ContextCapture Promises Accurate, High-resolution Reality Modeling Option for Infrastructure

The building industry has been seeking ways to build a model of as-built or existing conditions more accurately and quickly, and then integrate it into a design or use it for the beginning of a design. Each step in that direction provided more accuracy, resolution and speed, with some cost reduction. But the march to find something that can be done more simply, and without using expensive and cumbersome equipment continued.

City of Paris

One of the more interesting products released to General Access at Bentley’s Year in Infrastructure Conference, held this week in London, was ContextCapture, the first product release of the Acute3D software technology Bentley acquired earlier this year.

City of Graz, Austria

According to Huw Roberts, vice president of platform marketing with Bentley Systems, ContextCapture is changing the way architects, engineers, contractors, and owner-operators think about capturing real-world conditions around buildings, plants, bridges, or any type of infrastructure. The software provides highly detailed models of virtually any size or resolution, up to city scale, using simple photography. Depending on the level of accuracy needed, the photos can even be taken with a smartphone.

Plant model

ContextCapture generates a detailed reality mesh using the referenced photography. This results in a navigable 3D model that includes fine and photorealistic detail, sharp edges and precise geometric accuracy. ContextCapture is not only used for design, but for monitoring, surveying and inspection, and for geospatial use. This capability was able to generate a high-resolution 3D model of the city of Stockholm, which was then merged with GIS data to become a key tool for the city’s decision-making and as a way to communicate with its residents.

ContextCapture has native support within MicroStation CONNECT Edition and is available with the Bentley SELECT subscription program. It has level-of-detail generation and adaptive tiling, plus surface and volume measurement. There is also a Web browser plug-in to view models embedded in webpages, a desktop viewer application, and a way to distribute models as self-contained programs call ‘live cubes,’ within which users can view and navigate models without requiring any other software.

Users are able to deploy the software within their design and engineering environment, for all manner of design, documentation, and visualization. “By having ContextCapture live and active in your design environment, the relationship between ContextCapture and BIM activity gives you the real-world environment represented in a realistic way as context for your engineering work,” said Dr. Jean-Philippe Pons, co-founder of Acute3D, and now director of software development with Bentley. “Plus you can add engineering data and perform a variety of analysis – such as solar, volumetric, line of sight, traffic simulation, and so on – all within the context of a highly accurate model of the real-world condition.”

“In ContextCapture you can automatically create your reality mesh from a collection of photographs and then immediately utilize it within your engineering modeling application,” said Roberts. “Once it’s in that environment, you can work as you normally would, and incorporate the reality mesh into your design and documentation efforts, such as attach data attributes, measure, cut sections, detect clashes, and of course create great visualizations. You can also use it as a MicroStation mesh, and modify, edit, and map it to property data, and so on. When used as a MicroStation mesh, you can even include it in an i-model, for example to review in Navigator.

My immediate thought was that ContextCapture would supplant the need for laser scanning to a large degree. Pons and Roberts agreed.

“The difference between ContextCapture and LiDAR and laser scanning is that with ContextCapture there is no need for special equipment,” said Pons. “You can use regular cameras and computers with ContextCapture. The reality mesh has the same accuracy, precision, and quality that you get with point scan, and we expect people to use it more often.”

“ContextCapture can easily be much more integrated to the architectural engineering construction activity,” said Roberts. “The experience of interacting with a reality mesh is more like what you see in the movies – especially when combined with our new LumenRT. The interaction is much more immersive, and much closer to models you work with in design applications. Reality meshes also have awareness of appropriate level of detail. When you are viewing close to an area it will include and display the full level of detail, and as you move your viewpoint away it simplifies the model to match what’s needed – resulting in great performance, while maintaining full fidelity within the file.”

MicroStation knows what is needed for the view of the camera, so only the needed pieces of the reality mesh can be used without sacrificing any accuracy. As to accuracy, it’s really just a function of the scale and how many pixels and how many photo vantage points are used. With sub-millimeter accuracy, it is as accurate as laser scanning for typical infrastructure uses. Acquisition time is much shorter than for laser scanning, so that’s an area where time is saved.

Also, when scanning at a construction site, for example, you may be using an expensive scanner looking only in straight lines from one point, which limits your view of the construction site. With ContextCapture, you are using photos that give you a wider perspective of the area, and doesn’t limit you to five or ten viewpoints.

“There are some cases where laser scanning might be the better option, such as when requiring fine accuracy over a long distance, or at night where photos may be not feasible,” suggested Roberts. “Or if you want to scan ground through the leaves of trees, such as with LiDAR. But I can envision that for a majority of typical cases ContextCapture will be a preferable option.”

Pons said LiDAR and ContextCapture can be used together. A reason to use laser scanning with reality mesh, is to bring in data used by other consultants, or to include legacy data.

“Accuracy from a technical point of view, is driven by pixel size. From a reasonable distance, with a reasonable camera, ContextCapture provides you with more accuracy than you’ll typically need,” Pons said. “Using ContextCapture is often the better option because you can easily have more vantage points and therefore no occlusion. You can include the backside and underneath of things, resulting in a more complete model.”

With ContextCapture, the integration of Acute3D and Bentley, as tied into engineering and the rest of the Bentley products, takes this functionality to another level and opens up countless possibilities.

Bentley also announced General Access to ContextCapture Center for “grid computing” power to expedite processing over multiple computers on your network or in the cloud, dramatically reducing processing time, and supporting very large models that involve more than 30 gigapixels of imagery.

Tags: Bentley Systems, ContextCapture, YII2015

Categories: 2D, 3D, 3D PDF, AEC, AECCafe, Bentley Year In Infrastructure. YII2015, BIM, Cloud, collaboration, construction, engineering, field, file sharing, geospatial, GIS, mobile, site planning, sustainable design, terrain, traffic simulation, virtual reality, visualization

This entry was posted on Thursday, November 5th, 2015 at 10:10 am. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.