How to Get the Best Mesh Results?

Introduction

Generating a high-quality mesh depends primarily on how your data is captured, processed, and prepared before upload.
The most important factors are:

• The acquisition method: Static or Mobile (TLS vs. MLS)

• The input format

• The preservation of structure and metadata (scan positions, trajectory, panoramas)

Identify Your Capture Method

Before preparing your data, determine how it was captured:

• TLS (Terrestrial Laser Scanning) → static scan positions

• MLS (Mobile Laser Scanning) → continuous movement with trajectory

This decision defines the correct workflow and input format.

Use the Correct Input Format

Choosing the right format is critical for mesh quality.

Recommended formats:

• TLS data → use LGSx

• MLS data → use B2G

Supported alternatives:

• Structured E57 → acceptable

• MLS in LGSx → supported, but requires special configuration

Prepare TLS Data (Static Scanning)

For TLS workflows, follow these best practices:

Acquisition

• Use dense and regularly spaced scan positions

• Ensure good overlap between scans

• Minimize occlusions

• Capture panorama images

Data preparation

• Keep the dataset structured

• Clean the data (remove noise and reflections)

• Export directly from Cyclone REGISTER 360 PLUS

Prepare MLS Data (Mobile Scanning)

For MLS workflows, the trajectory and coverage are key.

Use B2G (Recommended)

B2G is the preferred format because it preserves:

• The scanner trajectory

• The continuity of acquisition

Acquisition best practices

• Maintain a smooth and continuous trajectory

• Avoid tracking loss

• Use loop closures where possible

• Capture areas multiple times from different angles

Special Case: MLS Data in LGSx

Using LGSx for MLS data is possible, but requires careful configuration.

Critical requirement: waypoint spacing

When importing MLS data into Cyclone REGISTER 360:

• Default waypoint spacing = 50 m → not suitable

• Required waypoint spacing = 1–2 m

Why this matters:

Correct waypoint spacing ensures:

• Stable geometry

• Accurate normal estimation

• Smooth mesh surfaces

• Better texturing

Incorrect spacing can lead to:

• Artifacts

• Loss of detail

• Inconsistent geometry

• Poor mesh quality

Using E57 Data

E57 can produce good results, but only if structured.

Works well:

• Structured E57 (scan positions preserved)

Should be avoided:

• Unstructured E57 (single merged point cloud)

Ensure Overall Data Quality

Regardless of the workflow, always verify:

Point cloud quality

• High density

• Clean and filtered

Coverage

• TLS → sufficient scan overlap

• MLS → continuous trajectory and multiple passes

Metadata (critical)

Ensure the following is preserved:

• Scan positions (TLS)

• Trajectory (MLS)

• Panorama images (if available)

Common Mistakes to Avoid

• Using unstructured point clouds

• Applying unnecessary format conversions

• Losing metadata (trajectory, scan positions, panoramas)

• Keeping default 50 m waypoint spacing for MLS

• Insufficient coverage or scan overlap

Final Recommendations

For best mesh results:

• Use LGSx for TLS workflows

• Use B2G for MLS workflows

• Keep data structured

• Preserve all metadata

• Avoid unnecessary processing steps or conversions