The Bruno Kessler Foundation participationin 3D-ICONS

As part of our series of partner highlights this issue features the Bruno Kessler Foundation (FBK) in Trento, Italy, which with its 3DOM research unit participates as a content provider in the 3D-ICONS project. This article details the work FBK is doing.

The Bruno Kessler Foundation (FBK – www.fbk.eu ) in Trento (Italy) is a research centre established by the government of the Autonomous Province of Trento. FBK conducts interdisciplinary research in Information Technologies as well as of other areas of sciences and technologies and of humanities. FBK employs 350 researchers, graduates, and post-doc students. FBK has generated 14 spin-offs, start-ups and joint-ventures, has developed a broad network of domestic and international alliances and strategic partnerships, hosted international congresses, conferences and symposia. Within the FBK research units, the 3D Optical Metrology (3DOM) research unit (http://3dom.fbk.eu ) has its expertise in 3D data collection and modeling, active and passive sensor characterization and integration as well as geo-referenced and geospatial data analyses.

The 3D Optical Metrology (3DOM) unit of FBK Trento is working within the 3D-ICONS activities as content provider. FBK-3DOM will deliver 3D contents and metadata belonging to different heritage sites, monuments and objects located all over Italy: medieval castles located in the Trentino region, archaeological artifacts located in different Italian museums (Trento, Paestum, Chianciano, etc.), underground Etruscan tombs, Greek temples in Paestum, etc. FBK-3DOM will provide only reality-based 3D models, so digital 3D representations produced starting from data acquired on-site using active or passive sensors.

The reality-based 3D digitization of the declared monuments and objects followed some fundamental steps necessary to satisfy all the project requirements, avoid problems on-site and deliver the best 3D representation (Figure 1).

Figure 1

Figure 1: The different steps in reality-based 3D surveying and modeling projects, in particular in case of large and complex sites and architectures: specifications, planning, acquisitions, processing and final representation.

The reality-based digital acquisitions are done using active and passive sensors, specifically TOF and triangulation-based laser scanner, total stations and digital cameras, often mounted on Unmanned Aerial Vehicles (UAV) (Figure 2). For all the case studies, the surveying operations had to face the typical on-site problems, like presence of occlusions, difficult light conditions, reflections, just to cite some of them.

figure2kopie

Figure 2: Different moments of data acquisitions on site, using different sensors and techniques.

The data processing follows the well-known pipeline for the geometric and appearance modeling of the surveyed objects or sites. Generally unstructured point clouds are converted into structured polygonal models to digitally recreate the geometric shapes of the surveyed objects and allow better photo-realistic visualizations. For certain case studies, a simple point cloud will be available, as a better way to interact and display the surveyed scene.

figure3kopie

Figure 3: Examples 3D textured models generated after the post-processing of laser scanning data.

F1gure4kopie

Figure 4: Examples of photogrammetric processing and products.

The produced 3D models (Figure 3 and 4) will be available on the web (Figure 5) using different technologies and viewers (PDF3D, Nubes, Nexus, Unity3D, Potree, etc.) and in different forms (point clouds or textured polygonal models). All these 3D contents will be coupled with the necessary metadata.(point clouds or textured polygonal models).

Figure5kopie

Figure 5: Web-based visualization of large and complex 3D models within NUBES, a platform developed by CNRS-MAP (http://www.map.archi.fr/nubes/NUBES_Information_System_at_Architectural_Scale/Home.html). The large models (more than 10 million polygons with more than 200 MB texture, are easily displayed on the web and queryable.

The 3D-ICONS project is giving FBK the possibility to face different open issues in the3D  R&D community, in particular:

- the creation, as automated as possible, of a large amount of 3D models. The latest developments at sensors (e.g. hand-held laser scanner, Kinect-like sensors, etc.) and algorithm (e.g. Structure from motion or Dense Image Matching) level are so far promising but a dummy use of such new technologies without knowing in details performances, limitations and best practices will lead to unfavorable results. “One-click” software or “all-in-one” sensors are not the right approach, even for the creation of simple Internet products.

- sharing and preservation of 3D heritage contents. Internet is offering great possibilities but IPR issues, data transfer limitations, metadata creation and conservation of (3D) digital material are all important matters for the content providers and as well user community.

Related publications:

- F. Remondino, F. Menna, A. Koutsoudis, C. Chamzas, S. El-Hakim:  “Design and implement a reality-based  3D digitisation and modelling project”. Digital Heritage 2013 Int. Congress, IEEE Proceedings.

- B. Jiménez Fernández-Palacios, F. Remondino, J. Lombardo, C. Stefani, L. De Luca: “Web visualization of complex reality-based 3D models with Nubes”. Digital Heritage 2013 Int. Congress, IEEE Proceedings.