5.1 Map Fusion Fundamentals

What Is Map Fusion

Map Fusion is a post-processing mode in LixelStudio that combines multiple independent L3, L2 Pro, or K2 scan segments into a single unified point cloud. Each segment is processed through SLAM optimization individually, then the fusion algorithm aligns and merges them using shared spatial reference: RTK coordinates, matching control point names, or both. The L3 does not require control points for Map Fusion: its sessions are matched by region selection. RTK and shared control points still work on the L3 when a job calls for them.

It is an alignment tool for sequential segments that together cover a complete site, not a stitching tool for redundant data. Segments must share spatial reference at their connection zones; that reference is what locks them together accurately.

When to Use It

Projects Exceeding Battery Life

The L2 Pro and K2 each provide up to 90 minutes of scanning time per battery charge. Large facilities, warehouses, campuses, and multi-floor commercial buildings require more time than a single battery supports. Plan segment breaks at logical transition points and join the segments in post-processing.

Projects Exceeding RAM Capacity

LixelStudio specifies 64 GB at both its minimum and its recommended configuration and publishes no memory-to-duration figure. The duration figures belong to LCC Studio reconstruction: 64 GB handles 30 minutes of capture data reliably, and 128 GB handles 60 minutes of capture data. If project scope exceeds your hardware, process shorter segments that fit in RAM, then fuse the results.

Multi-Floor Buildings

Moving between floors on stairways creates SLAM stress: narrow geometry, repetitive structure, brief feature interruption. Breaking at floor transitions and fusing produces more reliable results than carrying a single session through multiple stairways.

Long Linear Environments

Tunnels, mine galleries, and corridors exceeding 1,640 ft (500 m) are not suitable for a single Narrow Space session. Narrow Space raises the success rate within that distance; beyond it, break the run into segments of 1,640 ft or less and fuse them with relative control points. Field execution for this scenario is covered in 5.3 Field Collection.

Recurring Scan Projects

Facilities that require periodic rescanning (construction progress monitoring, facility management, asset tracking) use Map Fusion to merge new scan segments with existing baseline data. Only the changed zones need recapture; the new segments fuse with the unchanged baseline through shared control points.

On the L2 Pro and K2, this requires control point infrastructure planned into the baseline scan from day one. Control points cannot be added to a completed scan after the fact. If the baseline has no shared control points with the future rescan zones, fusion is not possible. See 5.2 Pre-Project Planning for permanent control point requirements.

Hard Limits

These are system constraints, not recommendations. Exceeding them causes fusion to fail or produce unusable output. Every figure here is the LixelStudio point cloud pipeline for the L2 Pro and K2.

• Maximum 10 maps per fusion job
• 200 minutes combined across the whole job, the same total LCC Studio enforces
• Each capture session under 20 minutes (target 12 to 15 for buffer). Longer segments demand significantly stronger processing hardware
• Minimum 50 ft (15 m) overlap between consecutive segments, with 50 to 100 ft (15 to 30 m) recommended
• At least one shared control point per junction with a matching name in the overlap region, or valid RTK instead. Plan 2 and use 3 where alignment quality matters; Narrow Space uses 2
• Same device type per fusion job. L2 Pro and K2 segments cannot be mixed, and 16-line and 32-line L2 Pro units cannot be fused together
• Coordinate transformation requires at least 3 absolute control points across all maps, not on a straight line

Connection Methods

On the L2 Pro and K2, Map Fusion uses one of three methods to establish alignment between segments. The method determines whether the fused output has global coordinates, relative coordinates, or no georeferencing. This decision must be made before collection begins; it cannot be applied retroactively.

The Four Valid Connection Patterns

Official documentation defines which configurations the fusion algorithm can resolve. Every L2 Pro and K2 segment must be linkable to at least one adjacent segment through one of these patterns.

1. All segments have RTK
2. No RTK: consecutive pairs share identically named control points
3. Segment 1 has RTK; segments 1 and 2 share control points; segments 2 and 3 share control points
4. Segments 1 and 2 have RTK; segments 2 and 3 share control points

A segment with no connection to its neighbors cannot be fused.

What Comes Out

The output of a successful Map Fusion job is a single unified point cloud spanning all input segments, saved as sub-maps in a separate result folder. If any segment carried valid RTK, the output has global coordinates. If only control points were used, the output has relative coordinates: internally consistent geometry with no absolute position.

The fused result exports from LixelStudio as LAS, LAZ, PLY, or E57. For Autodesk workflows, the built-in LAS to RCP converter handles up to 10 LAS files and targets AutoCAD and Recap 2020 or later, so no separate ReCap Pro license is required. The same export selection and quality considerations that apply to single-scan outputs apply here.

Processing time for Map Fusion is substantially longer than single-scan processing, and hardware requirements are higher.