IMAGO Visualization System: An Interactive Web ... - Semantic Scholar

1 downloads 0 Views 742KB Size Report
of an effective web-based 3D visualization system whose architecture offers an ..... for 3D models with higher resolution (e.g. Prophet). The. IMAGO 3D Viewer ...



IMAGO Visualization System: An Interactive Web-Based 3D Visualization System for Cultural Heritage Applications Caroline M. Mendes, Luciano Silva, Olga R. P. Bellon IMAGO Research Group, Universidade Federal do Parana (UFPR), Curitiba, Brazil

Abstract— Due to the evolution of technologies and methods for realistic 3D reconstruction of objects, in many projects it can be found efficient ways to make research results in digital preservation available on the Internet. 3D visualization of cultural heritage is highlighted in this scenario, helping to expand research activities in this field by providing proper tools to allow for example, remote access to historical artifacts. Thus, visualization systems must be able to handle important aspects in the context of digital preservation, such as user profiles, security and ease access to 3D models. This paper presents the development of an effective web-based 3D visualization system whose architecture offers an easy and fast interactivity with 3D models even when limited computer resources are available. The system has been successfully adopted in developing of 3D Virtual Museums in the Universidade Federal do Parana (UFPR) in Brazil, providing an important tool to promote research, educational, social and cultural activities.

I. I NTRODUCTION In the last decades there was a significant increase in the number of projects in digital preservation of art collections, introducing novel techniques for three-dimensional (3D) reconstruction of objects. Today, one can create a realistic 3D representation of an object to record its current state and make it available for future generations. 3D models can also assist restoration activities on cultural objects suffering natural degradation or even for physical reconstruction of them in case of disasters. Considering the fragile characteristic of many cultural objects, interactive 3D visualization is an effective tool to virtually explore these objects with great realism, thus avoiding a physical manipulation that may cause irreparable damages. Likewise, virtual museums contribute significantly to the remote access to collections, through web-based systems for 3D visualization, allowing visitors to explore and visualize artworks wherever they are, from large sculptures [1] to small artifacts manufactured by endangered communities [2]. Therefore, advanced systems to generate realistic 3D models have being developed [3], becoming fundamental tool to support activities in areas of education, entertainment and culture. Some issues need to be considered to make available cultural heritage 3D models on the Internet. The This work was supported in part by UNESCO, UFPR and CNPq.

© 2012 ACADEMY PUBLISHER doi:10.4304/jmm.7.2.205-210

copyrighted objects need to be preserved and the highresolution 3D model digitally preserved can not be accessed. The Digital Michelangelo Project [1] was pioneer in exemplify this issue, and the authors proposed a remote rendering technique to provide 3D interactive while protecting the high-resolution 3D model [4]. However, if high-resolution 3D models are available, techniques to store, transmit, and visualize large amounts of data need to be incorporated into systems. Progressive meshes [5] or progressive compression [6] allow the user interaction even while the 3D model has not been completely transmitted. In addition, the multi-resolution approaches like multi-triangulation [7] can improve performance to storing and managing the level-of-details. Digital preservation users, e.g. curators, professors and students, demand quick and easy access to 3D models. The difficulty in accessing and interacting with 3D models is certainly a demotivating factor for many users, especially in teaching and research activities. Unsuccessful software installations, complex user interfaces, slow response time and limitations in performing simple operations such as zooming or rotating an object, are found in the some web-based systems for 3D visualization. This paper presents a web-based system for 3D Virtual Museum that provides effective access to 3D models digitally preserved. The 3D Virtual Museum at the Universidade Federal do Parana in Brazil has successfully promoted the visualization of assets from several institutions, such as the Curitiba Metropolitan Art Museum and the UFPR Natural Science Museum. The web-based system was developed considering observations made by digital preservation users, such as computational limitations, cognitive ability and native language. The proposed 3D visualization tool was designed for intuitive 3D interaction. Furthermore, a remote rendering server is used to ensure security in the context where the high-resolution 3D models can not be available. This paper is organized as follows. Section II presents the use of 3D viewers in digital preservation projects. Section III introduces the web-based system for 3D visualization, as well as the 3D visualization tool developed for this context, followed by conclusions and future works in Section V.




