Building Tomography: Automatic Floor Plan and Radio Map Generation for Indoor Localization

The Hong Kong University of Science and Technology
Department of Computer Science and Engineering


PhD Thesis Defence


Title: "Building Tomography: Automatic Floor Plan and Radio Map Generation 
for Indoor Localization"

By

Mr. Junliang LIU


Abstract

Indoor localization is of great importance for a range of pervasive 
applications, attracting many research efforts in the past decades. Most 
radio-based solutions require 1) priori knowledge of buildings, usually 
represented as a floor plan and 2) a process of site survey, in which radio 
signatures of an interested area are annotated with their real recorded 
locations. Floor plan plays an essential role in many indoor pervasive and 
mobile applications, but its collection and on-site calibration are 
inconvenient and usually prohibitively costly for map providers. Site survey 
involves intensive costs on manpower and time.

In this study, we investigate novel sensors integrated in modern mobile phones 
and leverage user motions to release the human efforts in constructing the 
radio map of a building. We propose Building Tomography, which automatically 
explores indoor architectural layouts and samples radio fingerprints. Building 
Tomography generates a floor plan illustrating a number of key spatial elements 
like rooms, corridors, walls, and other physical features at one level of a 
building, and constructs corresponding radio map towards it. Our idea is based 
on human-centric sensing and crowdsourcing. The popularity of smartphones, with 
rich built-in sensors, enables fine-grained sensory records on human mobility 
and activity. Although the records from one user might be less useful, a large 
amount of contributing users enrich the records to an applicable level so that 
the interior layout of a building emerges.

Our main contribution is to design and implement a building tomography system 
that is inexpensive and pervasive. No building knowledge is required and all 
sensor readings are collected by off-the-shelf smartphones. To validate this 
design, we deploy a prototype system and conduct experiments in an office 
building. Results show that the generated floor plan accurately reflects real 
layout and is able to facilitate many pervasive applications, including indoor 
localization and navigation.


Date:			Wednesday, 19 June 2013

Time:			9:00am – 11:00am

Venue:			Room 3501
 			Lifts 25/26

Chairman:		Prof. Wenjing Ye (MECH)

Committee Members:	Prof. Yunhao Liu (Supervisor)
 			Prof. Kai Chen (Supervisor)
 			Prof. Qiong Luo
 			Prof. Ke Yi
 			Prof. Shiheng Wang (ACCT)
                        Prof. Jianping Wang (Comp. Sci., CityU)


**** ALL are Welcome ****