EGS technology represents the foundations for the development of high-end, powerful acquiring software used in architecture and urban management for large buildings 3D model generation.
In architecture and historical heritage management, most of the conservation techniques are based on a thorough knowledge and analysis of the pre-existing condition of things. The successful outcome of any intervention is strongly dependant on the quality of the starting information. One of the most interesting and promising surveying method is 3D scanning. The possibility of having a full, detailed and accurate model to study and elaborate is a great opportunity for conservation, restoration and enhancement.
Virtualgeo, company with a 30-years experience in the field, offers 3D landmarking solutions for very large objects, such as buildings, complexes and land features. A series of scans are carried out, each acquiring in detail a section of the whole object or complex, then the series of acquisitions, described as point clouds, are elaborated and connected to obtain a final model. The load of information to be processed is always extremely high.
This has always represented one of the most noticeable limitations in this technology: the quantity of data acquired, several billion points on average, can be so high that the construction of a 3D mesh obtained by merging the hundreds of point clouds can require weeks of computer processing time to be achieved, and could result in digital file so big and cumbersome that only the most advanced machines available could be fast enough for allowing further elaborations. The cost of the entire operation would be immense and uneconomical. The problem faced by Virtualgeo was challenging: to find a quick and accurate software system to process large amounts of data and to store them in a format that could still be useful. Virtualgeo knew that its task was to develop a software to match those seemingly incompatible requirements, so decided to join its know-how to that of EGS, active and specialized in the CAD field for 15 years.
The collaboration with EGS proved a success. EGS provided the software libraries upon which Virtualgeo constructed its solution: CloudCube. CloudCube uses a series of algorithms to match the point clouds and reduce its complexity without losing any accuracy. This is possible through scaling. The amount of accuracy required from an architectural 3D scan can be measured in millimetres: an error of one mm or less on surfaces up to tenths of metres in length and on tools which might have bigger errors is negligible. The scanners generally used for acquisition, however, are far more accurate than that. Basing CloudCube on EGS libraries granted the developers a solid base for fast processing, and allowed them to concentrate on matching the 1mm threshold of accuracy for the surfaces, maintaining a perfect level of detail but cutting the number of required points and polygons to a minimum.
The result is staggering: billions of points are reduced to less than 1 million polygons, and the complete and detailed 3D model generation are both fast to navigate and visualize and easy to archive in small files. The models at their minimum size, without any photograph included, can be less than 30 MB large.
The libraries provided by EGS were a crucial part of the success. One of the most important features is the speed at which the point clouds are matched together into the final model; slowing down and dilating the total working time, this part is one of the most expensive in the whole process. EGS libraries are also the base for the point clouds management system of CloudCube. Virtualgeo was able to make this one of the strongest advantages of its own software. Only a few days are required to process over a record-breaking 79 billion points and more than 1000 scans into a single, quick to visualize model.
The software offers other very important advantages. The model creation process does not outputs a single mesh, which is of little use in architecture, but provides the possibility to divide the model into a full and articulated set of superposed layers, allowing for fast and accurate measurements and easy accessibility of the interior areas. It is also possible to assign to every part a different material code, granting a quick and visual access to one of the most vital and useful piece of information in architecture.
“Segmentation is one of the main features of our software”, says Mr. Erminio Canevese, founder and president of Virtualgeo, “along with accuracy, acquisition and elaboration speed, for which EGS technology has shown crucial”. In the past years, the solutions developed through EGS technology was applied by Virtualgeo to several projects, for both Italian and foreign companies and institutions and has achieved several important results. Some of the most noticeable works are the scanning and rendering of the Peter’s Gate in the Peter and Paul fortress in San Petersburg (Russia), the complete 3D model generation of the Convent of Santo Domingo in Tecpatàn (Mexico), the scanning of the Red Castle in Tripoli (home to the Libyan Archaeological Museum), which extended for 38.000 square metres and required 45 billion points, and the 3D model of the Baku Opera and Ballet Theatre (Azerbaijan), composed with the data collected over more than 850 scans. Italy plays a very important role too: two of the most relevant works are located in Italy, and are the 3D model of the Ospedale al Mare, in Venice, which required more than 70 billion points to describe all the 29 buildings of the complex, and the measurements for the Autodromo del Mugello, whose asphalt layer was redeployed according to the new information provided by Virtualgeo.