Most digital preservation projects using VRML (Virtual Reality Modeling Language) viewers to provide 3D interaction in yours systems. VRML is a markup language for geometric modeling that allows to describe 3D objects and scenes into a text file [8]. Most recent, the X3D - Extensible 3D ( is the VRML improvement and incorporates advances, such as animations and multi-textures and Extensible Markup Language (XML) support. VRML and X3D files are usually visualized by standalone 3D viewer or plugin for web browsers. Octaga Player (, Cosmo Player ( and Cortona3D ( are some examples. VRML is widespread and shared format on the Internet. However, experiments performed in this work showed dissatisfaction of digital preservation users using VRML viewers. Some problems was observed: unsuccessful installations; some web browsers not support all VRML features; unexpected inoperability without output error message; intermittent delays in rendering; and complex user interface. VRML/X3D viewers provides a user interface, but it differ in the number of options available and the location of these options on screen. Figure II shows three different viewers tested. The Cortona3D displays a toolbar (Figure 1(a)), Cosmo Player presents a navigation panel (Figure 1(b)) and Octaga Player (Figure 1(c)), in some versions, does not present a visible interface, being necessary to click the left mouse button to display the options. Virtual Inspector [9] is a visualization software that allows users to inspect large 3D models. The tool is targeted to digital preservation users and provides easy and intuitive 3D interaction. The main purposes of this system are: provide performance using low-cost computers; simple and intuitive interaction for beginner users; high quality rendering of large 3D models; enable professionals of digital preservation to annotate information in the 3D models; allow local and remote access; and provide protection of 3D data. One of the main advantages of Virtual Inspector is its flexibility and configurability. The interface can be specified using XML, e.g. interface components that will be displayed on screen, including interactions parameters. The remote visualization can be done with a particular web browser, especially developed for this purpose. Therefore, users are not allowed to access 3D models using their preferred web browsers, such as Internet Explorer, Mozilla Firefox, among others. In the last decade, approaches to 3D visualization of cultural heritage were presented. The systems are included in practical context, where users can access 3D models on Internet. Inuit 3D, a project of the Canadian Museum of Civilization [10], was one of the first to allow 3D visualization in web browsers. The 3D models exposed in virtual exhibition rooms are visualized using VRML viewers. Czech Republic heritage (e.g. historical monuments, buildings and objects) is available on the © 2012 ACADEMY PUBLISHER

Internet ( using VRML models. The models are stored in a database and are dynamically sent to client after verification of the user position in the virtual city [11]. Other example, the VRML 3D model of ancient city of Xanthi in Greece is available on the Internet ( The high-resolution texture provides realism to 3D model and are progressively loaded according the viewpoint, thus allowing interaction even without the complete texture received [12]. In The Digital Michelangelo Project [1] is presented the 3D reconstruction of large Michelangelos artworks, in particular the sculpture David. Due to the cultural importance, the authors discuss the unauthorized distribution and the physical reconstruction of objects from images of the 3D models. They created a client-server visualization system [4] in which 3D models in both resolution (high and low) are stored in a remote rendering server. The system enables users to interact only with low-resolution 3D models, and client receives images from the same 3D model (high-resolution) rendered on the server every time that user stops the interaction. The standalone 3D viewer developed (client) provide a intuitive user interface and works in Windows. As reported in this section, digital preservation projects mostly use VRML viewers or standalone software for 3D visualization. The digital preservation users studied in this project reported some regards to developing of our web-based system, such as intuitive 3D interaction, comprehensible interface and easy access on web browser. The system developed is presented following. III. IMAGO V ISUALIZATION S YSTEM The IMAGO Research Group ( has a Virtual Museum to disseminate the results obtained in digital preservation of cultural and natural assets. The web-based 3D visualization system was developed to provide quick and easy access to 3D models for digital preservation users. Contact with users involved in digital preservation areas allowed the creation of a 3D visualization tool for intuitive 3D interaction. Besides, users can give feedback to the system, allowing further improvements. The system has three approaches to access 3D models and use remote rendering in order to allow the visualization even in computers with insufficient configuration (e.g. accelerated graphic card) through web browsers. In order to protect the 3D models, it is necessary to register to login in the system, and the visualization of the 3D models is limited according to the user access level. Visitors have limited access, but teachers or researchers may request less restricted access in order to visualize 3D models with higher resolution. The system uses a client-server architecture (organized as shown in the diagram of Figure III) that allows the viewing client on web browser, e.g. IMAGO 3D Viewer (Section III-C), while the server contains 3D models of high and low resolution. The client 1 can interact with 3D models locally. The client 2 can interact with 3D






