ArticlePDF Available

Abstract and Figures

This document summarizes the major milestones in mobile Augmented Reality between 1968 and 2014. Major parts of the list were compiled by the member of the Christian Doppler Laboratory for Handheld Augmented Reality in 2010 (author list in alphabetical order) for the ISMAR society. Later in 2013 it was updated, and more recent work was added during preparation of this report. Permission is granted to copy and modify.
Content may be subject to copyright.
contact: Clemens Arth arth@icg.tugraz.at
The History of Mobile Augmented
Reality
Developments in Mobile AR over the last almost 50 years
Clemens Arth, Lukas Gruber, Raphael Grasset, Tobias Langlotz,
Alessandro Mulloni, Dieter Schmalstieg, Daniel Wagner
Inst. for Computer Graphics and Vision
Graz University of Technology, Austria
Technical Report
ICG–TR–2015-001
Graz, May 11, 2015
arXiv:1505.01319v1 [cs.HC] 6 May 2015
Abstract
This document summarizes the major milestones in mobile Augmented Reality
between 1968 and 2014. Mobile Augmented Reality has largely evolved over the
last decade, as well as the interpretation itself of what is Mobile Augmented
Reality. The first instance of Mobile AR can certainly be associated with the
development of wearable AR, in a sense of experiencing AR during locomotion
(mobile as a motion). With the transformation and miniaturization of physical
devices and displays, the concept of mobile AR evolved towards the notion of
”mobile device”, aka AR on a mobile device. In this history of mobile AR we
considered both definitions and the evolution of the term over time.
Major parts of the list were initially compiled by the member of the Christian
Doppler Laboratory for Handheld Augmented Reality in 2009 (author list in
alphabetical order) for the ISMAR society. More recent work was added in
2013 and during preparation of this report.
Permission is granted to copy and modify. Please email the first author if you
find any errors.
Keywords: Technical Report, Mobile Augmented Reality, History
Introduction
This document summarizes the major milestones in mobile Augmented Re-
ality between 1968 and 2014. Mobile Augmented Reality has largely evolved
over the last decade, as well as the interpretation itself of what is Mobile
Augmented Reality. The first instance of Mobile AR can certainly be as-
sociated with the development of wearable AR, in a sense of experiencing
AR during locomotion (mobile as a motion). With the transformation and
miniaturization of physical devices and displays, the concept of mobile AR
evolved towards the notion of ”mobile device”, aka AR on a mobile device.
In this history of mobile AR we considered both definitions and the evolution
of the term over time.
Major parts of the list were initially compiled by the member of the The
list was compiled by the member of the Christian Doppler Laboratory for
Handheld Augmented Reality1in 2009 (author list in alphabetical order)
for the ISMAR society. More recent work was added in 2013 and during
preparation of this report.
Permission is granted to copy and modify. Please email the first author
if you find any errors.
(a) Research (b) Mobile PC (c) Mobile Phone (d) Hardware
(e) Standard (f) Game (g) Tool (h) Deal
Figure 1: Icons used throughout this report for a rough categorization of
related research, development and events.
1CDL on Handheld AR: http://studierstube.org/handheld_ar/
1
(a) (b) (c)
(d) (e)
Figure 2: (a): Sutherland’s system in [66]. (b): Conceptual Tablet Computer
by Kay in 1972 [31]. (c): First handheld mobile phone by Motorola in
1973. (d): Caudell and Mizell coining AR in 1992 [7]. (e): IBM smartphone
presented in 1992.
1968
Ivan Sutherland [66] creates the first augmented reality sys-
tem, which is also the first virtual reality system (see Fig.2(a)
left). It uses an optical see-through head-mounted display that
is tracked by one of two different 6DOF trackers: a mechanical tracker and
an ultrasonic tracker. Due to the limited processing power of computers at
that time, only very simple wireframe drawings could be displayed in real
time.
