ArcGIS Blog

Indoor GIS

ArcGIS Indoors

Create an Indoor GIS with Reality Capture Webinar: Answers to your Questions

By Sophie Frank and Jason Hine

Our recent webinar, “ArcGIS Indoors: Create an Indoor GIS with Reality Capture,” sparked significant interest and generated many questions from attendees. The webinar explored how organizations can leverage reality capture tools in ArcGIS Indoors to create indoor maps when CAD or BIM files are unavailable or out-of-date.

You can now watch the webinar recording.

In this blog, we’ll dive into the questions received during the live webinar and provide further insights into creating floor-aware indoor maps using lidar point clouds and importing 360-degree photos. If you cannot find the answers you were looking for, please reach out to us via the ArcGIS Indoors community.

Questions about ArcGIS Indoors and its licensing

1. Which version of ArcGIS Pro are the Indoors tools available to work with point clouds and 360° images?

The Generate Floor Plan From Point Cloud tool was released with ArcGIS Pro 3.4 but significant enhancements were made at ArcGIS Pro 3.5. The Import Indoor Images tool is available in ArcGIS Pro 3.3 and later.

2. When was ArcGIS Indoors first released and made publicly available?

ArcGIS Pro has had indoor GIS-specific capabilities since version 2.0. ArcGIS Indoors really came into its own with the release of ArcGIS Pro 2.5 in early 2020.

3. Do we need separate licensing for ArcGIS Indoors?

Yes, ArcGIS Indoors and the tools we demonstrated in this webinar (Generate Floor Plan From Point Cloud and Import Indoor Images) require an ArcGIS Indoors license. There are a few licensing options depending on what your organization wants to do. You can read more about it here.

In addition to an Indoors license, certain geoprocessing tools require other ArcGIS Pro extensions. The Generate Floor Plan From Point Cloud tool requires ArcGIS 3D Analyst as well as ArcGIS Spatial Analyst or ArcGIS Image Analyst. Learn more about licensing requirements for Indoors geoprocessing tools.

4. Are any Indoors tools available with basic licensing?

Most Indoors geoprocessing tools require at least a Standard license, and many require an Advanced license. A Basic license allows you to create the Indoors model using tools like Create Indoor Dataset, Create Indoors Network Dataset, and Create Indoor 3D Dataset. Learn more about licensing requirements for Indoors geoprocessing tools.

5. How is Indoor reality data supported in ArcGIS Online, in terms of publishing, curating, and consuming?

You can share Indoors layers, floor-aware maps and scenes, and oriented imagery layers to ArcGIS Online and use them in apps such as Indoors Viewer, ArcGIS Experience Builder, and Map Viewer (all three support floor-aware maps and include capabilities that allow viewing oriented imagery).

Though outside the scope of the workflows presented in this webinar, ArcGIS Online supports publishing additional reality products, such as publishing point cloud data from LAS files as point cloud scene layers and integrating them into web scenes and apps that support them. Learn more about supported items in ArcGIS Online.

6. What version of ArcGIS Enterprise do I need to leverage 360° photo support in Indoors Viewer?

ArcGIS Indoors Viewer added support for 360° photos at ArcGIS Enterprise 11.4. Oriented Imagery layers are also supported across ArcGIS, including an Oriented Imagery Viewer in Map Viewer (available at Enterprise 11.2) and an Experience Builder widget (available at Enterprise 11.5) among others. These capabilities are also available in ArcGIS Online.

7. How much does it normally cost to map a building like the example being shown about fires. Additionally, what is the cost of the extension in ArcGIS Pro?

A cost estimate for any given project will depend on many factors, including ones outside Esri’s control. To understand the options for licensing ArcGIS Indoors for your organization, please contact your Esri account representative.

8. What is the difference between the indoor reality workflows demonstrated and what ArcGIS Reality supports?

In the context of this webinar, the term “reality capture” is a generic reference to technologies that convert the physical world into digital data. This webinar focuses on doing “reality capture” in an indoor environment. ArcGIS Reality, on the other hand, is a suite of products designed for outdoor use with drone and satellite data for sites and larger geographic regions.

Questions related to hardware and scanning process

1. Does Esri offer building scanning services?

No, but there are a number of Esri partners who provide these types of services. Learn more about Esri’s Indoor GIS specialty partners.

