Cyber System:Collaborative Research: An Intelligent Network of Wireless Videos for Dynamic Scene Analysis

本提案的重点是开发一个通过无线通信信道连接的视频摄像机智能网络，用于动态场景理解。在各种应用中使用的现有视频传感器网络系统通常由一组视频摄像机组成，这些摄像机通过有线通信信道将数据传输到中央处理单元，在传感器本地几乎没有对视频进行智能处理。缺乏机动性和安装有线通信的困难严重阻碍了从大型视频数据网络中获得最大可能的利益。智力优势：实现这一目标取决于成功解决一些理论和实践挑战，包括摄像机网络的自组织及其操作，场景分析的学习行为模式，可用资源的最佳利用，处理动态拓扑和不同程度的移动性，分布式控制和决策制定，以及开发计算效率高的算法等。提出的研究通过开发一个在每个摄像机节点具有本地智能处理的无线摄像机网络的分层架构来解决这些挑战。更广泛的影响：集成了图像处理、计算机视觉和网络任务的无线视频传感器网络将有助于在视频和网络技术的交叉点开发一个新的、令人兴奋的、具有挑战性的科学研究领域。它将极大地惠及大量现有应用，并创造新的应用，从国土安全、核设施监控和边境安全等国家优先领域，到视频通信和多媒体娱乐等商业利益。这些pi将利用这个机会开发无线视频网络的跨学科课程，这些课程可以通过基于视频的远程学习技术在这两个机构教授。

This proposal focuses on developing an intelligent network of video cameras connected over a wireless communication channel for applications in dynamic scene understanding. Existing video sensor network systems employed in various applications usually comprise of a set of video cameras that transmit data over a wired communication channel to a central processing unit with almost no intelligent processing of the video locally at the senor. The lack of mobility and difficulty of installation of wired communication is a serious impediment to deriving the maximum possible benefit from large networks of video data.Intellectual Merit: Realization of this goal depends on successfully addressing a number of theoretical and practical challenges, including self-organization of camera networks and their operation, learning behavior patterns for scene analysis, optimal utilization of available resources, handling dynamic topology and varying degrees of mobility, distributed control and decision making, and developing computationally efficient algorithms, among others. The proposed research addresses these challenges by developing a hierarchical architecture of a wireless network of video cameras with local intelligent processing at each camera node.Broader Impact: Wireless video sensor networks which integrate image processing, computer vision and networking tasks will help develop a new, exciting and challenging scientific research area at the intersection of video and network technologies. It will immensely benefit a large number of existing applications and create new applications ranging from national priority areas like homeland security, monitoring nuclear installations and border security, to commercial interests like video communications and multimedia entertainment. These PIs will use this opportunity to develop inter-disciplinary courses on wireless video networks that can be taught at both institutions through video-based distance learning technologies.