A Robust Method for Resolving Incorrect Visual Occlusion in Dynamic Augmented Reality Environments of Animated Engineering Operations

解决动画工程操作的动态增强现实环境中不正确视觉遮挡的稳健方法

• 批准号: 0825818
• 负责人: Vineet Kamat
• 金额: $ 29.87万
• 依托单位: Regents of the University of Michigan - Ann Arbor
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2008
• 资助国家: 美国
• 起止时间: 2008-09-01 至 2013-08-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=0825818&HistoricalAwards=false
• 关键词: Robust; Method; Resolving; Incorrect; Visual

The objective of this project is to investigate and design a functional and accurate occlusion handling method that can be easily integrated into any Augmented Reality (AR) application. AR is an emerging visualization technology which blends computer generated scenes in the form of CAD objects with the views of the real environment. From the point of view of an observer of an AR scene, incorrect visual occlusion occurs when part or whole of a real object must block the observer?s view of a virtual (i.e. CAD) object. If not resolved accurately, incorrect occlusion leads to unconvincing visual artifacts that reduce the observer?s confidence and reliability in the AR visual simulations, and consequently prevents them and other decision makers from comfortably using the simulation results for making crucial decisions. This research proposes a novel integration of geometry capture, remote sensing, and data communication technologies to enable realistic, reliable, and visually convincing results in AR animations of engineering operations. The research objective will be achieved by developing an automated method to obtain, record, and retrieve the depth information for both virtual and real entities in an AR animated scene of an engineering operation using remote sensing devices, and a z-buffering technique to detect and resolve any identified incorrect visual occlusion instances in real-time. Success in the research will lead to significant increases in productivity, accuracy, and safety in performance of operations due to better understanding of the engineering environment in which projects take place, through the use of advanced AR visualization technology. The results will be applicable to solve a wide range of engineering problems in construction, civil engineering, manufacturing, mechanical design, medicine, shipbuilding, transportation, and other domains. Solving the occlusion problem is also necessary to enable future work on subjects such as collision detection, motion preemption, and physical interference between real and virtual objects in AR, all of which have been long standing problems that prevent the widespread application of dynamic AR in several engineering and scientific domains. This project will significantly help the career development of graduate and undergraduate students, including women and minority engineers who will actively participate in the research activities and prepare to become tomorrow?s engineers. The project will lead to new software tools and training materials that will be widely distributed to the engineering, education, research, and professional communities to enhance scientific and technological understanding.

该项目的目标是研究和设计一种实用且准确的遮挡处理方法，该方法可以轻松集成到任何增强现实 (AR) 应用程序中。 AR 是一种新兴的可视化技术，它将计算机生成的 CAD 对象形式的场景与真实环境的视图融合在一起。从 AR 场景的观察者的角度来看，当真实对象的部分或全部必须遮挡虚拟（即 CAD）对象的观察者视图时，就会发生不正确的视觉遮挡。如果不能准确解决，不正确的遮挡会导致不令人信服的视觉伪影，从而降低观察者对 AR 视觉模拟的信心和可靠性，从而阻止他们和其他决策者轻松地使用模拟结果来做出关键决策。这项研究提出了几何捕获、遥感和数据通信技术的新颖集成，以在工程操作的 AR 动画中实现逼真、可靠和视觉上令人信服的结果。研究目标将通过开发一种自动化方法来实现，该方法使用遥感设备来获取、记录和检索工程操作的 AR 动画场景中虚拟和真实实体的深度信息，以及使用 z 缓冲技术来检测和检索深度信息。实时解决任何识别出的不正确的视觉遮挡实例。研究的成功将通过使用先进的 AR 可视化技术更好地了解项目发生的工程环境，从而显着提高运营绩效、准确性和安全性。研究结果将适用于解决建筑、土木工程、制造、机械设计、医学、造船、交通运输等领域的广泛工程问题。解决遮挡问题对于未来在 AR 中的碰撞检测、运动抢占以及真实与虚拟对象之间的物理干扰等主题的研究也是必要的，所有这些都是长期存在的问题，阻碍了动态 AR 在多个领域的广泛应用。工程和科学领域。该项目将极大地帮助研究生和本科生的职业发展，其中包括女性和少数族裔工程师，他们将积极参与研究活动并为成为明天的工程师做好准备。该项目将带来新的软件工具和培训材料，这些工具和培训材料将广泛分发给工程、教育、研究和专业社区，以增强科学和技术的理解。