Smartphone Positioning with Street View Image Database

Autor:	Hao-hsiang Hu, 胡皓翔
Rok vydání:	2013
Druh dokumentu:	學位論文 ; thesis
Popis:	101 There have been various methods and research on navigation such as using global position system (GPS), accelerometer and gyroscope. However, those sensors possess limitations. In this thesis a new approach which combines street view image database with real-time image from a smart phone for positioning is developed. This thesis proposes a concept of image matching to generate image positioning. An image database with information such as photo position and direction is built up to cover a territory. An image of current location is obtained via a smart phone or a camera and is compared with the image database to select database images with the highest similarities. By estimating the position and correcting the path using pedestrian dead reckoning (PDR) and Kalman filter (KF), the experimental results show a higher navigation accuracy.
Databáze:	Networked Digital Library of Theses & Dissertations
Externí odkaz:	http://ndltd.ncl.edu.tw/handle/86354494678802350508 Zobrazit plný text záznamu