本文为美国大峡谷州立大学（作者：David Qorashi）的硕士论文，共40页。

国际民用航空组织（ICAO）将UAV定义为一种无须人类驾驶的飞行器，通常称为无人机。目前主要用于国防和安全部门，但正以惊人的强大功能和廉价形式向一般市场转移。虽然无人机目前仅限于在美国的非商业娱乐用途，但预计它们很快将被广泛应用于商业和消费。无人机可能会彻底改变各种商业领域，包括私人安全、农业实践、产品运输，甚至空中广告。《商业内幕》预测，未来十年全球累计花费98亿美元的无人机预计支出中，有12%将用于商业目的。

目前，大多数无人机都由某种经典的操纵杆或多点触控遥控器控制。虽然无人机制造商已经提高了其产品的整体可控性，但对于缺乏经验的用户来说，今天设计的大多数无人机控制仍具有相当的挑战性。为了降低无人机的可控性挑战，使学习曲线平坦化，姿态控制可以用来改善无人机的驾驶性能。本研究的目的是通过手势和其他非传统控制方式的使用，开发、评估一种改进的、更直观的无人机飞行方法。其目标是使用并测试一个端到端的无人机系统，为无人机新手用户提供一个简单易行的控制界面。我们期望通过实施基于手势的导航，新用户将有一个全面的愉快、安全体验，快速学习如何轻松导航无人机，并避免在飞行器处于初始学习曲线时丢失或损坏。在这项研究过程中，我们了解到，虽然这种方法确实提供了许多承诺，但仍有许多技术使这个问题比预期更具挑战性。本文详细介绍了我们解决这个问题的方法，分析了我们收集到的用户数据，并总结了经验教训。

Unmanned aerial vehicles (UAVs), commonlyknown as drones, are defined by the International Civil Aviation Organization(ICAO) as an aircraft without a human pilot on board. They are currentlyutilized primarily in the defense and security sectors but are moving towardsthe general market in surprisingly powerful and inexpensive forms. While dronesare presently restricted to non-commercial recreational use in theUSA, it isexpected that they will soon be widely adopted for both commercial and consumeruse. Potentially, UAVs can revolutionize various business sectors includingprivate security, agricultural practices, product transport and maybe evenaerial advertising. Business Insider foresees that 12% of the expected $98billion cumulative global spending on aerial drones through the followingdecade will be for business purposes.[28] At the moment, most drones arecontrolled by some sort of classic joystick or multitouch remote controller.While drone manufactures have improved the overall controllability of theirproducts, most drones shipped today are still quite challenging forinexperienced users to pilot. In order to help mitigate the controllabilitychallenges and flatten the learning curve, gesture controls can be utilized toimprove piloting UAVs. The purpose of this study was to develop and evaluate animproved and more intuitive method of flying UAVs by supporting the use of handgestures, and other non-traditional control modalities. The goal was to employand test an end-to-end UAV system that provides an easy-touse control interfacefor novice drone users. The expectation was that by implementing gesturebasednavigation, the novice user will have an overall enjoyable and safe experiencequickly learning how to navigate a drone with ease, and avoid losing ordamaging the vehicle while they are on the initial learning curve. During thecourse of this study we have learned that while this approach does offer lotsof promise, there are a number of technical challenges that make this problemmuch more challenging than anticipated. This thesis details our approach to theproblem, analyzes the user data we collected, and summarizes the lessonslearned.

1 引言

2 可控性问题

3 相关研究工作

4 论文研究中的假设

5 手势控制器的设计与实现

6 实验设计

7 实验结果分析

8 经验教训与本文贡献

9 进一步研究内容

更多精彩文章请关注公众号：