Abstract In this topic, the main issues and challenges related to creation, management, search and presentation of interactive 3D content are presented. The challenges result mainly from the continuous progress in development of 3D/VR technologies and systems. In particular, the following shifts are presented: from passive 3D content to active 3D applications, from 3D geometry to meaningful interactive virtual environments, from single-user to multi-user 3D content creation, from file search to 3D object search, and from 3D scenes to efficient 3D visualization interfaces.
Introduction
Standards enabling platform-independent declarative description of interactive multimedia 3D content, such as VRML/X3D and MPEG-4 described in Chap. 2, have been developed for a long time and have reached a high level of technological advancement and maturity. The 3D content standards make possible creation of highly complex content, in which both synthetic and natural elements are integrated in 3D virtual scenes. These standards enable rich forms of multimodal data presentation and provide complex interaction capabilities. Virtual scenes conforming to these standards can be accessed both locally and remotely over the network, similarly as in the case of typical Web content. Different hardware and software platforms can be used to access virtual scenes, starting from lightweight mobile terminals, such as smartphones and tablet PCs, through typical desktop computers to highly sophisticated immersive virtual reality (VR) installations.
The 3D content standards have a great potential to become important elements leveraging development of a new class of interactive secure 3D multimedia applications, available on-line in a distributed network environment and combining features of today’s Internet, television and games to provide users with rich experience in different domains, such as entertainment, education, training, tourism, e-commerce, engineering, design, and medicine.
Such 3D multimedia network applications can offer rich forms of 3D presentation and interaction, available today only in desktop applications, with flexibility and simplicity of networked access, typical to today’s Web. Access to such applications through the network—combined in some cases with cloud computing—offers great advantages over traditional desktop applications. Due to Web search mechanisms, such applications can be easily located and accessed. They can be used on any device, at any time and any place—including both on-request and broadcast publishing modes. The applications can be flexibly grouped, linked and merged to create more advanced 3D mashups. Most importantly, they can be based on usergenerated content. In many application domains user-generated content is critical. By “users,” we mean both end-users, accessing the final 3D content, and domain experts, who are not IT experts, but who are responsible for preparation of content for a particular application, for example, a virtual museum, a virtual store, or a VR-based learning system. Only involvement of non-expert users on a mass scale in content creation process guarantees that there will be sufficient amount of up-to-date relevant content and that the content will be of interest to users. Users should be able to easily create and share over the network reusable interactive 3D content elements—including geometry, associated multimedia objects, and behavior. Such content should be searchable on the network, including all of its important properties—semantics, geometry, interactivity, and behavior.
Similar evolution trends are well known from the development of the WorldWide Web, which started as a static content publication platform and evolved towards a universal application deployment platform and the largest shared repository of user-generated content (Web 2.0). Similar development path is expected for the 3D Web, however, the challenges which arise here are more complex because of the inherent complexity of the interactive 3D content. Therefore, before the vision of ubiquitous 3D network applications described above becomes a reality, there are still a number of issues that must be addressed. Most of them relate to the creation, management, search, and presentation of the interactive 3D multimedia content. The issues related to the following shifts are further discussed in this topic:
• from passive 3D content to active 3D applications,
• from 3D geometry to meaningful interactive virtual environments,
• from single-user to multi-user 3D content creation,
• from file search to 3D object search,
• from 3D scenes to efficient 3D visualization interfaces.
The above list is not a complete list of problems related to creation of interactive 3D applications. According to the authors of this topic, they are, however, the most important 3D content-related research challenges which are not sufficiently addressed by research community. Issues related to 3D scanning, modeling and development of content standards have been well covered by other research efforts.
The content representation standards, such as VRML/X3D and MPEG-4, described in Chap. 2, enable building passive 3D/VR systems, i.e., systems where the 3D/VR
technology is employed to visualize some pre-designed virtual environments in the form of three-dimensional scenes. These scenes may be animated and interactive but they are passive in the sense that their content structure is fixed. The set of objects contained in virtual scenes and their initial values of attributes are constant. These scenes can be—to some extent—dynamically modified by Script nodes, but again, the scripts and their initial values of parameters are fixed.
A real challenge in the area of 3D/VR systems based on content representation standards consists in building active applications, where 3D content is used not only to represent fixed 3D scenes but also serves as a rich interface for fully-featured interactive 3D applications or complex interactive virtual worlds. For this to happen, problems such as dynamic composition of virtual scenes, access to external data sources, efficient modeling and manipulation of complex virtual world data, persistency of virtual worlds, and continuous visualization must be addressed. To solve these problems, the following methods must be provided:
• methods of flexible selection of contents of virtual scenes based on various content properties, including geometrical, logical and temporal characteristics of content elements;
• methods of parameterization of virtual scenes, with preferences provided either by an end-user (implicitly or explicitly)—for interactive 3D Web applications built with the VRML/X3D standards, or by the content producer—for broadcast systems based on the MPEG-4 standard;
• methods of accessing databases and other data sources from 3D/VR applications, both during virtual scene creation and at run-time. These methods should enable retrieval of all elements of the virtual scenes from a database;
• methods of describing virtual scenes on a higher level of abstraction than geometrical, and creating libraries of reusable components on the base of object-oriented techniques;
• methods of modifying appearance of virtual scenes by selecting and adjusting data visualization method;
• methods of manual and automatic updating of virtual worlds, both from virtual scene instances and other applications;
• methods of enforcing user privileges for spatial or logical subsections of virtual environments; and
• methods of efficient database modeling of virtual objects, virtual scenes and virtual worlds whose extent—spatial, logical or temporal—may exceed a single virtual scene.
Numerous research and development works have been carried out with the aim to extend or supplement 3D content standards with the above features. In particular, the use of object-oriented techniques has been studied [10,24-26, 66, 67]. However, these works focus on behavioral aspects of content, leaving geometrical aspects as an open issue. A method of pre-processing VRML scenes has been proposed in [31]. The functionality of this solution is limited and the syntax is bound to the VRML97 textual notation. The problem of accessing databases from VRML scenes has also been investigated. The VRML Database Working Group has approved recommended practices for accessing databases [49]. However, the proposed solution provides only low level data access methods and is limited to controlling values of fields of existing nodes. As a consequence, both the virtual scene structure and the applied visualization method must be hard-coded in virtual scenes and only attributes of existing elements can be influenced by data coming from a database. For database modeling of virtual worlds, in most cases application specific solutions are used, although a simple general purpose spatiotemporal database model for VRML objects has been proposed in [82].
The works described above provide only partial solutions to selected problems related to building active 3D/VR applications and none of these methods have become widely accepted. A much more comprehensive and flexible approach is needed to enable building practical active 3D/VR applications.
The problem of building active 3D/VR applications is further discussed in Chap. 4, and an approach to this problem, called X-VR, based on dynamic database modeling of interactive 3D/VR content is proposed.
