本文为瑞典林雪平大学（作者：Piotr Rudol）的毕业论文，共115页。

在过去的几十年里，由于多个学科的不断发展，无人机系统（UAS）执行的任务范围一直在稳步增长。提高无人机自主性的目标是扩大无需或仅需极少外部帮助即可执行的任务范围。本文提出了在以摄像机为主要传感器的室外和室内环境中提高无人机自主性的方法。首先，提出了一种融合彩色图像和热像的无人机在室外工作时的目标检测、定位和跟踪方法。具体地，描述了一种用于构建显著性地图的方法，其中人体位置被标记为感兴趣的点。这种地图可用于紧急情况，以提高第一反应者或机器人系统本身的态势感知。另外，同样的方法也被应用于车辆跟踪问题。生成的履带车辆地理位置流允许对车辆超车、进入或离开交叉口等进行定性推理，从而提高了态势感知能力。其次，提出了两种在没有GPS情况下解决UAS室内定位问题的方法。两者都使用摄像头作为主要传感器，并实现室内自主飞行和导航。第一种方法利用与地面机器人的合作，为无人机提供定位信息。第二种方法使用基于标记的视觉姿态估计，其中所有的计算都是在小型飞机上完成的，通过不依赖外部计算能力来增加自主性。

The range of missions performed by UnmannedAircraft Systems (UAS) has been steadily growing in the past decades thanks tocontinued development in several disciplines. The goal of increasing theautonomy of UAS’s is widening the range of tasks which can be carried outwithout, or with minimal, external help. This thesis presents methods forincreasing specific aspects of autonomy of UAS’s operating both in outdoor andindoor environments where cameras are used as the primary sensors. First, amethod for fusing color and thermal images for object detection, geolocationand tracking for UAS’s operating primarily outdoors is presented. Specifically,a method for building saliency maps where human body locations are marked aspoints of interest is described. Such maps can be used in emergency situationsto increase the situational awareness of first responders or a robotic systemitself. Additionally, the same method is applied to the problem of vehicletracking. A generated stream of geographical locations of tracked vehiclesincreases situational awareness by allowing for qualitative reasoning about,for example, vehicles overtaking, entering or leaving crossings. Second, twoapproaches to the UAS indoor localization problem in the absence of GPS-based positioningare presented. Both use cameras as the main sensors and enable autonomousindoor flight and navigation. The first approach takes advantage of cooperationwith a ground robot to provide a UAS with its localization information. Thesecond approach uses marker-based visual pose estimation where all computationsare done onboard a small-scale aircraft which additionally increases itsautonomy by not relying on external computational power.

1. 引言

2. 融合热像和彩色图像进行目标检测、跟踪和定位

3. 小型无人机室内自主导航

4. 结论

更多精彩文章请关注公众号：