NeTS: Small: Integrating Electronic and Visual Signals for Accurate Localization

如今，越来越多的物体可以用它们的无线电子识别器来表示。例如，人们可以通过他们的手机或笔记本电脑的MAC地址来识别，而产品可以通过他们的RFID号码来识别。随着我们的社会日益数字化，使用电子识别器定位对象变得越来越重要。然而，传统的无线定位技术不能满足快速安装的精确和低成本定位的要求。其中一些定位需要昂贵的硬件来实现高精度，这对于大规模部署是不切实际的，而其他一些定位方法，如基于WiFi RSSI的定位不准确，对环境噪声没有鲁棒性。本项目设计了一种名为EVLoc的新的本地化方法。在EVLoc中，由于视觉定位的精度约为0.1-0.5米，因此使用视觉信号(V信号)来提高电子定位的精度。该技术充分利用了V信号的高精度和电子信号(E信号)的普及性。本课题进行了三个主要的研究工作：(1)在给定完整且清晰的E、V信号的情况下，进行目标定位；(2)在实际环境中，当E、V信号不完整且不清晰时，进行目标定位；(3)实现E-V方法，以验证其准确性和有效性。这项开创性的研究开发了一种变革性的方法，将E和V信号整合在一起，以实现精确定位。该方法可应用于许多新兴应用，如对老年人、儿童和那些出于医疗目的表现出危险行为的人进行疫情控制和跟踪。

Nowadays, an increasing number of objects can be represented by their wireless electronic identifiers. For examples, people can be recognized by their cellular phones' or laptops' MAC addresses and products can be identified by their RFID numbers. Localizing objects with electronic identifiers is more and more important as our society is becoming increasingly "digitalized". However, traditional wireless localization techniques cannot meet the rapidly mounting requirements of accurate and cost-effective localization. Some of them need expensive hardware to achieve high accuracy, which is impractical for massive deployments, while others such as WiFi RSSI based localization are inaccurate and not robust to environmental noise. This project designs a novel localization methodology called EVLoc. In EVLoc, visual signals (V signals) are used to improve the accuracy of electronic localization, as the accuracy of visual localization is around 0.1-0.5 meter. This technique fully leverages V signals' high accuracy and the pervasiveness of electronic signals (E signals). The project conducts three main research tasks: (1) perform object localization given complete and distinct E and V signals, (2) perform object localization in practical settings when the E and V signals are not complete and distinct, and (3) implement the E-V approach to validate its accuracy and efficacy. This pioneering research develops a transformative approach that integrates E and V signals for accurate localization. The approach can be applied to many emerging applications such as epidemic control and tracking of elderly people, children, and those exhibiting at-risk behaviors for healthcare purposes.