2. What tools are used to scan the buildings? 

There are a growing number of lidar scanning devices in the market that are capable of collecting point cloud data in indoor environments, ranging from handheld to wearable to tripod based. Esri business partners who offer such devices include Trimble, Leica Geosystems, NavVis, and others. The market for affordable handheld scanners is growing and this is by no means an exhaustive list of available options.

3. Towards the beginning of Sophie’s portion, there was a video of a man using a wearable scanner/camera. What was the model and name of that equipment?

In the slides and demos there was a GIF of a building scan being conducted with a NavVIS VLX wearable scanner. The presentation also featured an image of someone holding a Leica BLK2GO.

4. What parameters need to be considered when collecting indoor lidar point cloud? Are there any special requirements for geospatial control?

Considering things like line of sight, reflective surfaces, sensor range, and indoor and outdoor control points to use for georeferencing are just a few of the parameters to think about before conducting a lidar scan. This blog provides a great starting point for what aspects to consider when planning your building scan.

5. What should the precision of the lidar scan be (i.e., points per square meter) to produce the type of indoor reality work demonstrated in this webinar?

In this webinar we suggest an average point density of 1-2 points per square centimeter, which would be 10,000-20,000 points per square meter and an average point spacing of 7-10 mm.

6. Can you give an estimation of time to scan a building (per square meter)?

The scan time per square meter depends on the type of scanner (e.g., handheld, wearable, or tripod based), the number and quality of lidar sensors on the device, whether or not you’re simultaneously collecting 360° photos, and possibly other factors. Our recommendation is to check with the device vendor. For example, in this blog NavVis claims a scan rate of 600 square meters in 6 minutes with an accuracy of 6 mm.

7.  I heard some iPhones can collect lidar. Does this provide the adequate level of detail?

Depending on the software and the phone you use, you may be able to capture the level of detail and accuracy necessary to generate floor plans from point clouds using the ArcGIS Indoors tools. However, scan size is a limiting factor here and scalability is difficult, given storage limitations on phones, the amount of time required to obtain the necessary point density, and the potential need to stitch together many small scans.

Questions about 360-degree photos tools

1. How does the Import Indoor Images tool understand which photo belongs to which level? Which parameter is linking this between photos and the Indoors dataset?

The Import Indoor Images tool requires the Levels features from an Indoors model as input. It compares the elevation of the intersecting levels features to the elevation of the images in the E57 file (based on the georeferencing information available in the E57 metadata). The tool adds a Level ID field to the oriented imagery layer, if one does not already exist and populates it with the related level feature’s Level ID field value.

This is why accurately georeferencing the scan data in the vendor software prior to export is important. If the elevation or Level ID information of the imported photos is not correct, you can use the tool’s Elevation Adjustment parameter to compensate, or you can move the imported features up or down using the Adjust 3D Z Value tool in ArcGIS Pro and manually populate the Level ID field.

2. Can an E57 file contain several pictures?

Yes, an E57 file can contain multiple images in multiple formats. You can export all of the photos from a building scan into a single E57 file using vendor software. The Import Indoor Images tool will then add all of the spherical (or equirectangular) 360° JPEG images to an oriented imagery dataset in a single tool run.

3. Please provide more detail on HOW the e57 file is generated as there are tons of vendor software out there. 

When using vendor software to export processed scan data, there is commonly an option to choose E57 as the output format. There may be sub-options. For compatibility with the Import Indoor Images tool, look for and choose E57 export options to:

  • Include panoramic images
  • Store images as equirectangular or spherical representation (not pinhole, cubemap, etc.)
  • Use JPEG image file format

4. How are the indoor photos’ GPS location determined and how are they placed in the correct location if they are taken indoors?

You can georeference the building scan (both the point clouds and images captured) either during the scan itself or later in the vendor software depending on what scanning device you use. GPS doesn’t work indoors, so indoor scan data needs to be georeferenced to correctly position the photos. Most device vendors use anchor or control points to support georeferencing the scan data during processing.

5. If the scanning was done without a coordinate system, the 360 images do not contain latitude and longitude information. Can we georeference the images later in ArcGIS Pro?

