Feedback Control of Visual Appearance With Maximally Sensitive Sensors for Decentralized Event Detection and Security

使用最灵敏的传感器对视觉外观进行反馈控制，以实现分散式事件检测和安全

• 批准号: 0323693
• 负责人: Bijoy Ghosh
• 金额: $ 25.61万
• 依托单位: Washington University
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2003
• 资助国家: 美国
• 起止时间: 2003-08-15 至 2007-07-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=0323693&HistoricalAwards=false
• 关键词: Feedback; Control; Visual; Appearance; Maximally

In many surveillance problems, one would use several c.c.d. (charged coupled device) camerasconnected in a network, that would generate a stream of images. The problem is to process theimage stream so as to detect a typical event in the scene from the observed recording of the imagesequences. Such an event might be quantified, for example, by a sudden movement in the scene orexistence of an unfamiliar target. Difficulty in identifying an event is that the 'event characteristic'is hard to quantify and extracted from the complex scene imagery, a process that would typicallyrequired to be done in real time. Additional difficulty arises from the fact that a single view of thescene may not be enough to isolate an event - hence the need for multiple view over an interval oftime.We propose that every visual stream be quantified and represented locally by an internal dy-namicmodel that produces its own spatiotemporal sequence. The importance of this internal repre-sentationis that it does not represent the entire scene, or all the events in the scene. Rather, its solepurpose is to amplify specific intruding event anywhere in the scene and to respond as to 'when'and 'where' it has occurred. The design problem that we propose to investigate is to synthesizethe internal dynamics so as to respond 'maximally' in presence of an intruding event as opposedto somewhat more routine events. The internal dynamics would be implemented on a processorlocally connected to the sensor and in its simplest form would be a directionally selective flowmodel with inputs from the scene images taken by the camera. The model parameters are tunedto respond to specific events in the scene. In order to be able to synthesize the above described'maximally sensitive sensor', we subdivide this project into three distinct parts.The first is to introduce 'appearance models' to represent a sequence of images taken by acamera. Such an appearance model results in a suitable data-compression and is particularly usefulwhen a suitable set of appearances have to be isolated and detected in the scene. Our second goalis to introduce a suitable internal representation of the observed spatiotemporal signal using 'flowmodels'. The proposed flow models have flow velocities that are dependant on the direction ofpropagation and can be altered by the magnitude and position of the input target events. The flowmodels can be tuned by feedback to produce maximal activity to selected targets in the scene. Ourthird goal is to network a distributed set of cameras, together with associated internal models, fordistributed detection. The model activity from each camera sensor, confirming detection of a localevent, is fused together in order to detect a spatio-temporal global event.The intellectual merits of this project are described as follows. The first is Selective Encodingof the Spatio-Temporal Events in the Scene through Appearance Models. The second is InternalModelling and Feedback Tuning for Maximal Response. Finally the third is Decentralized Detec-tionand Feedback Control of the Network Structure.Broader impact of the proposal includes interaction between Signal Processing, Sensor BasedControl and Sensor Networks. Feedback control for sensor tuning and network reconfiguration aretwo research areas that this proposal makes the most impact.The project would be carried out by the PI with 2 PhD students at the Center for BioCyberneticsand Intelligent Systems and would also provide an interdisciplinary training ground for seniorundergraduates from Computer Science, Electrical and Systems Engineering.

在许多监视问题中，人们会使用几个c.c.d.。（电荷耦合器件）连接在网络中的相机，这将产生图像流。问题是处理图像流，以便从观察到的图像序列记录中检测场景中的典型事件。这样的事件可以量化，例如，通过场景中的突然移动或不熟悉目标的存在。识别事件的困难在于“事件特征”难以量化和从复杂的场景图像中提取，这一过程通常需要在真实的时间内完成。额外的困难来自于这样一个事实，即场景的单个视图可能不足以隔离一个事件-因此需要在一段时间内的多个视图。我们建议，每个视觉流被量化和表示本地的内部动态模型，产生自己的时空序列。这种内部再现的重要性在于它并不代表整个场景，或者场景中的所有事件。相反，它的唯一目的是放大场景中任何地方的特定入侵事件，并对“何时”和“何地”发生做出反应。设计问题，我们建议调查是synthesizethe内部动态，以便响应“最大限度地”在存在的入侵事件作为opposedto有点更常规的事件。内部动力学将在本地连接到传感器的处理器上实现，其最简单的形式将是一个方向选择性流动模型，输入来自相机拍摄的场景图像。调整模型参数以响应场景中的特定事件。为了能够综合上述的“最高灵敏度传感器”，我们将这个项目分为三个不同的部分。第一个是引入“外观模型”来表示由相机拍摄的图像序列。这样的外观模型导致合适的数据压缩，并且当必须在场景中隔离和检测合适的外观集时特别有用。我们的第二个目标是引入一个合适的内部表示所观察到的时空信号使用“流模型”。所提出的流模型具有依赖于传播方向的流速，并且可以通过输入目标事件的大小和位置来改变流速。可以通过反馈来调整流模型，以产生对场景中所选目标的最大活动。我们的第三个目标是将一组分布式摄像机以及相关的内部模型联网，以进行分布式检测。从每个摄像机传感器的模型活动，确认检测到一个localevent，融合在一起，以检测一个时空的全局事件。第一种是通过外观模型对场景中的时空事件进行选择性编码。第二个是最大响应的内部建模和反馈调整。最后是网络结构的分散检测和反馈控制，该建议的更广泛影响包括信号处理，基于传感器的控制和传感器网络之间的相互作用。传感器调谐反馈控制和网络重构是该项目最具影响力的两个研究领域，该项目将由PI与生物控制与智能系统中心的两名博士生共同完成，并将为计算机科学、电气和系统工程专业的高年级本科生提供跨学科的培训。