8.4 Quality Assessment and Point Cloud Editing
Verifying point cloud density, coverage, and coordinate accuracy using LixelStudio's Accuracy Check, then preparing data for delivery with clipping, denoising, slicing, and measurement tools.
Quality Review Before Export
Quality review in LixelStudio happens after processing and before export. Time spent verifying the point cloud catches problems that would otherwise surface hours into a downstream Revit or AutoCAD workflow, where fixing them requires returning to the source data or rescanning entirely.
- Check point density across the full model. Sparse or holey areas correspond to zones where scanning speed was too high, loop closure was incomplete, or the device was angled away from the geometry. If Point Cloud Enhancement was enabled during processing (5 mm or 1 mm spacing), verify that the enhanced areas rendered as expected.
- Verify coverage against the scanning route plan. Walk the point cloud room by room. Zones that were skipped in the field are missing entirely. There is no processing fix for uncollected data.
- Measure a known dimension. Pick a wall-to-wall or floor-to-ceiling distance you measured in the field or know from drawings. A 16 ft (5 m) corridor reading as 26 ft (8 m) indicates trajectory drift, not a measurement tolerance issue. The scan has a fundamental accuracy problem and likely requires rescanning. See 8.2 Single Scan for processing configuration that affects accuracy.
- Run the Accuracy Check if coordinate transformation was performed using control points or RTK. This is found under Project Processing > Accuracy Check in LixelStudio. It compares checkpoint coordinates against their surveyed true values and produces coordinate differences, planar error (max, min, average), and elevation error (max, min, average) for each point pair. The report exports to the "Report" folder in the project directory.
- Check for dynamic object artifacts. Moving objects scanned during acquisition (people, vehicles, equipment) appear as smeared or ghosted geometry. Operator body parts in the scan path produce a distinct repeating streak that follows the trajectory. Dynamic Object Removal in processing settings handles most of these automatically, but persistent obstructions may require manual clipping before export.
- For Map Fusion projects, inspect every segment boundary. Discontinuities at join zones indicate insufficient overlap or control point misalignment between segments. Map Fusion requires 50 ft (15 m) to 100 ft (30 m) of overlapping path in feature-rich areas and at least 1 shared control point between adjacent maps. See 8.3 Map Fusion for connection requirements.
Accuracy Check Workflow
LixelStudio's Accuracy Check (Section 5.3 in the user manual) verifies coordinate accuracy by comparing transformed checkpoint positions against their surveyed true values. This is available only for projects that used ground control points or RTK/PPK for coordinate transformation.
- Select the processed point cloud in the data layer, then go to Project Processing > Accuracy Check.
- Select checkpoints. If XGRIDS reflective targets were placed on checkpoints during acquisition, the software can select them automatically. Otherwise, left-click checkpoint positions manually in the point cloud (minimum 3 points).
- Import the checkpoint coordinate file (.txt or .csv). Format: Point Name, Easting, Northing, Ellipsoidal Height.
- Click Calculate. The results show coordinate differences for each point pair, plus planar error and elevation error statistics (max, min, average, and mean square error).
- Export the accuracy report to the Report folder for project records and client delivery.
Important Automatic target selection depends on scan quality at the target. Pausing briefly at each reflective target during field acquisition improves the density of points on the target surface, which directly affects whether LixelStudio can locate the target automatically during Accuracy Check.
The Accuracy Check report and the Map Fusion control point accuracy report are separate outputs. Map Fusion generates its own report in the Report folder when coordinate transformation is configured. Both should be reviewed for projects that use Map Fusion with absolute coordinates.
Point Cloud Editing Tools
LixelStudio provides editing tools under the Tools menu (Section 6 in the user manual) for preparing point cloud data before export. Most tools save their results as new point clouds rather than modifying the original processed data. The Align tool is the exception: it changes the point cloud's coordinates directly.
Clipping
LixelStudio has two clipping tools. Clipping Box (6.10) uses a 3D bounding box that can be translated, rotated, and scaled by dragging control handles. Use it to isolate a volume, such as extracting a single floor from a multi-story scan. Clipping (6.11) uses 2D rectangle or polygon selection. It offers 4 selection modes: Rectangle Selection, Polygon Selection, Rectangle Delete, and Polygon Delete. After selecting an area, choose Inner Clip (keep selected, delete rest) or Outer Clip (keep rest, delete selected).
A full-facility scan delivered to a contractor who needs only the mechanical room is not useful delivery regardless of data quality. Clip to the required area before exporting rather than sending the entire dataset. See 8.5 Export Formats for format-specific delivery guidance.
Noise Filtering and Denoising
Noise removal happens at two stages in LixelStudio, and it is important to understand which one applies to the situation.
