How to Register and Transform Point Clouds into a Project Coordinate System

Why Point Clouds May Not Align

When colored and uncolored point clouds do not align, first verify their source data, origins, coordinate systems, units, and XYZ order. Misalignment can result from different capture segments or reference frames, mixed use of meters and millimeters, or an incorrect coordinate order. If the spatial relationship is unknown, use manual corresponding-point registration; if the point cloud must be placed in a defined local coordinate system, use valid control points or independently confirmed transformation parameters. For SHARE SLAM S20, SHARE SLAM S20 SE, SHARE SLAM S100, and SHARE3DCAM C10 Series projects, first confirm the data source, units, coordinate axes, and delivery coordinate system, and then determine whether manual corresponding-point registration or coordinate transformation is required. Whether a model supports a particular field positioning configuration must be confirmed against the relevant product documentation rather than inferred solely from the software workflow.

Confirm Four Items Before Starting a Project

Confirm the delivery coordinate system, units, origin scheme, and verification method with the acceptance party, and document them in the project records.

When geographic coordinates are required, also define the source global navigation satellite system (GNSS) coordinate system, target coordinate system, and height-processing method. Control points are used as constraints or in transformations; independent check points must remain separate and be used only for verification. The two must not be used interchangeably.

How to Manually Register Two Point-Cloud Segments

If the spatial relationship between two point-cloud segments is unknown, begin with manual corresponding-point registration:

1. Select the reference point cloud that will remain fixed and the point cloud to be transformed.
2. In both viewports, pick corresponding features such as wall corners, building corners, fixed ground points, or targets.
3. Distribute the corresponding points across the full area, avoiding points concentrated in one local region or arranged in an approximately collinear pattern.
4. Check the match status, overlap, and errors; reconfirm any points with large errors.
5. After confirmation, generate original-name_registered.las. The source point cloud will not be overwritten. Point-cloud merging only combines already aligned files into one file. It does not solve for a spatial transformation and therefore cannot replace registration.

How to Transform an Existing Point Cloud into the Target Coordinate System

Coordinate transformation in

SHARE PointClouds Studio applies when real-time kinematic (RTK) was not used, RTK or control quality is insufficient, or a point cloud must be transformed into an existing target project coordinate system:

1. Verify the length units and confirm that X/Y/Z in the CSV correspond to easting, northing, and elevation.
2. Import the control points and match each one in the point cloud.
3. Once at least three valid control points are available, calculate the transformation parameters. Remove incorrect points or points with large errors, then recalculate.
4. If confirmed parameters are already available, you can instead enter X/Y/Z offsets, X/Y/Z rotations, and a scale factor.
5. Preview the result and save it as original-name_converted.las without overwriting the source point cloud.

Figure 1: The Coordinate Transform window shows the control-point table, offsets, rotation, scale, and preview controls.

When coordinates from existing drawings are used, the 3D transformation should likewise be completed with at least three well-distributed valid points whenever possible. Two points alone do not satisfy the requirements for the 3D transformation described here.

How External Post-Processed Trajectories and Satellite Positioning Fit into the Coordinate Workflow

When real-time RTK status is unsatisfactory, use an external post-processed kinematic (PPK) tool to process rover, base-station, and navigation data and generate an RTKLIB POS file. Check trajectory continuity, gaps, and solution status. SHARE PointClouds Studio only uses the resulting file as the Trajectory Source in GNSS Fusion; verify the source coordinate system, target coordinate system, and height offset before processing. A PPK trajectory cannot replace control points or independent check points.

Checking Trajectories, Photos, and Coordinates

If undistorted photos were generated during processing, image comparison can link the plan view, trajectory, photo locations, photos, and 3D viewport. This is useful for checking capture locations, site details, and anomalous areas, but it cannot independently demonstrate coordinate accuracy. After coordinate processing, perform the following checks separately:

• 3D Accuracy Check: Import the independent Checkpoint File, pick the corresponding points in the point cloud, run Matching Calculation, and use Export Report to generate the 3D accuracy report.
• Elevation Accuracy Check: Import the independent Checkpoint File, set Matching Neighborhood, run Matching Calculation, and use Export Report to generate the elevation accuracy report. Record the two types of checks separately. Control points must not serve as the sole basis for verification.

In Section Extraction, set the direction and project-required width. Choose one mode: use Batch Cutting with Number of Sections and Batch Interval, or use Custom Interval to define custom positions. Then verify the layout in Preview before extraction. CAD drawing export simultaneously generates drawing exchange format (DXF) and DWG files and includes them in the exported ZIP package; lines, annotations, units, and coordinates must be checked manually. RVT, IFC, and complete BIM models remain downstream professional workflows. 3D Gaussian splatting (3DGS) is intended only for photorealistic viewing and must not be used as a deliverable for coordinate, dimensional, or accuracy acceptance. The coordinate delivery package should include a description of the source data, the _registered.las or _converted.las file, coordinate-system and unit information, the control-point table, independent check reports, DXF review records, and the software version. Failed or canceled tasks can be retried. If an exception occurs, retain the project and logs; you must be signed in before uploading logs.

FAQ

Q: What should I do first when colored and uncolored point clouds are misaligned? First confirm whether they come from the same source, then check the origin, coordinate system, units, and XYZ order. If their spatial relationship is unknown, use manual corresponding-point registration.

Q: What is the minimum number of points required for coordinate transformation? At least three valid control points are required, and they should be distributed across the project area. Alternatively, confirmed XYZ offset, rotation, and scale parameters may be used.

Q: Will registration and coordinate transformation overwrite the original files? No. Registration outputs _registered.las, while coordinate transformation outputs _converted.las.

Q: Can the software generate a POS file directly for PPK? An external PPK tool generates the RTKLIB POS file. SHARE PointClouds Studio only uses that file as the Trajectory Source in GNSS Fusion.

Q: How can I demonstrate that the transformed coordinates are reliable? Use independent check points that were not involved in the transformation, perform separate 3D and elevation checks, and assess the results against the project acceptance criteria.