Figure 1. Examples of user interfaces of VRML viewers: (a) Cortona3D, (b) Cosmo Player and (c) Octaga Player.

models locally and can also view images obtained from the remote rendering. Finally, the client 3 only view images sent by the remote rendering server, so considering that it has no 3D accelerated graphics card. More details on this system are presented below. A. 3D Models Used The 3D models used in the system are generated in three steps: data acquisition, geometry reconstruction, and generation of high-resolution textures. In the acquisition stage two main devices are used: (1) a laser scanner (e.g. Minolta Vivid 910) obtains 2D color images with lowresolution (640x480) and depth images of the object, and (2) a digital camera (e.g. Canon EOS5D) captures highresolution images to generate textures. In the second step, the acquired views (obtained from different positions and stored as depth images) are aligned and integrated to obtain a 3D mesh. The complete reconstruction developed is composed of several specific steps, described in [3], [13]. Finally, the third phase generates and applies a high quality texture in the 3D model, as proposed in [2], [14]. B. IMAGO 3D Viewer One of the goals of the system was to present a 3D visualization tool alternative to VRML viewers, which was considered by users studied in this project as having a complex interface and non-intuitive interaction. The first initiative of this project was the IMAGO Plugin [15], but like any plugin, the additional installation is required. Observed Java acceptance, since Java Runtime Environment (JRE) is usually installed by default in user’s computer, was developed the IMAGO 3D Viewer, which uses Java applets and Java OpenGL (JOGL). IMAGO 3D Viewer can be used by other projects and are available at the project website. 1) Use Mode: To use the IMAGO 3D Viewer on web browsers we can use the Applet tag and the JNLPAppletLauncher resource in HTML web page. The viewer receives as input a zip file that must be specified in the input parameters. Other parameters must also be specified, such as the 3D model file name, among others. The zip file contains the 3D model file, the texture file and a configuration file. We use a simple 3D model format that can © 2012 ACADEMY PUBLISHER

be generated by any application. 3D models with texture (extension “.m”) and vertex color 3D models (extension “.m2”) are text file with following informations: the 3D coordinates (X, Y, Z) of each vertex; the normal vectors of each vertex; the set of faces, represented by 3 vertices each; color information or uv coordinates of each vertex (i.e. mapping the 2D texture coordinates for each vertex of the 3D model) for 3D models with texture; and color properties of the scanned object. The 3D model must be attached in the zip file, as well as the texture file (i.e. Targa File Format). More information can be found on the project website. 2) Interface Configuration: IMAGO 3D Viewer interface can be customized according to the target group of the application by editing a configuration file. The programmer can choose the operations to manipulate the 3D model (predefined events), interface components (e.g menu bar layout and/or toolbar icons) and activation mode of operations(e.g. via an icon on the toolbar and/or mouse, keyboard or other compatible input devices). Menu bar and toolbar are interface components widely used and are more understandable for beginners and unfamiliar users with virtual reality environments. Also, the default image of toolbar icons can be replaced for drawings or symbols that the programmer considers more intuitive for his/her application users. As shown in Figure III-B.2, the configuration file has two sections: Menus and Events. The menus are defined in the configuration file after the tag “# menu” and each line contains the label of the main menu followed by the labels of the options. The menus can contain western text, allowing the use and internationalization for different user groups. The events are defined after the tag “# Events” and each line contains a label followed by a predefined event (all predefined events are shown in Figure III-B.2) and the activation mode. The choice of devices and independence of language allow greater flexibility, being adaptive to the needs of the application or user. 3) Running: To provide a better rendering performance, the viewer uses OpenGL VBO. When it is not possible, triangles strips of the 3D model are generated in real time [16]. The use of VBO for static geometries (no animation) offers a considerable gain in performance, since it is possible to send data once to the graphics card



Figure 2. Architecture used in the visualization system of 3D Virtual Museum.