1972
The first conceptual tablet computer was proposed in 1972 by
Alan Kay, named the Dynabook [31]. The Dynabook was proposed
as personal computer for children, having the format factor of a tablet with
2
a mechanical keyboard (really similar design from the One Laptop per Child
project started in 2005). The Dynabook is probably recognized as being the
precursor of the tablet computers decades before the iPad (see Fig. 2(b)).
1973
The first handheld mobile phone was presented by Motorola and
demonstrated in April 1973 by Dr Martin Cooper [1]. The mobile
named DynaTAC for Dynamic Adaptive Total Area Coverage was supporting
only 35 minutes of call (see Fig. 2(c)).
1982
The first laptop, the Grid Compass21100 is released, which was
also the first computer to use a clamshell design. The Grid Compass
1100 had an Intel 8086 CPU, 350 Kbytes of memory and a display
with a resolution of 320x240 pixels, which was extremely powerful for that
time and justified the enormous costs of 10.000 USD. However, its weight of
5kg made it hardly portable.
1992
Tom Caudell and David Mizell coin the term ”augmented reality”
to refer to overlaying computer-presented material on top of the real
world [7] (see Fig.2(d)). Caudell and Mizell discuss the advantages of
AR versus VR such as requiring less processing power since less pixels have to
be rendered. They also acknowledge the increased registration requirements
in order to align real and virtual.
At COMDEX 1992, IBM and Bellsouth introduce the first smart-
phone, the IBM Simon Personal Communicator3, which was released
in 1993 (see Fig.2(e)). The phone has 1 Megabyte of memory and a
B/W touch screen with a resolution of 160 x 293 pixels. The IBM
Simon works as phone, pager, calculator, address book, fax machine, and
e-mail device. It weights 500 grams and cost 900 USD.
2http://home.total.net/~hrothgar/museum/Compass/
3Wikipedia: http://en.wikipedia.org/wiki/Simon_(phone)
3
(a) (b) (c)
Figure 3: (a): Chameleon system proposed by Fitzmaurice [15]. (b):
NAVSTAR-GPS goes live in 1993. (c): Apple Newton Message Pad 100.
1993
Loomis et al . develop a prototype of an outdoor navigation
system for visually impaired [38]. They combine a note-
book with a differential GPS receiver and a head-worn elec-
tronic compass. The application uses data from a GIS (Geographic Informa-
tion System) database and provides navigational assistance using an ”acous-
tic virtual display”: labels are spoken using a speech synthesizer and played
back at correct locations within the auditory space of the user.
Fitzmaurice creates Chameleon (see Fig.3(a)), a key exam-
ple of displaying spatially situated information with a tracked
hand-held device. In his setup the output device consists of a
4” screen connected to a video camera via a cable [15]. The video camera
records the content of a Silicon Graphics workstation’s large display in or-
der to display it on the small screen. Fitzmaurice uses a tethered magnetic
tracker (Ascension bird) for registration in a small working environment.
Several gestures plus a single button allow the user to interact with the mo-
bile device. Chameleon’s mobility was strongly limited due to the cabling.
It did also not augment in terms of overlaying objects on a video feed of the
real world.
In December 1993 the Global Positioning System (GPS, official
name ”NAVSTAR-GPS”) achieves initial operational capability (see
Fig.3(b)). Although GPS4was originally launched as a military ser-
vice, nowadays millions of people use it for navigation and other tasks such
as geo-caching or Augmented Reality. A GPS receiver calculates its position
by carefully timing the signals sent by the constellation of GPS satellites.
The accuracy of civilian GPS receivers is typically in the range of 15 meter.
4Wikipedia: http://en.wikipedia.org/wiki/Global_Positioning_System
4
(a) (b)
(c) (d) (e)
Figure 4: (a): Milgram Continuum [40]. (b): Rekimotos NaviCam system
[57]. (c): Rekimoto’s matrix marker [56]. (d) and (e): Touring Machine by
Feiner et al. [14].
