Sensor Signal Processing

信号处理的基本挑战是：如何最好地感知；如何在网络中分配数据的处理和通信以最大化性能和最小化成本；如何分析数据以提取显著信息。信号处理是将许多现代技术凝聚在一起的粘合剂。支撑移动通信、医学成像、游戏图像渲染和许多其他技术的算法都是在全球信号和图像处理研究社区内开发的。今天，世界是一个无处不在的相互关联的传感环境，需要从产生的大量数据中提取有用的信息。在国防、国土安全和环境监测的应用中，需要收集和组合来自一系列复杂程度迥异的传感器(例如，从卫星成像到地面运动探测器)的数据，以实现对感兴趣场景的持续广域监测。这有助于评估威胁，例如简易爆炸装置的放置、砍伐森林的长期生态影响，或监测诸如火灾或洪水破坏等关键时间事件。在道路上，闭路电视网络、基于路口的压力传感器和GPS对交通流量的外部监控与车载传感器(如激光雷达、雷达和视频)相结合，为驾驶员提供辅助并最终实现自动驾驶系统创造了机会。这项平台提案寻求为我们的研究团队应对这些挑战的基金会提供资金。

The fundamental challenges for signal processing are: how best to sense; how to distribute the processing and communication of the data within the network to maximize performance and minimize cost; how to analyze it to extract the salient information. Signal processing is the glue which holds together much of modern technology. The algorithms underpinning mobile communications, medical imaging, image rendering for games and many other technologies were all developed within the global signal and image processing research community. Today the world is an environment of pervasive interconnected sensing with the associated requirement to extract useful information from the large volumes of data that arise. In applications of defence, homeland security and environmental monitoring there is a need to collect and combine data from a range of sensors of widely differing complexity (e.g. from satellite imaging to ground based motion detectors) to achieve persistent wide area monitoring of a scene of interest. This can assist in the assessment of threats, e.g. the planting of improvised explosive devices, the long-term ecological effects of deforestation, or the monitoring of time critical events such as devastation by fire or flood. On the roads the external monitoring of traffic flow by closed circuit television networks, junction-based pressures sensors and GPS create an opportunity when combined with on-vehicle sensors (e.g. lidar, radar and video) to provide driver assistance and ultimately automatic driving systems. This Platform proposal seeks funding for a foundation for our research team in addressing these challenges.