Figure 3. Example of the configuration file.

memory, avoiding bus over-transfer. With Triangle Strips, each new triangle can be set with the only one extra vertex by sharing two vertices already defined in the previous triangle. Thus, the number of vertices sent to and stored in the graphics card may be reduced by almost one third.

C. Client To allow the access with minimum requirements, three visualization approaches were developed: 1) Interacting with 3D Models: This approach considers that users have the resources to render the 3D models, e.g. 3D accelerated graphics card. The low-resolution 3D model is sent only once, but with sufficient visual quality for a good investigation of the object. IMAGO 3D Viewer is used to provide interaction in a web browser. 2) Interacting with 3D Models and Viewing Images: This approach is used to serve users who can render 3D models but wish to observe more details of objects. It is similar to that seen in [4], allowing the user to interact with 3D models and view images remotely rendered, but the viewing is done in the web browser. The user interacts with the low-resolution 3D model and when stop moving it, a request containing the viewing parameters is sent to the server, following the image shown on the screen (Figure III-C.2). This approach can consume bandwidth, since multiple images are progressively required. © 2012 ACADEMY PUBLISHER

3) Viewing Images: This approach allows to access 3D Virtual Museum just in case users can not render 3D model and not have Java support. In this case, the user can view images of the high-resolution 3D models rendered remotely. The user chooses a 3D model and a web page is displayed containing a similar IMAGO 3D Viewer interface with the same operations. By clicking a button, a request is sent to the server which sends an image that is displayed on the screen. The Java Server Pages (JSP) and Asynchronous JavaScript and XML (Ajax) were used in the development. This approach also allows the visualization of 3D models on mobile devices. D. Server The server is responsible for storing and sending 3D models and images to the client. It consists of a request server, a rendering server and a database, which stores information about users profile and 3D models available. The access to the system is stored in log files, in order to identify which 3D model is being viewed and by which users. A request server working on ports 80 (HTTP) and 43 (HTTPS) was used for intermediate the communication between clients and the rendering server. Thus, access problems are avoided since these ports are commonly released in networks. The rendering server generates JPEG rendered images from the highest resolution 3D model using the viewing parameters of the client (e.g. 3D coordinates, lighting, zoom factor, among others) and sends these images to the request server. The remote server was developed with C++ and OpenGL. To improve rendering performance, we use the FBO programming technique to support offscreen rendering. Using the FBO, image data are sent directly to a texture, and the rendering is performed with less memory usage and more processing gain. IV. P ERFORMANCE The experiments reported in this section were performed in computer with the following configuration: Intel Core 2 Duo processor 6300 1.86Ghz, 1 GB RAM,





Figure 4. Visualization of a Protocyon fossil using IMAGO 3D Viewer. The Protocyon is an animal that inhabited caves in Brazil during the Pleistocene (period between 1.816.000 and 11.500 years ago). Fossil belongs to the UFPR Museum of Natural Sciences. (a) Low-resolution 3D model rendered locally on the client’s computer and (b) image of the high-resolution 3D model rendered on the server and transmitted to the client through the Internet.


Client 51292 37456 48084 79664

Number of Faces Server 873746 342520 666900 714696

NVIDIA GeForce 7300SE, Windows XP Service Pack 2 operating system. The compression ratio using zip format for 3D models tested was of around 70% and of configuration file (802 bytes) of around 64%. The input zip files were generated in a Windows using the Winrar software with standard compression method. Table I shows the comparison between the rendering performed by IMAGO 3D Viewer using VBO and Triangle Strip with three widely used VRML viewers: Cortona3D, Cosmo Player and Octaga Player. The results show that IMAGE 3D Viewer render the 3D models with higher fps compared to the VRML viewers tested, even for 3D models with higher resolution (e.g. Prophet). The IMAGO 3D Viewer has been tested on different operating systems (Linux, Windows and Mac), and web browsers, such as Firefox, Google Chrome and Internet Explorer. In remote rendering tests we used four 3D models available in our 3D Virtual Museum: Protocyon, Alamito, Stenzel and Carybe. 3D models informations are in Table II. Table III shows the tests performed on the client (IMAGO 3D Viewer). The 3D models are with three times of zoom in from the start position (Figure IV). This table presents the time spent on the rendering server to generate images, as well as the time spent between request and receiving the rendered images and display then on the screen. The little time spent on the server is due to offscreen rendering. On the client, the most time was spent to display the image on the screen, caused by the Java image library. © 2012 ACADEMY PUBLISHER





