What a $1,200 LiDAR Scanner Can Actually Do, and Why Intensity Matters
LiDAR is often associated with expensive survey-grade equipment, specialised operators, and project budgets that rule it out for routine work. The Livox Mid-360 changes that calculation — and this scan of an ordinary workspace is the demonstration.
The Scan
The scan below was captured in a standard residential workspace, with the field of view extending through an open doorway to the street frontage and surrounding vegetation. Trees at distances exceeding 30 metres are clearly resolved. The point cloud was processed and visualised in CloudCompare, with colouring applied by return intensity, not RGB. No camera. No targets. No post-processing beyond a standard FAST-LIO2 pipeline on Ubuntu.
What Intensity Colouring Actually Tells You
Every LiDAR pulse that strikes a surface returns a signal of varying strength. That strength, intensity, is a function of the material's reflectance properties, the angle of incidence, and the distance to the scanner. It is a physical measurement, not a visual one.
In the scan above, distinct intensity signatures are visible across different surface types:
Consistent mid-range return, clean and spatially uniform
Low or absent return, appears as voids or sparse regions
Moderate return, variable with surface grain and angle
Characteristically low and noisy, partial penetration through foliage
High intensity return, often saturated, clearly distinguishable
Distinct mid-to-low return, separable from wall surfaces
This means that even without a single camera pixel, the scanner is already performing a basic material classification. With the right processing pipeline, those intensity bands become the basis for semantic segmentation, automatically separating walls from floors from vegetation from structure, purely from physics. That is not a post-processing trick. It is inherent in the data from the moment of acquisition.
Practical Use Cases
These capabilities translate directly into routine built environment and civil engineering workflows:
As-Built Room Survey
Full room scan captures wall dimensions, ceiling height, window and door positions, and floor area in a single acquisition, typically under five minutes. Suitable for renovation documentation, heritage recording, and building compliance checks.
Tree Location & Canopy Mapping
Individual trees resolved at 30+ metres. Trunk position, diameter, and canopy extent extractable from a single scan position, useful for boundary disputes, arborist reports, and council vegetation audits.
Construction Verification
Scan before finishing trades begin. Verify door and window placement against architectural drawings before plastering or cladding conceals deviations. Simple application, clear cost-benefit case.
Dilapidation & Condition Records
A timestamped, spatially accurate point cloud is a more defensible record of property condition than photographs. Applicable to insurance assessments, pre-lease dilapidation surveys, and dispute resolution.
Façade & Roof Geometry
Ground-level scanning resolves roof pitch, eave geometry, and façade irregularity without scaffolding or drone permits. Surface condition anomalies become visible in intensity data alongside geometry.
Service Corridor Documentation
Intensity differences between pipe materials, conduit types, and surface coatings can distinguish service types without visual inspection. Useful for facilities management and infrastructure documentation.
Setup Notes
The Livox Mid-360 runs on Linux via ROS2, with FAST-LIO2 providing the SLAM-based mapping pipeline. Setup on Ubuntu is straightforward if you are comfortable with ROS2 workspaces, the main configuration steps involve correct IMU calibration, topic remapping, and tuning the FAST-LIO2 parameter file for indoor versus outdoor environments.
Point cloud visualisation and intensity analysis in this article used CloudCompare, which is free and open source. For batch processing and pipeline automation, PDAL handles the heavy lifting well.
Happy to share configuration notes for anyone working through the same ROS2 + FAST-LIO2 pipeline on Linux, feel free to get in touch or raise a question via LinkedIn.
The gap between entry-level and professional-grade LiDAR is narrowing quickly. A $1,200 scanner with a well-configured pipeline on a Linux machine now produces data that would have required a $30,000+ instrument five years ago.
The bottleneck is no longer the hardware. It is the workflow knowledge to extract value from the data.
That is the gap TerraIO is focused on closing.