More accuracy can be gained by using Differential GPS (DGPS) that uses
correction signals from fixed, ground-based reference stations.
The Apple Newton Message Pad 100 was one of the earliest com-
mercial personal digital assistant (PDA)5. Equipped with a
stylus and handwritten recognition, and feature a screen in black and white
of 336x240 pixels (see Fig. 3(c)).
1994
Steve Mann starts wearing a webcam for almost 2 years. From
1994-1996 Mann wore a mobile camera plus display for almost every
waking minute. Both devices were connected to his website allowing online
visitors to see what Steve was seeing and to send him messages that would
show up on his mobile display6
Paul Milgram and Fumio Kishino write their seminal paper ”Tax-
onomy of Mixed Reality Visual Displays” in which they define the
Reality-Virtuality Continuum [40] (see Fig.4(a)). Milgram and
5Wikipedia: http://en.wikipedia.org/wiki/MessagePad
6S. Mann, Wearable Wireless Webcam, personal WWW page. wearcam.org
5
Kishino describe a continuum that spans from the real environment to the
virtual environment. In between there are Augmented Reality, closer to the
real environment and Augmented Virtuality, which is closer to the virtual
environment. Today Milgram’s Continuum and Azuma’s definition (1997)
are commonly accepted as defining Augmented Reality.
1995
Jun Rekimoto and Katashi Nagao create the NaviCam, a
tethered setup, similar to Fitzmaurice’s Chameleon [57] (see
Fig.4(b)). The NaviCam also uses a nearby powerful worksta-
tion, but has a camera mounted on the mobile screen that is used for optical
tracking. The computer detects color-coded markers in the live camera im-
age and displays context sensitive information directly on top of the video
feed in a see-through manner.
Benjamin Bederson introduced the term Audio Augmented Re-
ality by presenting a system that demonstrated an augmentation
of the audition modality [4]. The developed prototype uses a MD-
player which plays audio information based on the tracked position of the
user as part of a museum guide.
1996
Jun Rekimoto presents 2D matrix markers7(square-shaped bar-
codes), one of the first marker systems to allow camera tracking
with six degrees of freedom [56] (see Fig.4(c)).
1997
Ronald Azuma presents the first survey on Augmented Reality.
In his publication, Azuma provides a widely acknowledged definition
for AR [3], as identified by three characteristics:
it combines real and virtual
it is interactive in real time
it is registered in 3D.
7http://www.sonycsl.co.jp/person/rekimoto/matrix/Matrix.html
6
Steve Feiner et al. present the Touring Machine, the first mo-
bile augmented reality system (MARS) [14] (see Fig.4(d) and
Fig. 4(e)). It uses a see-through head-worn display with integral orientation
tracker; a backpack holding a computer, differential GPS, and digital radio
for wireless web access; and a hand-held computer with stylus and touchpad
interface8.
Thad Starner et al. explore possible applications of mobile aug-
mented reality, creating a small community of users equipped
with wearable computers interconnected over a network [65].
The explored applications include an information system for
offices, people recognition and coarse localization with infrared beacons.
Philippe Kahn invents the camera phone9, a mobile phone which is
able to capture still photographs (see Fig.5(a)). Back in 1997, Kahn
used his invention to share a picture of his newborn daughter with
more than 2000 relatives and friends, spread around the world. Today more
than half of all mobile phones in use are camera phones.
Sony releases the Glasstron, a series of optical HMD (optionally
see-through) for the general public. Adoption was rather small,
but the affordable price of the HMD made it really popular in AR research
labs and for the development of wearable AR prototype (see Fig. 5(b)).
1998
Bruce Thomas et al. present ”Map-in-the-hat”, a backpack-
based wearable computer that includes GPS, electronic com-
pass and a head-mounted display [70] (see Fig.5(c)). At this
stage the system was utilized for navigation guidance, but it later evolved
into Tinmith, an AR platform used for several other AR projects10.
