}

Point cloud processing:
result that can be used from raw data

A laser scanner The raw point cloud captured by cannot yet be used directly. It contains noise, reflections, unnecessary points and is not optimized for design software. innoScan 3D's proprietary processing pipeline brings data into a clean, optimized, import-ready form. Tailored to your software and your needs.

Request a quote for custom processing

Why process the point cloud?

The laser scanner captures millions of points per second. This represents a huge amount of data that is almost useless when unprocessed:

Noise and reflections: the scanner captures all surfaces, including false points reflected from glass, metal.

Incompatibility: parts made from dozens or hundreds of instrument positions must be integrated into a single coordinate system.

File Size: the point cloud of a medium office building can be measured in hundreds of GBs, the design software cannot open it.

Format problem: the raw data format may not be compatible with your software.

Processing is what makes data a usable tool.

Our processing steps

1. Cleansing

We remove reflections, noise, moving objects (people, vehicles) and unnecessary points. Cleaning is partly automated, partly manual, and critical areas are finalized by manual inspection by our colleagues.

2. Matching

Details made from instrument positions are combined into a single coordinate system. The matching accuracy is validated by both automated and manual verification. The multi-stage quality control system ensures that inaccuracies from the fitting do not accumulate.

3. Optimisation

We use proprietary algorithms to reduce file size while maintaining visual quality and geometric accuracy. This is especially critical for large projects: the point cloud of a 40,000 m² office building is terabyte in size unprocessed, and after processing, the design software receives it as a manageable file. One building survey Large amounts of data recorded during this process become manageable for design offices.

4. Segmentation

We divide the building into parts: we make separate files per floor, room, area. You do not need to load the point cloud of the entire building into the software, but only the part you are working with. We also provide a consolidated point cloud for the entire area.

The direct input of the purified segmented point cloud is Scan-to-BIM Modeling and supports mechanical systems separate treatment as well.

5. Eksport

We prepare the final files in the target software-specific formats. We adapt the format and optimization to the specifics of the software.

Pontfelhő feldolgozás számítógépen

Own algorithms and unique solutions

Why don't we work with generic trading software? The point cloud processing software available on the market (CloudCompare, ReCap, Cyclone) are designed for general purpose. However, for high volumes, custom requirements and browser-based rendering, we run into limitations.

That is why we have developed our own workflow in three main areas:

Display Optimization: a browser-based viewer We optimize the data separately for you. Balance of speed and dot density: the viewer loads quickly, but details are not lost.

Export by building part: automated segmentation, from which separate and merged files are created in a single step.

Automated Quality Control: the fitting and cleaning steps are also algorithmically controlled in addition to manual validation.

CAD software compatibility

ArchiCAD

.e57, .xyz format. Point cloud divided into building parts, optimized for import.

Revit

.e57, .rcp format. Processing adapted to LOD level, can be imported natively.

Tekla

.e57, .las format. Optimized output for structural surveys.

AutoCAD

.dwg, .e57 format. Both 2D and 3D, reference point alignment.

The for architects We adapt custom exports to the design workflow so that the import can be used immediately.

Frequently Asked Questions

In what format do I receive the point cloud?

According to your software: .e57, .xyz, .las, .rcp, .dwg. Multiple formats are provided in parallel at no extra cost.

The direct basis of the processed point cloud is implementation plan It can be used for the preparation and documentation of complete building survey projects.

What are the expected file sizes?

There is typically a size difference of 3 to 10 times between the raw data and the processed result. A 200 m² apartment has a point cloud of a few GB, and a 40,000 m² office building can have up to 1 TB in raw data. After processing, the size of the files intended for the design software becomes well manageable.

How long is the processing time?

It is typically 3-10 working days depending on the size. For large projects (10,000+ m²), this can take 2-4 weeks. The exact timeline is included in the offer.

Do you process scan data from other companies?

Yes. If the on-site survey is made by another company or with our own instrument, we can take and process the data as long as the format and quality are appropriate, either from a drone survey derived data as well. We carry out a preliminary data check.

What happens if there is too much noise in the data?

The cleaning pipeline is partly automated and partly manual. For extremely noisy data (poor surface reflection, moving objects), the manual rate increases, which can also increase processing time and cost. Preliminary evaluation is carried out at the request for quotation.

Request for quotation

For all requests for proposals, we provide a free, complete offer tailored to the project, contact us with confidence!

Request for quotation for 3D laser scanning