4.1 When Georeferencing Matters
Not every scan needs to be tied to a real-world coordinate system. But when it does, the method must be chosen and set up before you scan. This decision cannot be corrected after the fact. This page lays out what each option means, when to use it, and what it requires.
Two Kinds of Accuracy
These terms are used interchangeably on job sites but they mean different things and are addressed by different tools.
Relative Accuracy
- How accurately the point cloud represents the internal geometry of the scanned space
- Measured as the dimensional error between two points within the same scan
- Achieved through good technique, adequate loop closures, and consistent speed
- Does not require RTK or GCPs
- Appropriate for as-built documentation, renovation planning, virtual tours, and any deliverable that does not need to register to an external coordinate system
Absolute Accuracy
- How closely the point cloud registers to a real-world coordinate system
- Measured as the positional error at a point with a known survey coordinate
- Requires RTK, PPK, or surveyed ground control points
- Good technique is still required, georeferencing does not compensate for poor SLAM data
- Required for survey deliverables, site topo, cadastral work, as-built vs. design comparisons on georeferenced models, and any project where the data must align with existing survey control
This decision cannot be made after the scan. RTK must be configured and active before scanning begins. GCPs must be physically placed and marked during the scan session. A scan captured without any georeferencing setup cannot be retroactively tied to a real-world coordinate system. No processing step can add absolute accuracy that was not captured in the field.
Two Sources of Elevation Error
Understanding why georeferencing works the way it does requires understanding what causes error in the first place. There are two distinct error sources in XGRIDS scans, and different georeferencing methods address different ones.
IMU leveling error originates from the inertial measurement unit's inability to establish a perfectly level reference plane at initialization. This introduces a systematic tilt into the entire dataset, the floor that should be flat has a slight pitch, and the elevation changes across the scan reflect that tilt rather than the real geometry. RTK corrects this by providing an external absolute coordinate anchor that eliminates the IMU's leveling inaccuracy. Absolute GCPs (GCP1) with surveyed Z coordinates also correct it by providing external elevation references distributed across the scan that LixelStudio uses to remove the tilt.
SLAM accumulation error is the drift that builds up along the scan trajectory as feature-matching errors compound over distance. This causes geometry far from the initialization point to be displaced horizontally or vertically relative to its true position. Ground control points with surveyed coordinates correct this by providing anchor positions distributed throughout the scan that constrain the SLAM trajectory at regular intervals. The denser the control points, the smaller the residual error between them.
RTK and GCPs are not interchangeable, but they are not as cleanly separated as they first appear. RTK eliminates IMU leveling error by providing an external vertical reference, but does not constrain SLAM drift accumulation along the trajectory. Absolute GCPs (GCP1) with surveyed coordinates, including elevation, eliminate IMU leveling error and constrain SLAM drift. They are the only single method that addresses both error sources. Relative GCPs (GCP2) constrain drift but do not correct leveling because they carry no absolute coordinate reference. The hybrid approach (RTK plus absolute GCPs) provides redundant correction of IMU leveling and adds continuous coordinate alignment between discrete GCP anchor points.
The Five Positioning Options
Relative Coordinates Only
No RTK, no GCPs. The point cloud uses a local coordinate system with the initialization point as origin. Position in that system is arbitrary, it has no relationship to real-world coordinates.
Appropriate for as-built renovation documentation, virtual tours, internal space analysis, and any deliverable that does not need to register to an external coordinate system.
No additional setup requiredRTK Georeferencing
The RTK module attaches to the scanner and receives GNSS corrections via NTRIP. It provides absolute coordinate alignment during scanning by continuously logging satellite positions with centimeter-level corrections applied. The result is a trajectory anchored to a real-world coordinate system.
Requires Fixed satellite status throughout the scan (not Float, not Single Point), more than 10 valid satellites, at least 10 meters of movement while Fixed, and more than 100 valid RTK data points logged. RTK rarely achieves Fixed status indoors, confirm in LixelGO before relying on it.
RTK module + NTRIP credentials requiredGround Control Points with Surveyed Coordinates
Physical markers placed at positions with known surveyed coordinates are identified in LixelGO during the scan and applied in LixelStudio during processing. The SLAM trajectory is adjusted to pass through each marked point's known position, constraining drift across the dataset.
This is the correct method for indoor environments, underground spaces, urban canyons, and any location where GNSS is unreliable or unavailable. Spacing must be at most 100 m for the L2 Pro and at most 50 m for the K1. Points cannot all be collinear, three-dimensional distribution across the scan area is required.
Total station or GNSS survey required beforehandXGRIDS Relative Control Points
XGRIDS sticker targets are placed in the scan environment and marked in LixelGO, but they carry no surveyed coordinates. They function as internal consistency anchors, they constrain SLAM drift and improve internal accuracy without tying the scan to any real-world coordinate system.
Appropriate when absolute georeferencing is not required but the scan covers a large area where drift without any constraint would be unacceptable for the deliverable. Common for large interior documentation projects, multi-floor building scans, and any scenario where relative accuracy must be maintained over a long trajectory.
No survey required. Sticker targets onlyHybrid: RTK and Ground Control Points
RTK eliminates IMU leveling error and provides continuous coordinate alignment. Absolute GCPs also eliminate IMU leveling error and constrain SLAM drift at discrete anchor points. Together they provide redundant leveling correction and the tightest overall constraint on both error sources, producing the highest accuracy achievable with the XGRIDS platform for any given project.
Use this approach for survey-grade deliverables, high-accuracy as-built comparisons, and any project where the client's accuracy specification requires addressing both error sources. Requires both an active RTK connection during scanning and pre-placed surveyed GCPs marked during the scan.
RTK module + surveyed GCPs requiredWhen to Use Which
Use Relative Only (Option A or D)
- As-built documentation for renovation or fit-out
- Virtual tours and 3D visualization deliverables
- Interior space analysis, room measurements
- BIM modeling where the model will be placed manually in project coordinates
- Any project where the client has not specified a coordinate system
- Indoor environments where RTK is unavailable and survey budget does not allow GCPs
Use Absolute Georeferencing (Options B, C, or E)
- Survey deliverables requiring registration to a national or local coordinate system
- Site topography where the data must align with existing survey control
- As-built vs. design comparisons on georeferenced BIM models
- Multi-site projects where scans must register to each other via a common coordinate system
- Any project with a stated accuracy specification in absolute terms
- Construction progress monitoring where spatial consistency between sessions is required
©2026 Alpine Reality Capture LLC • XGRIDS Pro Guide™ • Site Disclaimer