1999
Hirokazu Kato and Mark Billinghurst present ARToolKit, a pose
tracking library with six degrees of freedom, using square fiducials
and a template-based approach for recognition [30]. ARToolKit is
available as open source under the GPL license and is still very popular in
the AR community (see Fig. 5(d)).
8MARS: http://graphics.cs.columbia.edu/projects/mars/mars.html
9Wikipedia Camera Phone: http://en.wikipedia.org/wiki/Camera_phone
10Tinmith webpage: http://www.tinmith.net/
7
(a) (b)
(c) (d) (e)
Figure 5: (a): Camera Phone Development by Kahn. (b): Sony Glasstron
optical HMD in 1997. (c): Thomas et al.’s Tinmith system [70]. (d): AR-
ToolKit for pose tracking in 6DOF [30]. (e): Palm VII, the first consumer
LBS device.
Tobias H¨ollerer et al. develop a mobile AR system that allows
the user to explore hypermedia news stories that are located at
the places to which they refer and to receive a guided campus
tour that overlays models of earlier buildings [25] (see Fig. 6(a)). This
was the first mobile AR system to use RTK GPS and an inertial-magnetic
orientation tracker.
Tobias H¨ollerer et al. present a mobile augmented reality sys-
tem that includes indoor user interfaces (desktop, AR tabletop,
and head-worn VR) to interact with the outdoor user [26] (see
Fig. 6(b)). While outdoor users experience a first-person spatialized mul-
timedia presentation via a head-mounted display, indoor users can get an
overview of the outdoor scene.
Jim Spohrer publishes the Worldboard concept, a scalable in-
frastructure to support mobile applications that span from low-end
location-based services, up to high-end mobile AR [64]. In his paper,
Spohrer also envisions possible application cases for mobile AR, and social
implications.
8
(a) (b) (c)
Figure 6: (a): H¨ollerer et al .’s MARS system [25]. (b): ollerer et al .’s user
interface [26]. (c) Benfon Esc! NT2002, the first GSM phone with a built-in
GPS sensor.
The first consumer LBS device was the Palm VII, only support-
ing zip code based location services (see Fig.5(d)). 2 years later,
different mobile operators provided different location based services using
private network technology11.
Benefon Esc! NT200212, the first GSM phone with a built-in
GPS receiver is released in late 1999 (see Fig. 6(c)). It had a black
and white screen with a resolution of 100x160 pixels. Due to limited
storage, the phone downloaded maps on demand. The phone also included a
friend finder that exchanged GPS positions with other Esc! devices via SMS.
The wireless network protocols 802.11a/802.11b13 - commonly known
as WiFi - are defined. The original version - obsolete - specifies
bitrates of 1 or 2 megabits per second (Mbit/s), plus forward error
correction code.
2000
Bruce Thomas et al. present AR-Quake, an extension
to the popular desktop game Quake [69] (see Fig. 7(a)).
ARQuake is a first-person perspective application which
is based on a 6DOF tracking system using GPS, a digital compass and vision-
based tracking of fiducial markers. Users are equipped with a wearable com-
puter system in a backpack, an HMD and a simple two-button input device.
The game can be played in- or outdoors where the usual keyboard and mouse
commands for movement and actions are performed by movements of the user
in the real environment and using the simple input interface.
11Wikipedia: http://en.wikipedia.org/wiki/Palm_VII
12http://www.benefon.de/products/esc/
13Wikipedia: http://en.wikipedia.org/wiki/802.11
9
(a) (b)
(c) (d)
Figure 7: (a): ARQuake by Thomas et al. [69]. (b): mPARD system by
Regenbrecht and Specht [53]. (c): BARS system by Julier et al. [28]. (d):
First commercial camera phone in 2000.
Regenbrecht and Specht present mPARD, using analogue
wireless video transmission to a host computer which is taking
the burden of computation off the mobile hardware platform
[53] (see Fig. 7(b)). The rendered and augmented images are sent back to
the visualization device over a separate analog channel. The system can op-
erate within 300m outdoors and 30m indoors, and the batteries allow for an
uninterrupted operation of 5 hours at max.
