Scan Post-Processing – From Raw Scans to Accurate 3D Models (Terrestrial Laser Scanning)

Introduction

Industries like offshore installations, refineries, shipyards, and processing plants often use 3D laser scanning to capture precise as-built conditions in complex and tight spaces. But collecting this data is only the first step toward delivering what clients need.

The true value of a scan is realized not in the field, but during post‑processing. Raw scan data, typically consisting of massive point cloud datasets, must be processed before it can be trusted for engineering, fabrication, installation planning, or safety assessments.

3D scan post‑processing is an essential step in industrial digitalization workflows. It transforms raw measurements into clean, aligned, and correctly referenced datasets that engineers can rely on. Tasks such as scan registration, noise removal, optimization, and quality control ensure that the point cloud accurately represents real-world geometry. Any disturbing factor during 3D scanning, such as vibration, reflective surfaces, limited access, ongoing operations, or moving objects, must be taken into consideration during the post-processing process.

In industry where every downtime is extremely costly, accuracy, quality, and completeness of the data are critical. Any incorrect data can lead to clashes, rework, or costly downtime; robust post‑processing is essential. As industrial projects increasingly depend on digital twins and data‑driven workflows, well‑structured and repeatable 3D scan post‑processing has become a key factor in delivering reliable, engineering‑grade outputs.

Definition and Purpose

• 3D Scan - A 3D scan is a digital representation of a physical object, structure, or environment, achieved by measuring its geometry in three dimensions. In industrial applications, 3D scanning is predominantly conducted using laser scanners or LiDAR (Light Detection and Ranging) technology. These methods capture millions of points with coordinate X, Y, Z coling point-cloud, enabling the accurate mapping of real-world surfaces
• Point-cloud - A point cloud is a digital dataset composed of millions or billions of individual points, each characterized by precise X, Y, and Z coordinates within a three-dimensional space. Collectively, these points provide an accurate spatial representation of a physical object, structure, or environment.
• 3D Laser Scanner - A laser scanner is a tool that allows for precise measurement of the position of millions of points representing the area the device is located in. This operation is performed using a laser beam, allowing for quick measurement of a large area by assigning X, Y, and Z coordinates to each point. The data collected by the scanner is called a point cloud.
• LIDAR - Light Detection and Ranging, is an advanced remote sensing technology that employs laser light to accurately measure distances to various objects and surfaces. This methodology facilitates the creation of highly precise three-dimensional (3D) models of physical environments, thereby enhancing the understanding and analysis of complex spatial information.

Steps in Scan Post-Processing

• Point-cloud preparation
• Registration
• Quality control
• Point-cloud formats
• Application of point-cloud

Point-cloud preparation - reducing Noise and Artifacts

In the context of 3D laser scanning, noise and artifacts refer to unwanted points or distortions that fail to accurately represent the actual geometry of an object or environment. These issues are prevalent in raw point cloud data and may arise from various factors, including surface reflectivity, extensive scan distances, environmental conditions such as dust and fog, vibrations, and moving objects in the scanned area.

The process of noise reduction is a first step in the post-processing of 3D scan data. This process aims to identify and eliminate outliers, random measurement errors, and moving duplicate objects while preserving the integrity of true surface geometry and sharp features. Without proper filtering, noise will reduce the quality of final point-cloud registration and negatively impact on any other process we would like to use of the point cloud data, including 3D modeling, mesh preparation, clash detection, and dimensional control.

Basic filtering of raw point clouds is most often implemented in software provided by 3D scanner manufacturers. Industry leaders such as Leica Geosystems and Faro ensure good noise filtering and automatic or semi-automatic removal of duplicate or unwanted objects. At CAPNOR, based on our many years of experience, we have developed a methodology and created our own internal programs to optimize this process, ensuring high-quality final point-clouds.

Registration

Point cloud registration is the process of merging individual 3D scanner stations into a single, unified point cloud that represents the entire scope of work. This registration can be accomplished using surveyed control points, artificial targets, or by aligning overlapping geometry between individual scans using cloud‑to‑cloud algorithms.