Yes. If the E57 is not properly georeferenced, you can move the features in the oriented imagery layer after importing them using tools such as Transform, Adjust 3D Z, and Move (or Move To). This is often less accurate and may require manually populating certain field values in the Oriented Imagery layer such as Camera Heading.

Questions about what’s possible to do with these tools

1. Will there be capability of tagging an item in the photos for it to appear on the map?

You can use the Map-Image Location tool on the oriented imagery layer ribbon in ArcGIS Pro to click a location in the image and view the location on the map and vice versa. There is not yet a way of creating a new feature directly from the Oriented Imagery viewer.

2. Is it possible to modify the orientation of the images (like their true north)?

The image orientation information (heading, pitch, and roll) is automatically populated based on metadata in the E57 file, but you can modify it in the Oriented Imagery layer if necessary.

3. Is Esri considering gaussian splat models as an option for display in the software?

Yes, Gaussian Splat layers are on Esri’s road map.

4. Can you enable an ArcGIS Indoors app with 360° photos taken on an outdoor space?

Yes. You can integrate both indoor and outdoor photos in a floor-aware map. Ensure the Level ID field value is Null for features located outdoors for proper interaction with the floor filter.

5. Can floor plans (polylines/units) be generated from 360° photography?

This is not currently supported using ArcGIS Indoors tools.

6. Is it possible to export the generated floor plan as an IFC or to BIM? Is it in 3D?

Floor plans generated using the Generate Floor Plan From Point Cloud geoprocessing tool are z-enabled 2D polylines. They are not 3D objects and therefore cannot be exported to BIM (IFC, etc.).

7. How do you keep the floor plans square with this method?

Our tool includes algorithms to square corners as much as possible. It works best with 90 degree angles.

8. How do you speed up the point cloud in rendering within ArcGIS Pro?

To optimize performance of point clouds in ArcGIS Pro it is recommended that you 3D thin, tile, and build pyramids on your LAS dataset. Learn more about optimizing performance of point clouds in ArcGIS Pro.

9. Can you make the Oriented Imagery layer available for other applications to consume?

Yes. The Oriented Imagery layer is shared as a feature service to an ArcGIS Organization and can be consumed in custom apps and viewers that support feature services.

10. Is there any utility in combining LAS data with existing CAD data? For example, our campus CAD data is on a relatively slow update schedule, so I’m wondering if collecting point cloud data would be helpful for updating an Indoor GIS between CAD cycles.

It is possible to update existing indoor GIS data using point cloud data, but it is currently a manual process. The workflow would involve importing the point cloud to GIS features in a temporary dataset, then replacing the existing GIS features using ArcGIS Pro’s editing tools. Care must be taken to ensure that feature IDs, names, and other important attribute information get updated too.

11. How easy is it to remove noise from point clouds in GIS, like people and furniture?

The Generate Floor Plan From Point Cloud tool uses point density to identify floor plan features such as walls. Choosing a Z Range that minimizes people and furniture present and maximizes walls and mullions that are present further improves removal of noise in the output polylines.

12. Is it possible to layer 3D feature data on top of the lidar data? For instance, line data representing communication cables.

Yes. You can add other feature layers to the same scene as a point cloud layer. If the features are hidden behind or within the point cloud, you can try symbolizing the features as 3D tubes and increase the tube width until the features poke through.

Questions about other ways of capturing data

1. How do you process oriented imagery from a consumer standalone camera, like an Insta360, if you want to add it to an existing Indoors data model without having a E57 file with embedded images?

When capturing images using a 360° camera there are a few options available, including:

  • Load images to an Oriented Imagery dataset using tools in ArcGIS Pro. A prerequisite for this approach is creating a table that contains information about each photo using the oriented imagery table schema.
  • Capture points in a floor-aware map using ArcGIS Field Maps along with the 360° photos and later convert them to an oriented imagery dataset.

2. When taking pictures with a 360° camera (such as Insta360 or GoPro), does each room need to be captured? One photo per point?

Each point in the oriented imagery layer should be associated with a single image. Determining whether to include a photo of each room depends on your organization and the audience for whom you are building floor-aware maps and apps.

Want to learn more? Review these essential resources for Indoor GIS.

Share this article

Subscribe
Notify of
0 Comments
Oldest
Newest
Inline Feedbacks
View all comments