Smart On-Board UAV Thermal Signature Analysis for Surveillance, Tracking and Navigation

在过去十年中，无人驾驶飞行器（uav）一直是世界航空航天工业中最具活力的增长领域，预计未来十年的支出将翻一番，从目前全球每年66亿美元的支出增加到2020年的114亿美元。尽管军事应用迄今为止主导了技术开发和市场扩张，但无人机在民用应用中的使用预计将推动这一增长。无人机比有人驾驶飞机更小，更容易储存、操作和运输。它们也非常适合执行长期监测任务，能够在对人类居住者有害的环境中发挥作用。无人机也可以用于支持政府应用，如边境控制/安全，执法和监视，这是预期的具体增长。尽管比有人驾驶飞机更具成本效益，并且对每个飞行模式的平均探测率为70%，但目前用于无人机监视的光学和热成像系统仅限于预定义的固定坐标，由操作员编程。图像检测也是一个关键的限制，高度依赖于飞行高度，相机方向和操作员的经验。RNC航空电子公司（RNC）的目标是开发一种新型集成机载成像系统，该系统将采用“热特征”分析算法，用于检测车辆/船舶留下的热迹，并实时（与其他机载传感器一起）探索这一点，以调整其飞行计划。一旦确定了感兴趣的对象，将使用视频分析技术分析其运动方向。研究结果将用于自主改变无人机预定义的监视飞行计划，以确保目标被持续跟踪。这种方法为当前能力提供了一个阶梯式的改变，具有改变无人机监视中使用热成像方式的巨大潜力。

Unmanned Aerial Vehicles (UAVs) have been the most dynamic growth sector of the worldaerospace industry in the last 10 yrs with expenditure forecast to double over the next decadefrom current worldwide expenditures of $6.6 bn annually to almost $11.4 bn by 2020.Although military applications have to date dominated technology development and marketexpansion, it is the use of UAVs in civil applications that is expected to drive much of thisgrowth.Being smaller than manned aircraft, UAVs are more easily & cost effectively stored, operatedand transported. They are also well suited to perform long monitoring tasks with an ability tofunction in environments hazardous to human occupants. UAVs can also be utilised to supportgovernmental applications such as border control/security, law enforcement and surveillanceand it is here that specific growth is expected.Although more cost effective than manned aircrafts and boasting an average detection rate foreach flight pattern of >70%, current optical and thermal imagery systems used in UAVsurveillance are limited to predefined fixed coordinates, programmed by the operator. Imagedetection is also a key limitation and highly dependent on flight altitude, camera orientationand the experience of the operator.RNC Avionics (RNC) aim to develop a novel integrated on-board imaging system that willemploy algorithms for 'thermal signature’ analysis, for detecting heat trails left by avehicle/vessel and explore this in real time (together other on-board sensors) for adjustment ofits flight plan. Once the object of interest has been identified, the direction of its movementwill be analysed using the video analytics techniques. The results will be used toautonomously alter the UAV pre-defined surveillance flight plan to ensure that the object iscontinuously tracked. This approach offers a step change in current capabilities withsignificant potential to change the way thermal imaging is used in UAV surveillance.