3D point cloud software in the geospatial industry has been steadily improving as LiDAR and 3D imaging technologies have rapidly evolved. New innovative breakthroughs have now made it possible to move from traditional, limiting data management techniques to instantaneous data access and visualization.
Recent technological developments such as Semi-Global Matching (SGM) and Structure from Motion (SfM) are creating enormously dense and massive 3D point clouds. These technologies along with hardware advancements of terrestrial, mobile, and aerial LiDAR scanners are generating huge datasets that contain an enormous amount of exploitable detail. End users are often not able to take advantage of all the data.
A new revolutionary approach offers a solution to access and visualize unlimited amounts of data density and size in real-time on a typical workstation or laptop. By providing real-time visualization and 3D data retrieval, decision makers are afforded unbounded access to all full resolution data for rapid and enhanced decision making.
This solution now provides the ability to host 3D point cloud data in a central repository for enterprise dissemination and visualization. Fast and usable internet streaming of dense 3D data is also now available. As this technology matures, user specific data analysis tools will be developed providing instant data access and manipulation allowing for improved data quality and more efficient project turnaround. The grey line between real world and virtual world is rapidly becoming smaller.
Computer power doubles on average every four years, but computer storage requirements increase by up 200 times every four years. The world has moved from megabytes to gigabytes (1000 megabytes) then from gigabytes to terabytes (1000 gigabytes), and already we are hearing the word “Petabyte” (1000 terabytes). This presents challenges for the visualization and sharing of large volumes of Big 3D Data.
- “Big Data is a pressing issue for the National Geospatial-Intelligence Agency (NGA),” and “expects storage requirements to grow exponentially over the next decade because of planned intelligence collection sensors.” Canadian GIS and Geomatics
- “…Big Data” is emerging as a major research theme for the academic community.” ESRI
- “…the Department of Defence alone is investing $250M annually into Big Data research projects.” Sensors and Systems
Most 3D point cloud applications apply traditional (and limited) display and rendering methods to visualize massive amounts of data. These methods include familiar techniques such as loading points into RAM, low resolution block technique, and point decimation. Euclideon Pty Ltd of Brisbane, Australia has solved some of the biggest limitations facing users of point cloud data, saving time and money, increasing productivity, and changing the ways that it can be used.
At the heart of the technology is a 3D search algorithm that can render unlimited quantities of point cloud data in real-time. Because this algorithm efficiently retrieves only one point for every screen pixel, it can display models of previously unimaginable sizes at interactive frame-rates, without the need for a powerful CPU or graphics card. By removing the hardware bottleneck from the process of 3D visualization, Euclideon has simultaneously removed any limitations on the amount of detail that a scene can contain.
Euclideon’s Unlimited Detail technology (UD), used in its Geoverse software, is based on the concept of atoms in 3D space – that is, storing objects as collections of 3D pixels. This is a better way of representing real-world objects, as it doesn’t ‘fill in’ objects with homogenous textures and repetitive shapes. Rather, it represents every point of a scanned object just as it appears in the real world. Just as Google or Yahoo are search algorithms that go searching for words on the Internet, Geoverse technology looks for points in point cloud data. It only retrieves one point per screen pixel and it retrieves them straight from the hard drive so there is no need to load data into RAM.
The power of this technology in the current Geoverse COTS software is apparent, allowing users to rapidly disseminate their data across the country from one central storage server. Developers of analytical geospatial software are now embracing the new breakthrough. Hexagon has their new point cloud format – .HPC for streaming across the Internet.
Merrick & Company is currently developing its long standing LiDAR production software MARS© to incorporate the Unlimited Detail (UD) technology which will dramatically increase production times and allow our users to more efficiently manage their LiDAR projects. Other special development geospatial projects are ongoing such as the Ball Aerospace / Merrick & Company “Server in the Sky” – placing UD at the sensor head for real-time streaming from the sky to ground.
The future is exciting and the possibilities are endless. Scan away, we can now handle any amount of point cloud data!
Please contact Josh Beck, GISP for additional information – Josh.Beck@Merrick.com