Processing-time Filter Level is set in Advanced Settings before processing begins (Section 5.1.3). It has 3 levels: Strong, Normal, and Weak. Strong removes more noise but may strip fine structural detail such as railings. Normal provides balanced results. Weak preserves the most geometry with minimal denoising. For high-density scans with fine detail, use Normal or Weak. For low-density scans, Weak is safest because Strong can remove real geometry that is already sparse.
Warning The processing-time Filter Level cannot be changed after processing completes. If the wrong level was selected, the project must be reprocessed from raw data with the corrected setting. Choose the filter level before clicking Process.
Post-processing Denoising (Tools > Denoising, Section 6.4) is a separate tool applied to already-processed point clouds. It uses neighborhood analysis with 2 configurable parameters: Neighborhood Points (number of adjacent points analyzed per point) and Standard Deviation Multiple (noise sensitivity threshold; smaller values classify more points as noise). The result is auto-imported as a new denoised point cloud. For sparse point clouds, increase the Neighborhood Points value to avoid stripping real geometry.
Pro Tip: Start with the Denoising tool's default parameters and inspect the result before adjusting. Aggressive settings remove real edge geometry along with noise, particularly on thin elements like handrails, cable trays, and pipe runs. Compare the denoised result against the original to verify nothing critical was removed.
Horizontal and Vertical Slicing
Horizontal Slice (6.14) generates horizontal cuts through the point cloud for floor plan creation. The tool displays 3 views and slicing is performed in the front view. Default slice thickness is 0.5 m (1.6 ft). The Step function auto-generates multiple slices at a configurable spacing (default 0.5 m). Each slice can be previewed in a floating window before export. Output formats are LAS and RCP.
Vertical Slice (6.15) generates vertical cuts for elevation and section drawings using the same workflow in the top view.
Coordinate preservation: Slicing uses the internal view adjustment and does not change the point cloud's coordinates. If the point cloud needs to be rotated before slicing (for example, to align walls with the slice plane), use the Align tool first. Note that Align permanently changes coordinates in the saved output.
Measurement and Data Recording
Profile Analysis (6.9) provides cross-section viewing and measurement within a selected rectangular area. Set a buffer width (for example, 0.1 m) for fixed-width selections. The measurement tool within Profile Analysis supports click-and-drag distance measurement. Supports profile views from any angle as of v3.5.
Data Recording (6.16) combines measurements, text annotations, and screenshot capture into a single workflow. Annotations support English characters and numbers only. After completing annotations and measurements, configure a report with organization name and screenshots-per-page setting, then export. The output includes a screenshot folder, an annotation CSV with coordinate data, and the complete report document.
Panorama Overlay (6.12) is available for point clouds colored using panoramic cameras. It provides measurement tools (point, distance, angle, area, multi-point) within the combined point cloud and panoramic image view. Floor slicing within this tool lets you analyze data level by level.
Note LixelStudio measurement tools are for quality verification, not survey-grade deliverables. Use them to spot-check known dimensions against the point cloud before export. Formal measurement deliverables should come from the downstream modeling or survey software.
Plane Draw (v3.5)
LixelStudio v3.5 introduced Plane Draw, which allows creating line work directly on point clouds or floor plans. It includes AI-assisted vector extraction to help convert 3D point cloud data into 2D drawings. Extracted vectors can be refined using split, delete, and trim tools, then exported as DXF files for use in CAD workflows.
Additional Tools
Resampling (6.3) reduces point count for delivery. Useful when a client's software cannot handle the full-density point cloud or when file size constraints apply. Smoothing (6.5) applies a neighborhood radius algorithm to even out surface irregularities. Use default parameters for structured point cloud data. Merge (6.6) combines multiple point clouds (LAS v1.1 to 1.4) into a single file with configurable attribute selection.
Supplementary QA with CloudCompare
CloudCompare is a free, open-source point cloud viewer and editor that handles LixelStudio export formats (E57, LAS, LAZ, PLY) without conversion. It is not a replacement for LixelStudio's built-in tools, but it provides additional review capabilities and is useful when a client or collaborator needs to inspect point cloud data without an Autodesk license or LixelStudio installation.
Misalignment and drift artifacts that are subtle in LixelStudio's viewer can be more visible when navigating the same data in CloudCompare. Generating horizontal cross-sections at multiple floor elevations is a fast way to check for slanted walls or floors caused by tracking drift. Cloud-to-cloud distance comparisons between overlapping scans can also reveal systematic offsets that are difficult to catch visually.
©2026 Alpine Reality Capture LLC • XGRIDS Pro Guide™ • Site Disclaimer