Project Processing

This chapter covers the post-processing workflow for raw scanner data, including project processing, map fusion, and accuracy verification.

Project Processing

Project Processing post-processes raw data from XGRIDS handheld scanners (L2 Pro, L3, K2) to produce the point cloud results you need.

Project Settings

1. Click File and select the project data to process. The project path must point to the project directory — the folder copied from the scanning device.
2. In Project Processing, confirm the imported data path.
3. Set the mapping time range as needed. To limit processing to a specific window, drag the timeline or choose a custom time range.
4. Choose a viewing angle in the preview area. Switch between Top, Front, and Left views to inspect the data range and trajectory.
5. The preview supports zoom and pan for inspecting project details.
6. Choose the mode that matches the project scene.
7. Choose the carry method that matches the capture device.
8. Set the output path for the processing results.
9. Check Auto-load project after processing to load the project automatically once finished.
10. Check Point Cloud Segmentation and set the parameter to enable segmentation.

Note: Project Processing supports only XGRIDS handheld scanner data with firmware 3.0.2 or above. Data with firmware between 2.4.1 and 3.0.2 requires Lixel Studio v3.6.1.2; data with firmware between 1.4.0 and 2.4.1 requires Lixel Studio v3.3.1.2; data from firmware earlier than 1.4.0 requires Lixel Studio v2.4.2 or earlier. For L1 devices, firmware 1.4.0 and above requires Lixel Studio v2.5.2.1.

Mode options:

Mount method: Handheld, Backpack, Vehicle, UAV.

Basic Parameters

Basic Parameters configures point cloud optimization options during processing — including dynamic object removal, filter enhancement, point cloud enhancement, and high-precision optimization — to improve the quality and usability of results.

1. On the Project Processing page, select Basic Parameters in the left panel.
2. Adjust the parameters as needed.

A higher denoising level removes more noise but also more structure. The examples below compare filter levels.

Railing 1 (high density, fully scanned):

• Strong: More denoising, may remove part of the railing.
• Normal: Balanced denoising, more of the railing intact.
• Weak: Less denoising, minimal damage to the railing.

Filter level: Strong

Filter level: Normal

Filter level: Weak

Railing 2 (low density, incomplete scan detail):

Filter level: Strong

Filter level: Normal

Filter level: Weak

Point Cloud Enhancement: Available in 5mm and 1mm modes for L2 Pro, L3, and K2 (point spacing ≈ 5mm or 1mm respectively). Enabling this reduces processing efficiency and requires adequate hardware, especially for 1mm enhancement which needs sufficient disk space.

Region Selection: If GCPs or POIs were collected in the field, click Region Selection to enhance only the point cloud within specific point ranges.

Enhancement Range: Adjust the size of the local enhancement range, currently 1–5m.

Preview: Click Preview to view the panoramic image of each point, making it easy to confirm the location of each GCP or POI.

Note: Only selected points are enhanced locally. To restore global point cloud enhancement, click Reset.

Coordinate Transformation

Convert scanned point clouds from the original coordinate system to a target coordinate system. Import a control point file, choose a GNSS data source, set the source and target systems, and complete the transformation through control point matching. Both preset and custom coordinate systems are supported.

1. On the Project Processing page, select Coordinate Transformation in the left panel.
2. Import a control point file. Configure Skip Rows and Delimiter as needed. Check GNSS to use GNSS data.

3. Choose the GNSS data source — automatic detection from the RTK module, or a GNSS file.
4. To match scanned control points with local references, click Edit GCP. In the GCP editor, the left panel shows scanned control points and the right panel shows local reference points. Use the list at the bottom to assign correspondences, then click Confirm to save.

RTK Module: Set Horizontal Accuracy, Satellite Count, and Maximum Tilt, and choose the RTK mount type (Default or Custom).

PPK Module: Uses PPK results to enable PPK-SLAM optimization and coordinate transformation when trajectory data is valid. RINEX 3.0 and above only. For data captured in PPK mode, the GNSS module auto-selects PPK during post-processing — click PPK Settings, load the observation and navigation files, adjust base station info, select satellite systems, and click Calculate.

Note: Base station coordinates in observation files from short-duration base station sessions often contain some error. To ensure the accuracy of the PPK solution, confirm that the base station coordinates are correct first.

GNSS File: Use third-party software to transform coordinates to a target local system.

1. Copy gnss.csv from the project's device records.

2. Transform it in third-party software, then save the result back to the project's external_data folder.

3. Check GNSS and select GNSS File. The software reads the file from external_data. The required format is: gps_time, Northing, Easting, Elevation.

