I-Corps: Transforming 2D Video into an Interactive 3D Viewing Experience

I-Corps：将 2D 视频转变为交互式 3D 观看体验

• 批准号: 1243737
• 负责人: Zhigang Zhu
• 金额: $ 5万
• 依托单位: CUNY City College
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2012
• 资助国家: 美国
• 起止时间: 2012-07-01 至 2013-08-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1243737&HistoricalAwards=false
• 关键词: Corps; Transforming; 2D; Video; into

A novel method is developed to convert long video sequences from a single camera into 3D panoramic visualizations. The approach holds the promise to work at near real-time on a laptop. The 3D content allows users to change view angles and it also supports panoramic views. The system has two real-time software components: (1) the image matcher that generates multi-view panoramic mosaics from 2D image sequences; and (2) the 3D renderer that can provide an end user a virtual fly-through or walk-through of a scene using commercially available 3D displays. The approach promises real-time 3D visualization for a broad range of camera types (high end/low end color, IR, even gamma-ray/r-ray imaging), camera motions (linear, circular), motion platforms (airborne, ground vehicle, hand-held), and scene distances (from a few inches to thousand feet). This technology has matured from research into environmental monitoring, security and defense applications, where video was captured from a single camera on board an aircraft or ground vehicle. The system is able to produce 3D visualizations of the videos for further video data analysis, scene understanding and scientific discovery. The technology has applications not only in defense and environmental monitoring, but also in entertainment, medical imaging, inspection, real estate and land survey/development industries. There are many 3D ready devices, from TVs, to smartphones in the consumer market already. But the content is still missing. To address this unmet need, the team is building a smartphone application to enable user-generated 3D content, without requiring a 3D camera.

提出了一种将单摄像机拍摄的长视频序列转换为3D全景可视化的新方法。 这种方法有望在笔记本电脑上接近实时地工作。3D内容允许用户改变视角，它还支持全景视图。该系统具有两个实时软件组件：（1）图像匹配器，其从2D图像序列生成多视图全景马赛克;以及（2）3D渲染器，其可以使用商业上可获得的3D显示器向终端用户提供场景的虚拟漫游或漫游。该方法承诺实时3D可视化广泛的相机类型（高端/低端彩色，红外，甚至伽马射线/r射线成像），相机运动（线性，圆形），运动平台（机载，地面车辆，手持），和场景距离（从几英寸到几千英尺）。这项技术已经从环境监测、安全和国防应用的研究中成熟起来，在这些应用中，视频是从飞机或地面车辆上的单个摄像头捕获的。该系统能够生成视频的3D可视化，用于进一步的视频数据分析，场景理解和科学发现。该技术不仅应用于国防和环境监测，而且还应用于娱乐、医学成像、检查、真实的房地产和土地测量/开发行业。 消费市场上已经有许多3D设备，从电视到智能手机。但内容仍然缺失。为了满足这一未满足的需求，该团队正在构建一个智能手机应用程序，以支持用户生成的3D内容，而无需3D相机。