Regardless of the method chosen, it is essential to thoroughly inspect both the individual scan connections and the overall integrity of the registered scan network. Local misalignments, weak constraints, or poorly distributed overlaps may not be immediately apparent at a global level, yet they can significantly affect downstream applications such as dimensional control, modeling, and change analysis. For this reason, quality control includes reviewing residuals, RMS values, overlap consistency, and visual inspection of critical geometries throughout the dataset.

Despite continuous advancements in registration algorithms and the increasing computational power available to modern software, the registration process still relies heavily on the expertise and judgment of the person performing it. Understanding the principles of operation of the scanner, target placement, and software limitations is essential for identifying potential errors that automated routines may not detect. Human supervision remains a key factor in deciding whether a registration result is not only mathematically acceptable, but also geometrically reliable and fit for purposes.

Based on our experience, the registration process has become significantly more automated and streamlined over the past 15 years, which has played a key role in the growing adoption and popularity of 3D laser scanning across industrial sectors. However, as automation increases, the importance of competent quality checks has grown in parallel. A structured quality assurance process ensures that automation enhances efficiency without compromising data accuracy, confidence, and traceability.

Quality control

In our experience, this is the most frequently overlooked step in the entire data processing process, and in our opinion, the most crucial. Given current production requirements and the short turnaround time, it might seem like skipping this step or performing it cursorily is a good idea, but in our opinion, the more thoroughly the quality control is performed, the fewer problems there will be in the future.

A structured quality check process is applied throughout scan processing and registration to verify data accuracy, consistency, and completeness.

Quality checks focus on verifying internal alignment accuracy between scans, typically evaluated through registration reports, residuals, and RMS values. Misalignments, excessive residuals, or poorly constrained connections are identified and corrected to ensure the overall registration accuracy remains within the defined tolerance limits for the selected scanning accuracy class. These requirements may be determined by our customers or the internal guidelines of each 3D scanning company.

An essential aspect of registration quality control is the validation of constraint geometry and redundancy. Each scan should be supported by sufficient overlap, multiple independent connections, and, where required, survey‑controlled reference targets. Cloud‑to‑cloud constraints with excessive error are excluded to prevent local distortions and cumulative alignment drift across the dataset.

Following internal registration, quality checks are extended to georeferencing, when required. Surveyed control points are compared against the registered point cloud to confirm that global positioning and orientation meet project and client requirements. Any deviations beyond acceptable limits trigger corrective actions or partial re‑registration to maintain traceability to the master coordinate system.

In addition to geometric accuracy, data integrity and coverage are verified during processing. This includes checking for gaps, shadowed areas, noise, unwanted artefacts, and inconsistencies in point density that could impact downstream modeling or measurements.

The final stage of quality control involves documentation and reporting. Registration statistics, accuracy classifications, and applied constraints are documented as part of the data delivery, providing transparency and confidence in the scan results. This ensures that the delivered point cloud is a reliable, high‑quality representation of the scanned asset and is ready for engineering workflows, clash detection, and digital twin integration.

Point-cloud formats

3D laser scanning generates large volumes of data in the form of point clouds. To ensure efficient data exchange, long‑term usability, and compatibility with downstream engineering tools, point clouds are delivered in standardized file formats. Each format serves different purposes depending on accuracy requirements, software environment, project scope, and most importantly, client requirements.

One of the most widely used formats is E57. It is an open, vendor‑neutral format designed specifically for 3D data. E57 supports high‑precision point coordinates, color information, intensity values, and scan metadata, making it well-suited for long‑term archiving and data exchange between different software platforms. The openness of the E57 format makes it the most frequently and eagerly used format for 3D data, allowing the creation of custom software solutions for point-cloud edition.

PTS, PTX, and XYZ formats represent simpler, ASCII‑based point‑cloud formats. They store point coordinates—and optionally color or intensity—in a plain text structure. While these formats are easy to read and widely supported, they result in significantly larger file sizes and limited metadata support. As a result, they are typically used for data exchange, troubleshooting, or compatibility with legacy software rather than as final master datasets.

For project environments centered around Autodesk products, RCP/RCS formats are commonly used. These are indexed point‑cloud formats optimized for visualization and performance in Autodesk applications such as Revit, AutoCAD, and Navisworks. RCP files act as project containers, referencing one or more RCS files, which store the indexed scan data. While highly efficient for design and coordination workflows, RCP/RCS are proprietary and are usually generated from original scan formats such as E57 or native scanner data.