4. Review the source coordinate system, and click Details to view its parameters.

5. Click the target coordinate system to open the selector. Choose a preset, or click + to define a custom system (ellipsoid, projection, datum transformation, plane transformation, elevation fitting, geoid, vertical grid, horizontal grid).

6. To apply elevation anomaly correction, check Elevation Anomaly and enter a value.
7. Click Start to run the transformation, or Cancel to discard.

Note: When using GCP and GNSS together, make sure the GCP ground truth and the target coordinate system selected for GNSS are the same coordinate system. When Height anomaly is used, the software applies the control point constraints first and then the Height anomaly, so in the control point report the Z of the ground truth and the Z of the transformed coordinates will differ by the Height anomaly value.

Coloring and Mesh

Configures the visual output of project processing results, including optimized visual pose, panoramic image output, and mesh generation with the corresponding resolution and format.

Map Fusion

Map Fusion stitches multiple point cloud datasets into a single map.

For successful fusion, adjacent maps must have a valid connection:

1. Adjacent projects must overlap by at least 15m (15–30m recommended), and the overlapping paths must remain mutually visible. Plan overlaps in feature-rich scenes and avoid open areas, long corridors, and smooth tunnels.
2. For RTK-based fusion, RTK data must be valid for every project.
3. For connection-point-based fusion, adjacent maps must share at least one connection point with a matching name (a control point collected in the field) in the overlap region, and control point names must not be duplicated across projects.

Fusion limitations:

1. Up to 10 maps can be fused at once, and each capture session must be under 20 minutes.
2. A single complete full-map point cloud is output after successful fusion. When a single LAS file exceeds 180GB, it is automatically split into multiple LAS files.
3. Fusion supports only data from the same device type.

Note: If coordinate transformation is required, all maps must have valid connections (connection points or continuous scan), and at least 3 absolute control points (not collinear) must exist across all maps.

Project Settings

1. Click Add Map to import the maps to be fused.
2. The list displays each map's point name, file path, and mode.
3. Adjust per-map processing in the Mode dropdown.
4. Click the trash icon to remove a single map, or Clear to remove all.
5. Set the base map — choose Auto or designate one.
6. Set the output path.
7. Optionally check Auto-load project after processing and Point Cloud Segmentation.

Connection status: If your data has no RTK info and you forgot to collect connection points with matching names, click Connection status to configure the connections manually.

1. On the Map List page, if different maps recorded control points with matching names during capture, the software automatically builds connections from those matching connection points. If two different connection points were accidentally given the same name in the field, you can also fix the connections through manual configuration.

2. Click Configure to see the trajectory and point cloud of each of the two maps in two separate views.

3. In the left menu panel, adjust the trajectory point size, display color, and display range, or adjust the point cloud display range by elevation or time.

4. In the bottom toolbar, switch to single-view mode.

5. Under the map name, choose Rotate or Translate to transform the corresponding data and complete a rough manual registration.

6. The bottom toolbar also provides a lasso tool to keep only the part you need: use the lasso to select overlapping trajectories, or lasso-deselect to clear the selection.

7. A depth wall is also supported: data behind the depth wall is hidden and cannot be selected. To adjust the depth wall orientation, click Reset.

8. In the registration view, you can also delete connections you have already set up.

Note: When completing fusion with overlapping trajectories, make sure the two trajectory segments between the different maps are mutually visible, and that only one set of overlapping trajectories is selected between each pair of maps.

Basic Parameters

Same as Project Processing.

Coordinate Transformation

Same as Project Processing. Switch between maps to review.

Coloring

Same as Project Processing.

Accuracy Check

Verify scanned data accuracy by comparing checkpoint coordinates against a set of true values after transformation.

1. Select the point cloud data requiring verification, then click Project Processing → Accuracy Check.
2. Select checkpoints automatically or manually.
3. Review the automatically selected points, and press Esc to exit selection once confirmed correct.
4. Click Calculate to complete the calculation.

Checkpoint selection in the point cloud

Parameter settings:

Note: Automatic point selection requires XGRIDS reflective targets placed on checkpoints during acquisition. Pause briefly at each target while scanning to ensure enough target points are captured.

Manual point selection: Left-click checkpoint positions in the point cloud (a green label appears). To replace a point, reselect near it and click Yes when prompted. After selecting at least 3 points, press Esc to confirm and exit.

The software then calculates the coordinate differences, planar error, and elevation error (max, min, average) for each point pair, along with the mean square error for plane and elevation. Click Export to save the accuracy report, which defaults to the "Report" folder in the project directory.