Figure 5. Starting position of 3D models used in the remote rendering test: (a) Protocyon, (b) Alamito (c) Stenzel and (d) Carybe.

V. C ONCLUSION This work presented the web-based 3D visualization system developed to provide quick and easy access to 3D models for digital preservation users. Currently, some digital preservation projects have dedicated on digitalization process and techniques to generate realistic 3D models, and VRML viewers has been used as a practical approach. It presents the IMAGO 3D Viewer, an alternative visualization tool to VRML viewers, which was considered nonintuitive for beginners users to 3D interaction. IMAGO 3D Viewer interface can be adapted to different target audience, providing advanced operations for experienced users while keeping a simple interface for beginner users. Contexts where 3D models can not be transmitted, the remote rendering turn the applications more secure and reliable. Web-based applications such as virtual museums need, in addition to security, make their systems more accessible. The visualization approaches developed allow the users to visualize 3D content on web browser, regardless of computer configuration. The interaction with low-resolution 3D models or viewing remotely rendered images does not affect the perception of object details.




Number of Faces 342520 714696 593584 1409939 873746 666900 792809 1210630

Alamito Carybe Rooster Prophet Protocyon Stenzel Vase Vitor



IMAGO 3D Viewer (VBO) 75.0 fps 53.0 fps 62.0 fps 29.0 fps 44.0 fps 56.0 fps 48.0 fps 33.0 fps

IMAGO 3D Viewer (T. Strips) 43.0 fps 22.0 fps 25.6 fps 11.4 fps 18.0 fps 24.2 fps 20.0 fps 13.3 fps

Cortona3D 6.0 9.3 fps 5.0 fps 5.87 fps 2.69 fps 4.07 fps 5.27 fps 4.53 fps 3.25 fps

Octaga Player 12.06 fps 5.97 fps 7.31 fps 3.12 fps 5.0 fps 6.49 fps 5.48 fps 3.66 fps

Cosmo Player 2.1.1 16.0 fps 9.0 fps 10.7 fps 4.6 fps 7.4 fps 9.6 fps 8.1 fps 5.3 fps

TABLE III. R EMOTE RENDERING TESTS . 3D Model Protocyon Alamito Stenzel Carybe

Image Size 124214 bytes 107627 bytes 67319 bytes 82241 bytes

Server Total Time 125 111 120 112

ms ms ms ms

Our project is now working in the preservation of baroque sculptures made by Antonio Francisco Lisboa, known as “O Aleijadinho” [17], in collaboration with UNESCO. The 12 near real sized sculptures represent twelve biblical prophets and are located in the Bom Jesus de Matosinhos Sanctuary in Congonhas, Brazil. Thus, we aim to develop methods to visualization of complex 3D models for our web-based system. Also, the WebGL is a recent technology to provide immediate access by 3D graphics support for web browsers natively. R EFERENCES [1] M. Levoy, K. Pulli, B. Curless, S. Rusinkiewicz, D. Koller, L. Pereira, M. Ginzton, S. Anderson, J. Davis, J. Ginsberg, J. Shade, and D. Fulk, “The digital michelangelo project: 3D scanning of large statues,” in SIGGRAPH 2000, Proceedings of the 27th Annual Conference on Computer Graphics, New Orleans, USA, 2000, pp. 131–144. [2] B. T. Andrade, O. R. P. Bellon, L. Silva, and A. Vrubel, “Digital preservation of brazilian indigenous artworks: Generating high quality textures for 3D models,” Journal of Cultural Heritage, vol. 13, no. 1, pp. 28–39, 2011. [3] A. Vrubel, O. R. Bellon, and L. Silva, “A 3D reconstruction pipeline for digital preservation,” in Proceedings of IEEE Conference on Computer Vision and Pattern Recognition, Miami, USA, 2009, pp. 2687–2694. [4] D. Koller, M. Turitzin, M. Levoy, M. Tarini, G. Croccia, P. Cignoni, and R. Scopigno, “Protected interactive 3D graphics via remote rendering,” ACM Transactions on Graphics, vol. 23, no. 3, pp. 695–703, 2004. [5] H. Hoppe, “Progressive meshes,” in Proceedings of the 23rd Annual Conference on Computer Graphics and Interactive Techniques, ser. SIGGRAPH ’96, New York, USA, 1996, pp. 99–108. [6] P. Alliez and M. Desbrun, “Progressive compression for lossless transmission of triangle meshes,” in Proceedings of the 28th Annual Conference on Computer Graphics and Interactive Techniques, ser. SIGGRAPH ’01, New York, USA, 2001, pp. 195–202. [7] P. Cignoni, F. Ganovelli, E. Gobbetti, F. Marton, F. Ponchio, and R. Scopigno, “Batched multi triangulation,” in Proceedings IEEE Visualization, Minneapolis, USA, 2005, pp. 207–214. © 2012 ACADEMY PUBLISHER

