Web Visualization of the Function-defined Virtual Worlds
چکیده
This paper describes how the web visualization can be greatly improved using the function-based shape modeling technique. The improvement is possible because the proposed function-defined VRML shape node allows the content creators to describe any complex models with relatively small functions compared to the large-size polygonal mesh based VRML nodes. These functiondefined shapes can be used together with the common VRML shapes. The proposed node has a few implementations capable of visualizing geometric shapes defined with HyperFun language as well as in any proprietary function-defined data format. For fast visualization of the function-defined shapes, we have developed an improved continuation polygonization algorithm specifically designed for VRML visualization. The design, the implementation details, and the application examples of the proposed node are discussed. 1. Web visualization Web visualization services first appeared when some platform-independent programming languages and data exchange formats like HTML, VRML and Java were freely distributed. A variety of web visualization systems have been developed. Progressive reconstruction and isosurface transmission suggested by Engel et al [1] aimed to reduce the amount of data to be reconstructed and transmitted during visualization. In order to improve the progressive reconstruction approach, Engel [2] has demonstrated the construction of stripped surface representations and adaptive hierarchical concepts used to minimize the number of vertices that have to be reconstructed, transmitted, and rendered. Seidel et al [3] have presented a framework to acquire high quality 3D models of real world objects including both geometry and appearance information presented by textures, bump maps or bidirectional reflectance distribution function used to describe the way a surface reflects light. Fogel et al [4] have proposed a web architecture for progressive delivery of 3D content, which is based on a progressive compression representation integrated into X3D framework. Taubin et al [5] have proposed another popular approach to delivering a 3D content over the Internet within a reasonable time, besides introducing a new adaptive refinement scheme [6] for storing and transmitting the manifold triangular meshes in progressive and highly compressed form. Recently, Pajarola and Rossignac [7] have proposed a compressed progressive meshes approach, which uses a new technique to refine the topology of the mesh in batches. Among these services, the polygon mesh representation is often used to represent geometric models. Thus the most popular 3D web content format Virtual Reality Modeling Language (VRML) [8] uses this representation for complex objects. Although there are many workarounds to improve the overall experience in 3D web visualization, the mainstream of the current web visualization is based on the scenario where the publisher creates a 3D model rather than an image and sends it to the viewer, which then renders and manipulates the model. As the rendering is done at the client’s side, the smoothness of presenting a complicated scene, that is formed by millions of polygons, is much dependent on the Internet bandwidth. In many cases, the quality of the scene has to be compromised so that the 3D modeling file size is kept within an optimized range. From the different web visualization models or techniques available today, some techniques are adopted to improve the performance of visualization; while some are used to transmit the visualization data interactively based on the level of details and the network load. However, the root of the problem is the actual representation of the visualization data to be transmitted over the network. The compactness of the visual representation directly affects the overall user experience in interactive web visualization. In this project, we design and implement a generic function-based shape node for VRML. In Section 2, the function-based approach to shape modeling in web visualization is addressed. In Section 3, the proposed VRML node for a generic function model is introduced. In Section 4, we describe an implementation and test results for the function-based shape node developed for the particular function model F-Rep with its dedicated description language HyperFun. In Section 5, we illustrate how a proprietary function model can be plugged-into our architecture. In Section 6, the performance analysis is given. In the last section, further issues are discussed and conclusions are made. 2. Function-based shape modeling over the Internet 2.1. Function-based shape modeling Function-based shape modeling is becoming increasingly popular in computer graphics. The idea of function-based approach to shape modeling is that complex geometric shapes can be produced from a "small formula" rather than thousands of polygons. Usually, parametric or implicit functions and their modifications are used to define the shapes. For rendering such function-defined shapes, either ray tracing or polygonization followed by fast polygon rendering is used. Alternatively, the function-defined shapes can be voxelized and rendered as a set of points. It must be admitted, though, that function defined models may sometimes suffer a serious problem which is large time needed to evaluate the defining functions. Nevertheless, many works have been done in the direction of accelerating the function evaluation when performing their rendering, and now many function-based models, which existed rather theoretically just a few years ago, revolutionize the ways of shape representation. In the next subsections, we select the direction of expanding function-based modeling to web visualization, overview the existing projects attempting such visualization, and propose our own solution to this problem. 2.2. VRML and its existing extensions VRML is the most common format adopted to define 3D objects in the virtual web spaces. We have chosen it to integrate with function-based modeling technique for web visualization. Besides its popularity in defining 3D web content, VRML suffers an obvious performance drop whenever it is used to describe complex scenes. This drawback directly encourages designing a new VRML shape node based on the function-based approach. A generic function-based geometric node can be defined in VRML since this language is highly extensible. There are various VRML extensions that have been implemented successfully. For example, Alexa et al [9] have suggested the Morph node that helps to interpolate among several geometries. Another example is customized GeoVRML [10] nodes that are defined to provide a suite of solutions for representing and visualizing geographic data using a standard VRML97 browser. Wyvil and Guy [11] have introduced a new VRML extension for skeletal implicit surfaces with a limited set of operations like blending, warping and Boolean operations. Besides that, Pittet et al [12] have proposed a new real-time Isosurfacing node. This node, which is based on the so-called Marching Cubes algorithm [13], allows real time rendering of an isosurface from 3D source data. Grahn et al [14] have introduced trimmed NURBS in VRML. This adopted approach allows visualization of complex CAD models in VRML. Meanwhile, Ginis and Nadeau [15] have presented new VRML extensions for scientific visualizations. The suggested nodes are not mere collections of visually complicated parts, but can offer significant help in simplifying a user’s job when visualizing vector fields. 