What's a point cloud?
🕐 Read time: 5 min
Written By Clark Yuan
Last updated 8 days ago
Overview
If you are new to Stitch3D, you may be encountering point clouds for the first time. This article explains what a point cloud is, how one is created, what the data inside it means, and why it is the primary format for spatial data delivery in industries like surveying, construction, mining, and utilities.
What is a point cloud?
A point cloud is a collection of data points, each representing a precise location in three-dimensional space. Think of it as a digital dot-to-dot of the physical world: millions or billions of individual points, each one recording an exact X, Y, and Z coordinate. Together, these points form a highly accurate three-dimensional representation of a real-world surface, structure, or landscape.
Point clouds represent the foundational "raw" spatial data produced by LiDAR sensors and photogrammetry pipelines. Other outputs such as orthomosaics, 3D meshes, digital elevation models are all derived from or built on top of the underlying point cloud. This is why point clouds are preferred for precision work; they preserve the original measurement data without the interpolation or simplification that derived products involve.
Unlike a photograph, which captures a flat 2D image of a scene, a point cloud captures the actual geometry of a space. You can rotate it, zoom into it, navigate through it, and take real measurements directly from it with survey-grade accuracy.
ℹ️ A useful analogy: Imagine spray-painting millions of tiny dots onto every surface of a building at once, then removing the building and leaving only the dots. What remains is a point cloud: the shape of the building, preserved in three-dimensional space, without the building itself.
How are point clouds created?
Point clouds are generated by two main technologies: LiDAR and photogrammetry. Both produce point clouds, but they work differently.
LiDAR (Light Detection and Ranging)
LiDAR sensors emit laser pulses and measure how long each pulse takes to return after bouncing off a surface. The technology uses an airborne laser that collects billions of returns while flying, with each return recording the precise distance from the sensor to the surface. The result is an extremely dense, accurate three-dimensional model of everything the laser hits (terrain, buildings, vegetation, utility infrastructure, etc.).
LiDAR is capable of capturing surfaces that are difficult to photograph, such as ground beneath forest canopy, and produces data that includes additional attributes like intensity, classification, and scan angle alongside the raw coordinates.
Photogrammetry
Photogrammetry generates point clouds from overlapping photographs. Software analyzes common features across hundreds or thousands of images and reconstructs the geometry of the scene in three dimensions. Imagery with sufficient quality, overlap, and texture can generate very dense and highly accurate point clouds; accuracy and density are the strengths of this process.
Photogrammetry-derived point clouds carry RGB color information from the original imagery, making them visually rich and intuitive to interpret.
LiDAR vs. photogrammetry at a glance
What data is stored in a point cloud?
Each point in a point cloud stores at minimum an X, Y, and Z coordinate. Most point clouds from modern sensors and processing workflows also include additional attributes:
These attributes are what make point clouds so powerful for analysis. You can color a point cloud by elevation to reveal terrain, switch to intensity to highlight surface materials, or filter by classification to isolate bare ground.
Common point cloud file formats
Point clouds are stored in standardized file formats designed for efficient storage and transfer of large datasets:
💡 Tip: If you have LAS files, convert them to LAZ before uploading to Stitch3D. The files are significantly smaller, upload faster, and require less storage on your plan. The data is identical.
Using point clouds in other software
Once you download a point cloud from Stitch3D, it can be opened in a wide range of industry-standard tools for further analysis, modeling, and reporting:
ℹ️ Note: LAS and LAZ are the most widely accepted formats across GIS and CAD software. If a specific tool requires a different format, check the software's documentation for recommended import settings before exporting your file.
What can you do with a point cloud in Stitch3D?
Once uploaded to Stitch3D, a point cloud is instantly viewable in the browser-based Viewer — no software installation needed for you or anyone you share it with. From there you can:
Navigate the scene in 3D from any angle, on any device
Measure distances, areas, volumes, and elevations directly in the Viewer
Annotate specific locations with points, notes, and incident reports
Adjust display attributes including color, intensity, elevation, and classification to suit your analysis
Share the data with clients, engineers, or stakeholders via a secure link or QR code — no Stitch3D account required to view
Who uses point clouds?
Point clouds are used across a wide range of industries wherever accurate three-dimensional data is needed: