SLAM Processing

Requires an LP360 Land Basic or higher license.

The SLAM Processing on the TVGO tab opens the SLAM Processing wizard, which allows users to configure settings for Trajectory Processing and Geocoding for handheld lidar and imaging scanners, such as the TrueView GO handheld scanners.

SLAM - Trajectory Processing Settings

The first page of the SLAM Processing wizard (Figure 1) is used to select the desired processing mode and specify any TGCP (TVGO Control Points) that were collected during the data collection process using the handheld scanner.

Processing Mode

• SLAM: Uses only SLAM algorithms for the reconstruction. Recommended for indoor datasets. It can use a local/arbitrary or a projected coordinate system for the project Coordinate Reference System (CRS).
• GNSS: Will mostly use GNSS + IMU georeferencing information for the reconstruction, however, it may use the SLAM algorithm to enhance areas with poor GNSS solutions. Recommended for outdoor datasets where RTK corrections were not available during the data acquisition. The user needs to select the Geographic Coordinate Reference System (CRS) for WGS84 + Ellipsoidal (Meters), Epoch 2015. 
• RTK: Requires RTK corrections recorded during the data acquisition. It will mostly use GNSS + IMU georeferencing information for the reconstruction, however, it may use the SLAM algorithm to enhance areas with poor GNSS solutions. Recommended for outdoor datasets. 
  • RTK Geographic CRS & Epoch: The user needs to select the Geographic Coordinate Reference System (CRS) of the RTK correction service. Normally, the RTK corrections are based on a Global Geographic CRS (Latitude, Longitude, and Elevation) in meters with an applicable epoch for that CRS. The most common CRS used by RTK services are:
    • World: WGS84 + Ellipsoidal (Meters), Epoch 2015
    • North America: NAD83(2011) + Ellipsoidal, Epoch 2010
    • Europe: ETRS89 + Ellipsoidal (Meters), Epoch 1989
• PPK: Requires a base station observation file to post process kinematic corrections for the sensor trajectory solution. The user needs to add Base Observation Files that cover the duration of the Cycle and select the applicable Base Station Location from the Survey Nail Manager. Selecting PPK will mostly use GNSS + IMU georeferencing information for the reconstruction, however, it may use SLAM algorithm to enhance areas with poor GNSS solutions. Recommended for outdoor datasets. 
• Realtime: Display real-time point cloud data output obtained during the collection.

SLAM - Geocoding Settings

The second page of the SLAM Processing wizard (Figure 2) contains settings that were previously referred to as the TVGO Advanced Settings in LP360 versions 2025.1 or older.

Lidar Processing

The Lidar Processing section includes options for the point cloud generation, along with the Scene Types that affect the algorithms used for the reconstruction.

• Point Cloud Optimization: Reduces thickness of the point cloud to make the display of local structures more precise.
• Moving Objects Filter: Will try to clean moving objects from the point cloud.
• Point Cloud Denoising: Removes floating noise within the point cloud. This option may remove some point data in sparse and unevenly distributed areas.
  • Note: Recommended when using a TVGO132S. With a TVGO116S, selectively enable based on the scenario.

Scene Type

This setting helps the algorithm to improve the 3D reconstruction. Note: During data collection the user can select the scene type, however, that will not affect the data processing. The scene type selected in this section will be the one used for processing overriding any such setting used during data collection.

• Indoor (No GNSS): will be automatically selected if the user selects the “SLAM” processing mode. Recommended for indoor areas. 
• Outdoor (GNSS): will be automatically selected if the user selects "GNSS", “RTK” or “PPK” mode. Recommended for Outdoor areas with GNSS reception.
• Mixed Indoor and Outdoor: Recommended for projects with areas both outdoor and indoor, where most of the work is done outside with GNSS reception, however, there will be small areas surveyed indoors without GNSS reception.
• Narrow, Closed Spaces: Recommended for indoor corridors or narrow spaces.
• Sparse in Features: Recommended for datasets with few vertical objects and a lot of open spaces. The algorithm will need to use the ground heavily as a reference as there are not a sufficient number of features to use for the SLAM algorithm.

Laser Filter Settings

Filter settings can be used to help prevent noise, lower accuracy, or unwanted points from being generated in the point cloud. Similar filters may also be applied later in LP360 to remove the points from a generated dataset.

• 3D Distance Filtering (m): Filters point cloud based on the distance from the scanner to the point.
  • Minimum (0.5~10.0)
  • Maximum (5.0~120.0)
• Intensity Filtering: Filters point cloud based on intensity.
  • Minimum (0~65535)
  • Maximum (255~65535)

Photo Processing

• Remove Uncolored Points - Not typically recommended but can be useful to reduce data density and remove points at the ends and edges of scan lines as it will remove any points that could not be assigned a color.
• Update Photo Exif Tags - Recommended to create True Pose Photos to enable the Image Explorer , the Immersive Image Explorer , Modeling , and other downstream uses of the photos.

Use Street View Photos - available when the TVGO is equipped with a TrueView GO 360 Photo Kit. Processes the Insta360 imagery for use in the Immersive Image Explorer , and other downstream uses of the photos. Check this article to learn more about the How to use the Insta360 with the TVGO.

• Enable the "Use Street View Photos" checkbox to include street-level imagery in your SLAM processing workflow.
• Device Type: Select the appropriate device type from the dropdown menu based on your camera system:
  • Panoramic - Use this option for standard panoramic camera systems.
  • Dual FishEye - Use this option for dual fisheye camera systems. Select "Dual FishEye" if you plan to generate Gaussian Splat outputs.

Note: The Device Type setting determines how the imagery will be processed and integrated with your SLAM data. Ensure you select the correct option that matches your camera hardware.

Colorization Method

• None: generates a point cloud without any color assigned.
• Fast: Recommended for a colorized point cloud.
• High Precision: is a highly intensive colorization that better preserves the real colors of features; however, it is time-consuming and can only be run on higher end graphics cards.

Note: This setting requires a NVIDIA graphics card. It is recommended a 3050Ti series and above, but not a 50 series GPU, and an Intel Core i7 and above CPU.