2.3. Function-defined geometric node In this project, a generic function-based node for VRML is proposed as a plug-in to a VRML browser. The idea of introducing such a node is to replace large polygonal representations of complex shapes with small formulae, which will be transmitted through the network much faster and therefore will improve the efficiency of web visualization. When a client tries to view a complex function-defined geometric model via the Internet, the VRML browser plug-in sends a view request to the remote content server. A small VRML file with a function-defined model description is sent back to the client. The size of the VRML file is small since the complex geometric model is described using little functions. When the VRML data reaches the client machine, the VRML plug-in starts to parse the VRML immediately. The scene graph traversal module proceeds and renders the scenes during the traversal. Once the scene graph traversal module finds the customized function-defined node, the browser plug-in will download the model description according to the specified URL. Alternatively, the description can be embedded in the VRML code. The polygonization of the function model is completed at the client machine, so that no large amount of rendering data is transferred across the network. Immediately after the polygonization has completed, the complex geometric model will be presented to the client. With the proposed architecture design, a new functionbased geometric node, called FShape, has been proposed for VRML. The FShape node is designed in such a way that it offers a generic infrastructure to support various function-defined models to be plugged-in during runtime. This can be achieved if the respective parsing and polygonization modules are implemented according to the defined plug-in interface. With this node, the content publishers can present any proprietary function-defined model to the viewers since they can develop the respective parsing and polygonization modules. 3. FShape node implementation A VRML node extension can be achieved either via plug-in customization or through VRML EAI interface [16]. For plug-in customization, the overall performance of the extended node will be much better than the other one, since the module is implemented with a low level programming language. The performance improvement becomes obvious when involving complex geometric models, which may require complex computations and algorithm evaluations. Although the plug-in customization may closely tie the node extension with the supported VRML browser plug-ins, the overall performance may need to be considered first if the node extension involves heavy computation processes. Another possible approach to implement the extension could be via a VRML Script node and its EAI interface. The implementation using JavaScript or Java may allow extension across different browsers, however this approach may have a serious performance setback. The polygonization implemented using JavaScript language will only be interpreted when it is about to be executed. Therefore, functions represented using a script language will always need more time to get executed as compared with precompiled modules. As overall performance is very important in the context of web visualization, the proposed FShape node is VRML browser plug-in dependent, in which the VRML extension is implemented via customizing a VRML browser plug-in. The proposed function-based geometric node could be extended via different browser plug-ins, e.g. Blaxxun Contact 3D [17], OpenVRML [18], FreeWRL [19], and other customizable browsers. In order to prove the concept proposed in this paper, we have extended Blaxxun Contact 3D to support FShape node. With the extension, viewers are able to visualize function-based objects in VRML paradigm. The extension of other VRML browsers can be done in a similar way, and we are working on it. 3.1. FShape node definition The proposed FShape node definition is listed in Figure 1. sourceString contains the definition of a functiondefined object. This field has a higher priority than sourceURL, i.e. if the field is defined, it will be used as the object textual definition source, instead of using the URL specified in the sourceURL field. sourceURL defines the actual path of the functiondefined object source file. It can be set to either a URL or a physical file path, in which the function-based content could be retrieved accordingly. Figure 1. FShape node definition with default values objectName specifies the model name for a functiondefined object. The default value for this field is “my_model”. sourceType identifies the type of the function-defined model referred in the scene. This field is prepared for the future extension to different types of function languages. The value of this field is used to associate the respective parser and polygonizer components with the main VRML plug-in module. ccw specifies whether the points of a face are presented in a counter-clockwise (TRUE), clockwise or unknown order (FALSE). convex specifies if the faces, being defined in an internal list of coordinate indexes defining the faces to be drawn, are convex or not. creaseAngle specifies an angle threshold. If two adjacent faces make an angle bigger than the creaseAngle, then the observer can see clearly where the two faces meet, the edge linking the two faces is sharp. Otherwise the edge linking the two faces will be smooth. normalPerVertex specifies whether the normal of each vertex is computed or not. solid determines whether the browser should draw both sides of a face or just the front side. reduce determines whether the polygonization process will produce the least number of polygons. The default value for this field is set to FALSE. searchVertex turns on recursive searching for vertex positions. If this option is not invoked, then BiLinear Interpolation is used to determine a vertex position for non-snapped vertices. FShape { ExposedField SFString sourceString [] ExposedField SFString sourceURL []