Scanner manufacturers also provide native proprietary formats, such as FLS, ZFS, or others, depending on the hardware and software ecosystem. These formats preserve full scanner metadata and are used during registration and processing stages. However, they are typically converted to open or platform‑specific formats before delivery to ensure interoperability and long‑term accessibility.

In practice, professional scanning workflows often use a combination of formats: proprietary formats for processing, E57 or LAS as neutral master deliverables, and RCP/RCS or similar for design and review purposes. Selecting the appropriate point‑cloud format is a key part of a data management strategy, ensuring that the scan data remains accurate, usable, and compatible throughout the asset lifecycle.

Application of point-cloud

How we use point clouds depends only on our imagination. Increasing the availability of scanning devices and point cloud processing capabilities open up endless possibilities for users. As a company operating primarily in the oil and gas industry, we will focus on these applications, highlighting only the potential for use in other industries.

In the industrial sector, point clouds are used for as-built documentation, engineering support, and asset digitalization. Facilities such as oil and gas installations, refineries, power plants, and manufacturing sites often have complex environments where accurate documentation is essential. Point clouds offer a precise geometric reference that captures the asset's current state, including deviations from design and undocumented changes.

Point-cloud data supports engineering and design workflows by providing a reference for 3D modeling, retrofit planning, and layout verification. Engineers use point clouds directly or convert them into 3D models for clash detection, equipment replacement studies, and constructability reviews. In offshore and brownfield settings, this reduces the need for site visits, lowers safety risks, and improves design accuracy.

Another key application is dimensional control. Point clouds enable precise measurements, alignment checks, and tolerance analysis of structures, piping, and equipment. This is particularly important for prefabrication, module installation, and life‑extension projects where dimensional accuracy directly impacts cost and schedule.

Point clouds also play an essential role in digital twin and asset management platforms, where verified scan data is combined with panoramic imagery, engineering models, and operational data. This creates a spatially accurate digital representation of the facility that supports maintenance planning, inspections, and long‑term data reuse.

In construction and infrastructure projects, point clouds are used for progress tracking, quality assurance, and as‑built validation. Scanned data is compared against design models to assess construction accuracy and detect deviations early. Infrastructure assets such as bridges, tunnels, roads, and rail corridors benefit from point‑cloud‑based documentation due to the scale and complexity of their geometry.

In architecture and the built environment, point clouds support building documentation, renovation, and heritage preservation. Accurate scan data allows architects and designers to work with reliable geometry for refurbishment projects, especially where original drawings are incomplete or outdated. In heritage applications, point clouds provide a non‑intrusive method for documenting historic structures while preserving fine geometric detail.

In mining and surveying, point clouds are used for volume calculations, deformation monitoring, and terrain analysis. Repeated scans enable change detection over time, supporting safety assessments and operational planning. Environmental applications include coastline monitoring, landslide analysis, and forestry studies, where point clouds provide detailed spatial datasets for natural environments.

Point‑cloud technology is also applied in transportation, automotive, robotics, and virtual reality, where accurate spatial data is essential for simulation, navigation, and visualization. In urban planning and smart‑city initiatives, point clouds form the geometric backbone for large‑scale city models and spatial analysis.

Summary

Capnor specializes in the acquisition and processing of 3D laser scanning data. Our core expertise lies in large‑scale projects that involve thousands of scan positions captured on a single asset, which naturally creates significant challenges during data processing and registration.

In recent years, rapid advances in hardware and software have greatly improved data processing capabilities. While this progress has increased efficiency, it can also create a misleading impression that data processing is straightforward and fully automated. In practice, achieving reliable and high‑quality results still requires experience, robust procedures, and expert oversight.

With over 20 years of industry experience, Capnor has developed proven workflows, detailed procedures, and proprietary software solutions that optimize data processing while maintaining strict quality standards. Our teams are involved at every stage of the project lifecycle—from planning and field scanning, through data processing and quality control, to the delivery of final reports, registered point clouds, or 3D models.

This end‑to‑end involvement enables a holistic approach to each project and ensures a clear understanding of both the capabilities and limitations of 3D scanners and processing software.

At Capnor, we are committed to continuous improvement and to meeting our clients’ expectations by consistently delivering accurate, reliable, and high‑quality data.