رسالة جامعية
Indoor Positioning System for Smart Devices
| العنوان: | Indoor Positioning System for Smart Devices |
|---|---|
| المؤلفون: | Yang, Yuan |
| بيانات النشر: | The Ohio State University / OhioLINK, 2021. |
| سنة النشر: | 2021 |
| المجموعة: | Ohiolink ETDs |
| Original Material: | http://rave.ohiolink.edu/etdc/view?acc_num=osu1618789654703694Test |
| مصطلحات موضوعية: | Engineering, indoor positioning, Wi-Fi fingerprinting, visual localization, particle filter, map updating |
| الوصف: | With the proliferation of personal smart devices in the last decade, mobile applications of indoor location-based services (ILBS) have been widely used in public buildings, such as hospitals, malls, school campuses and museums for patient monitoring, security management, asset tracking and indoor navigation, etc. As the core component of ILBS, indoor positioning systems (IPS) have gained increased attention. Since global navigation satellite systems (GNSS) are generally denied in indoor environments, as an alternative solution, many Radio Frequency (RF) based approaches of IPS have been proposed. However, these solutions either need to work in a controlled environment with customized RF infrastructure or could only offer low accuracy localization at the meter level. This work focuses on developing an innovative framework of IPS which is able to perform in any indoor environment without extra infrastructure and offers a robust and accurate localization estimation for smart device users.To achieve this goal, this work initially introduces a mean peak method that is subsequently combined with classification-based Wi-Fi fingerprint positioning (WF) techniques to deal with the challenges of positioning in a weak RSS environment. As the next step, the database updating problem for WF, including an innovative Bayes-inference-based sensor fusion framework which integrates WF and visual fingerprint positioning (VF) is addressed. The method provides both a location estimation and a heading direction estimation. In order to further improve the accuracy of the localization, a state-of-the-art visual localization algorithm, InLoc, is introduced in combination with WF for constructing a new IPS. Additionally, with the help of WF, the computational cost of InLoc is reduced compared to the original one. Moreover, to improve the robustness of the system, a particle filter and map updating function are introduced to optimize localization results and address the RSS variance problem. In summary, the proposed methodologies of this work define a framework for constructing an innovative IPS, a new path to build indoor positioning systems which are robust, easy to use and easy to deploy. |
| Original Identifier: | oai:etd.ohiolink.edu:osu1618789654703694 |
| نوع الوثيقة: | text |
| اللغة: | English |
| الإتاحة: | http://rave.ohiolink.edu/etdc/view?acc_num=osu1618789654703694Test |
| حقوق: | unrestricted This thesis or dissertation is protected by copyright: all rights reserved. It may not be copied or redistributed beyond the terms of applicable copyright laws. |
| رقم الانضمام: | edsndl.OhioLink.oai.etd.ohiolink.edu.osu1618789654703694 |
| قاعدة البيانات: | Networked Digital Library of Theses & Dissertations |
| الوصف غير متاح. |