Catálogo de publicaciones - libros

In this paper, we propose a visual assistance system that shows the scene behind a vehicle to its driver in the form of a virtual mirror by utilizing a surveillance camera. The virtual mirror will be displayed on a windshield display as if it were floating in the air in front of the vehicle. The driver entering an intersection can confirm the safety around the vehicle by checking the image in the virtual mirror while facing forward, because it shows the image of the scene behind the vehicle in an optically correct manner. We utilize a surveillance camera set up on a signal pole. The image in the virtual mirror was formed by applying geometric its transformation to the image taken by the camera so that it satisfy the optical constraints of a mirror.

We propose a scalable framework for virtual reality systems in a distributed environment. As the application scope of and member participation in a virtual environment increase, information sharing among geographically distributed users becomes critical and challenging. In the proposed framework, we partition the virtual environment into a group of cells and upload them to a number of heterogeneous Internet nodes. When a user sends a request to explore the distant virtual environment, visible cells will be identified and processed in parallel to produce a minimal amount of imagery results for remote transmission. To ensure scalability, we extend our scalable occlusion culling scheme using Plenoptic Opacity Function to speed up the identification process of visible cells in a virtual environment. We perform effective occlusion culling in two passes based on a non-binary opacity definition. Our experimental results justify both the efficiency and scalability of our framework in exploring large-scale virtual environments.

A collaborative molecular modeling environment based on virtual reality techniques is proposed in this paper. The environment consists of a VR based molecular modeling system (VRMMS), a distributed processing system (DPS) and a web service server (WSS). The proposed environment was evaluated in terms of accuracy, performance, and collaboration. The accuracy of the simulation was examined by comparing the simulation results with those produced by the most popular simulation tool, Insight II. The distributed processing system of 4 computers showed good computing performance. The collaborative works of molecular modeling were successfully exercised and the total processing time of the collaboration was 3 times faster than that of a single user’s performance.

In large-scale DVE (Distributed Virtual Environment), data filtering is a critical factor to consider in connection with the real-time communications between simulators geographically separated. However, several existing data filtering mechanisms such as region-based, grid-based and hybrid ones result in limitations in computational complexity, filtering accuracy, or scalability. Aiming at providing an extensible filtering approach and reducing the overhead on both host and network, we propose an interest filtering mechanism based on LoI (Layer of Interest). The mechanism performs data filtering by means of multicast region publication, distributed receiver-side matching, and LoI-Based data filtering. We have implemented the mechanism into our RTI (Run-Time Infrastructure) software, and experiments are taken to demonstrate the effectiveness of it.

The use of Collaborative Virtual Geographic Environments (CVGE) is one of the most promising uses of virtual reality in geographic field. Data collaboration can support multi-directional sharing of data and files, allowing ideas and thoughts to be communicated interactively among users. Thus how to integrate diverse and separated data and offer more convenient and intuitive data collaboration service becomes increasingly significant for implementing collaborative work in the CVGE. In this paper, we firstly design a Grid-based CVGE service framework and discuss the data searching mechanism and accessing control. A data collaboration service was built to integrate and share system internal resources and improve collaboration work efficiency. Finally, we built a collaborative virtual environment and implemented data collaboration service in silt dams system planning on a case study area, Jiu-Yuan-Gou watershed of Loess Plateau, China. Experiment results prove that the scheme addressed in the paper is efficient and feasible.

The traditional video teleconferencing systems provide a “video in a window” interface paradigm, it’s not sufficient for naturally interactive interface for the collaborative work. We design a collaborative space called VCS. In VCS, a virtual collaborative space is built based on immersive teleconferencing. The remote conferees can discuss in the virtual space as similar to do it in the local rooms. VCS can be used to provide a new paradigm for the remote cooperative work. This paper presents the spatial model and video object extraction technique of VCS.

HLA-based Distributed Simulation technology is employed widely in Virtual Reality (VR) applications, such as military simulation, internet games, roaming etc. To support larger number of participants, scalability is becoming a key issue of VR applications. In this paper, we explore the characteristics of distributed simulation, and analyze the scalability of servers and participants, and classify our approach to improve scalability of VR applications into three aspects: a three-tier node management mode to accommodate more participants, an efficient management of servers to manage scalable number of clients and software development interface to achieve reusability and interoperability of VR applications. We present our middleware platform, HIVE, providing a scalable HLA-based distributed simulation framework for VR applications, on which users can develop VR applications easily and quickly. Then we give the method and view of application integration with HIVE. Finally an experimental demo is given.

Usual multimedia contents are represented as scenarios with markup language such as SMIL, and then delivered to be played back in terminal machines, e.g., PC or STB(Set-Top Box). However, these multimedia presentations are likely limited within single machine presentation where only the speakers and display devices attached to same computer system are used, so that the spatial reality is necessarily degraded consequently. In order to enhance even audio scenario presentation with such spatial reality, each audio clip for each audio actor, dubbing artist or background music should be assigned a separate speaker mostly close to the location where the author intended. In this paper, we propose a new architecture called L-DEAR() to support the spatial reality using location information of each node in a network environment. The core of the L-DEAR is IBEE() which is a set of nodes called out for playing back at least one assigned clip at a given time during the whole presentation. We will present the functional modules of L-DEAR and procedures for building IBEE, as well as functional requirements for L-DEAR such as extending the SMIL and using a global clock.

Providing haptic cues can generate increased levels of presence in users as they interact with tangible objects. In this paper, we present button effects delivered by air-jet displays with the aim of providing click-like sensations when virtual buttons in an AR environment are pressed. We derive the profile of the haptic output by measuring the force profile of real physical buttons. To validate our concept, we have built an AR system featuring a cellular phone model which users can tangibly manipulate through a physical AR marker object. When users press a button on the model, they experience a corresponding click-like feeling at their fingertip. In this way, our system enables users to press a button on an AR model and experience a force profile generated by our pneumatic array directly derived from the act of pushing a button in the real world.

This paper introduces a novel haptic rendering technique devised to perceptualize a “haptic edge” correctly with respect to its stiffness and height models. Our previous study showed that the traditional penalty-based haptic rendering methods are not adequate to the collocated data of surface topography and stiffness since surface topography perceived by the user can be distorted from its model. In order to overcome the problem, we have developed a based on the theory of which states that the user maintains a constant contact force when s/he strokes a surface to feel its topography. To the best of our knowledge, our technique is the first of its kind that explicitly considers the effect of user exploratory patterns in haptic rendering. Computationally, the algorithm is adaptive and efficient, not requiring any preprocessing of original data. We also demonstrate the performance and robustness of the proposed algorithm through a psychophysical experiment.