Request Time 214 ms 159 ms 170 ms 121 ms


Client Response Image Drawing Time 646 237 231 247

ms ms ms ms

[8] J. Zara, “Virtual reality and cultural heritage on the web,” in Proceedings of the 7th International Conference on Computer Graphics and Artificial Intelligence, Limoges, France, 2004, pp. 101–112. [9] M. Callieri, F. Ponchio, P. Cignoni, and R. Scopigno, “Virtual inspector: A flexible visualizer for dense 3D scanned models,” IEEE Computer Graphics and Applications, vol. 28, no. 1, pp. 44–54, 2008. [10] F. Corcoran, J. Demaine, L.-G. Dicaire, M. Picard, and J. Taylor, “Inuit 3D: An interactive virtual 3D web exhibition,” in Proceedings of the Conference on Museums and the Web, Boston, USA, 2002, pp. 163–169. [11] Z. Jiri and S. Pavel, “Cultural heritage presentation in virtual environment: Czech experience,” in Proceedings of the 14th International Workshop on Database and Expert Systems Applications, Prague, Czech Republic, 2003, pp. 92–96. [12] A. Koutsoudis, F. Arnaoutoglou, and C. Chamzas, “On 3D reconstruction of the old city of xanthi. A minimum budget approach to virtual touring based on photogrammetry,” Journal of Cultural Heritage, vol. 8, no. 1, pp. 26–31, 2007. [13] J. S. Junior, O. Bellon, L. Silva, and A. Vrubel, “Improving 3D reconstruction for digital art preservation,” in ICIAP 2011, Proceedings of the 16th International Conference on Image Analysis and Processing, Ravenna, Italy, 2011, pp. 374–383. [14] B. T. Andrade, O. R. P. Bellon, L. Silva, and A. Vrubel, “Enhancing color texture quality of 3D models for digital preservation of indigenous ceramic artworks,” in Proceedings of the 12th IEEE International Conference on Computer Vision, Workshop on eHeritage and Digital Art Preservation, Kyoto, Japan, 2009, pp. 980–987. [15] C. M. Mendes, D. R. Drees, L. Silva, and O. R. Bellon, “Interactive 3D visualization of natural and cultural assets,” in Proceedings of the 2nd Workshop on eHeritage and Digital Art Preservation, Firenze, Italy, 2010, pp. 49–54. [16] F. Evans, S. Skiena, and A. Varshney, “Optimizing triangle strips for fast rendering,” in Proceedings of the 7th Conference on Visualization, San Francisco, USA, 1996, pp. 319–326. [17] B. T. Andrade, C. M. Mendes, J. de Oliveira Santos Jr., O. R. P. Bellon, and L. Silva, “3D preserving xviii century barroque masterpiece: Challenges and results on the digital preservation of aleijadinho’s sculpture of the prophet joel,” Journal of Cultural Heritage, 2011, in press.

Suggest Documents