The History Of The Progressive Mesh Representations

Published Date: 02 Nov 2017

The existence of computer aided design (CAD) has led to the discovery of many computing technologies that help collaboration in CAD environment among designers in different locations. Collaboration among designers is one of the most important technologies developed for CAD, which allows designers to collaborate from different locations in real time. Recently, collaborative design capabilities in computer aided design (CAD) systems, required effective communication to share and transmit CAD models among other designers in distributed design environment (Rong et al, 2005). However, current software solutions still cannot fully address these needs, even though several information technologies have been successfully commercialized since the late 1990s such as teleconferencing tools, collaborative whiteboard, Web-based visualization, product data management (PDM), and Internet-based collaborative CAD. Among all this technologies, Internet-based collaborative CAD has not received much attention from the research community [4]. Internet based collaboration can be organized as either a horizontal or a hierarchical systems. Horizontal collaborative CAD technologies are mostly focused on synchronous integration of heterogeneous design data, processes, and software systems over the Internet. They either apply sophisticated interfacing technologies to connect multiple CAD systems [4], or develop proprietary framework for networked CAD [5]. On the other hand, most work concerning CAD integration in the hierarchical category has concentrated on improving the interoperability of design data among various activities in a product life cycle [4]. However, most of the technology developed for horizontal and hierarchical systems did not fully addressed the issues of CAD data transfer, especially large files transfer across the networks with limited bandwidth, which is a problem yet to be solved in the practice [4]. The problems encountered among designers is that CAD models are always too lag and are not easily transfer in collaborative environments. According to (Rong et al, 2005), the following problems were encountered with CAD transfer over the network: The size of data file the CAD models to be transmitted is generally quite larger than that of the other data file over the internet; The CAD model data are progressively modified and updated until the final design is finally fixed and due to the large size of model data, there is a significant delay during the transmission over the network.

In order to successfully transfer the CAD file over the internet, several important methods was developed, some of which are as follows:

The CAD model can be transfer using 3D streaming technology, which streams the data files into several segments.

Minimize the data to be transfer using WINRAR, which convert them into more compact file.

This academic paper is specifically to present a discussion on the topic of emerging computer technologies for CAD. Our main goal here is to elaborate more on 3D streaming technology for CAD based on Multi-LOD (level of details). The second section will discuss the literature review of the recent computing technology for CAD and the approaches/method used in dealing with it.

The third section will particularly talk about 3D streaming technology and how the technology is applied to CAD. Section four will present the discussion, which will lead to the conclusion.

2.0 literature review

2.1 Overview

Horizontal systems

Han et al. built a pilot realtime 3D system to promote the Internet-based collaborative engineering design (modelling) [9]. Product models are stored in database according to the STEP standard. The system provides a web-based search tool with the concept of metadata for navigating the product data. Sun and Gramoll [10] proposed an engineering analysis environment for remote users to analyze, review, and design a product collaboratively. They adopted a thin-server and thick-client architecture, in which the client has full analysis and computing capabilities and the server simply controls the communication in the collaboration process. Sharing of design information among collaborators is accomplished by transmission of commands rather than computed geometric data, thus minimizing the network traffic.

Hierarchical systems

Tang [14] enhanced the linkage between stamping product designer and the die maker in a distributed collaborative environment constructed with multi-agent technologies. The goal is to shorten the lengthy development time in the metal stamping industry. However, fewer studies have addressed the issues of CAD data transfer, especially large files transfer across the networks with limited bandwidth, which is a problem yet to be solved in the practice. The majority of the collaborative CAD systems adopt simple FTP (File Transport Protocol) functions with some secured features for data exchange (e.g. encryption, password protection, firewalls, . . ., etc.). Some vendors claim that their systems employs

3D streaming for distribution of CAD models, but they just borrow the streaming method originally developed in computer graphics, which gradually transmits triangular meshes from a simplified level to a complex level [15], or the so-called progressive meshing

[16]. Mesh-based 3D streaming requires less computing load on the client side, but its shortcomings include (1) the exact shape cannot be described and (2) the sequence of data streaming is not controlled by design content. It can only allow incremental display of 3D shape as a monolithic object, as the streaming method is based on two algorithms: mesh simplification and mesh refinement [17–19]. Li et al. proposed a 3D model simplification and refinement algorithm which can be used in 3D streaming [19]. It demonstrates a good performance in reducing the model size while maintaining good visualization quality.

RELATED WORK

2.1 Progressive Mesh Representations

PM [Hoppe 1996, 1998] is the first proposed progressive compression technique. After a sequence of edge collapse operations, the original mesh is simplified to the base mesh. Edge collapse removes one vertex at a time. The connectivity and geometric information of the removed vertex are stored in the refinement. The reverse process is called a vertex split. Here, a new vertex is inserted into the mesh. PM provides a fine-grain control of the accuracy and supports view-dependent refinement. However, PM is not efficient in the sense of compression ratio. PFS [Taubin et al. 1998] is much more efficient in encoding at the expense of looser granularity. It groups the refinements into a batch to achieve a high compression ratio. PFS cuts the mesh through.[3]