sum) and transformed into an output raster in which each cell contains a 1-to-5 value representing the overall landslide susceptibility score (Remap: Classify Results).
- Finally, labels are attached to the 1-to-5 numeric values to represent 5 classes of risk (Very Low, Low, Moderate, High, Very High) and are symbolized with a relevant color scheme (Attribute Table).
- Optionally, double-click some of the raster functions in the RFT to see how they are set up.
Note:
Optionally, you could click the Save As button to create your own copy of this RFT and edit it in the raster function editor. However, in this tutorial, you will just apply the existing RFT to your data without changing it.
- When you are done with your review, switch back to the web browser tab that contains your Grenada landslide analysis web map with the Browse raster function templates window displaying.
You will now apply the RFT to your data.
- In the Browse raster function templates window, select Landslide Susceptibility Grenada and click Confirm.
After a few moments, the RFT appears as a tool on the side of your map, listing its required parameters.
- In the Landslide Susceptibility Grenada pane, choose the following input parameter values:
- For Distance to Rivers, choose Distance to rivers.
- For Land Use, choose Land use.
- For Elevation, choose Elevation.
- For Soil Types, choose Soil types.
- Under Result layer, , for Output name, type Landslide susceptibility followed by the initials for your name.
- For Result type, verify that it is set to Tiled imagery layer.
- Click Estimate credits.
After a few moments, the estimated number of credits it will cost to run the tool on your data appears: 1.46.
- Click Run.
After three to four minutes, the result layer appears.
- In the Layers pane, ensure that all layers have their Visibility off except the Landslide susceptibility result layer.
- On the Content toolbar, click Legend to view the color symbology for the five Landslide susceptibility classes.
- Zoom in and pan to explore the result layer.
The red areas are the most susceptible to landslides, and the green areas the least.
In this section, you examined a raster function template and ran it to create a landslide susceptibility raster layer.
Add building footprints and compare layers
You will now compare visually the building footprints you extracted earlier to the landslide susceptibility layer with the goal of identifying at-risk structures. First, you'll add the building footprints layer to the current map.
- If necessary, on the Contents toolbar, click Layers. In the Layers pane, click Add.
- In the Add layer pane, ensure that My content is selected and locate Grenada_buildings in the list. For the Grenada_buildings layer, click Add.
The layer appears on the map. To see it better, you will zoom to it.
- Next to Add layer, click the back button to return to the Layers pane.
- For the Grenada_buildings layer, click the Options button and choose Zoom to.
The map zooms in to the area where you extracted the buildings using the deep learning model.
Note:
To ensure the brevity of this tutorial, you only extracted the building footprints for a portion of the island. In a real-life setting, you might choose to extract the building footprints for the entire island instead.
- Zoom and pan through the map to identify buildings that are in high-risk areas (represented in red or orange).
It appears that most buildings in Grenada are in lower-risk areas. However, some buildings are in at-risk zones (orange). You will now save your map.
- On the Contents toolbar, click Save and open and choose Save.
As possible next steps, on the Contents toolbar, you could click Share map and share your web map with your colleagues and community. You can also retrieve the layers resulting from your analysis at any time by signing in to your ArcGIS organizational account and clicking Content. You could reuse these layers in many ways.
Note:
For instance, you could display them in a 3D web scene. Note that in that example scene, the buildings display as 3D objects because the 2D building footprints have a 10-meter extrusion applied. You can learn more about 3D scene workflows in the Create a scene tutorial.
In this tutorial, you used imagery and other types of raster data to study structures at risk of landslide on the island of Grenada. First, you uploaded 16 aerial TIFF images in ArcGIS Online and gathered them all into a tiled imagery layer. Then you accessed a pretrained deep learning model in ArcGIS Living Atlas and used it to extract building footprints from the imagery layer. Next, you used a raster function template to perform raster analysis and classify the landscape according to landslide susceptibility. Finally, you compared the landslide susceptibility layer and the extracted building footprints to visualize the structures at risk. The resulting web map can now be shared with your colleagues and community.
You can find more tutorials in the tutorial gallery.