Fritsch et al. introduce a general architecture for large scale AR
system as part of the NEXUS project. The NEXUS model in-
troduces the notion of augmented world using distributed data man-
agement and a variety of sensor system [16].
Simon Julier et al. present BARS, the Battlefield Augmented
Reality System [28] (see Fig. 7(c)). The system consists of
a wearable computer, a wireless network system and a see-
through HMD. The system targets the augmentation of a battlefield scene
with additional information about environmental infrastructure, but also
about possible enemy ambushes.
10
Sharp corporation releases the first commercial camera phone
to public (see Fig. 7(d)). The official name of the phone is J-SH0414.
The phones’ camera has a resolution of 0.1 megapixels.
At ISAR, Julier et al. described the problem of information overload
and visual clutter within mobile Augmented Reality [27]. They pro-
posed information filtering for mobile AR based on techniques such
as physically-based methods, methods using the spatial model of interac-
tion, rule-based filtering, and a combination of these methods to reduce the
information overload in mobile AR scenarios.
2001
Joseph Newman et al. present the BatPortal [48], a PDA-
based, wireless AR system (see Fig.8(a)). Localization is per-
formed by measuring the travel time of ultra-sonic pulses be-
tween specially built devices worn by the user, so-called Bats, and fixed
installed receivers deployed in the floors ceilings building-wide. The system
can support an HMD-based system, but also the more well known BatPortal
using a handheld device. Based on a fixed configuration of the PDA carried
and the personal Bat worn, the direction of the users view is estimated, and
a model of the scene with additional information about the scene is rendered
onto the PDA screen.
Hara et al. introduce TOWNWEAR, an outdoor system that
uses a fiber optic gyroscope for orientation tracking [61] (see
Fig.8(b)). The high precision gyroscope is used to measure the
3DOF head direction accurately with minimal drift, which is then compen-
sated by tracking natural features.
urgen Fruend et al. present AR-PDA, a concept for build-
ing a wireless AR system and a special prototype of palm-sized
hardware [17] (see Fig.8(c)). Basic design ideas include the
augmentation of real camera images with additional virtual objects, for ex-
ample for illustration of functionality and interaction with commonly used
household equipment.
Reitmayr and Schmalstieg present a mobile, multi-user AR
system [54] (see Fig.8(d)). The ideas of mobile augmented
reality and collaboration between users in augmented shared
space are combined and merged into a hybrid system. Communication is
14http://k-tai.impress.co.jp/cda/article/showcase_top/3913.html
11
(a) (b)
(c) (d) (e)
(f) (g)
Figure 8: (a): BatPortal by Newman et al. [48]. (b): TOWNWEAR system
by Hara et al. [61]. (c): Wireless AR setup concept by Fruend et al. [17]. (d):
Multi-user AR system by Reitmayr and Schmalstieg [54]. (e): ARCHEOGU-
IDE by Flahakis et al. [72]. (f): Mobile AR restaurant guide by Bell et al.
[5]. (g): First AR browser by Kooper and MacIntyre [34].
12
performed using LAN and wireless LAN, where mobile users and stationary
users are acting in a common augmented space.
Vlahakis et al. present Archeoguide, a mobile AR sys-
tem for cultural heritage sites [72] (see Fig.8(e)). The sys-
tem is built around the historical site of Olympia, Greece.
The system contains a navigation interface, 3D models of ancient temples and
statues, and avatars which are competing for the win in the historical run
in the ancient Stadium. While communication is based on WLAN, accurate
localization is performed using GPS. Within the system a scalable setup of
mobile units can be used, starting with a notebook sized system with HMD,
down to palmtop computers and Pocket PCs.
Kretschmer et al. present the GEIST system, a system for
interactive story-telling within urban and/or historical envi-
ronments [35]. A complex database setup provides information
queues for the appearance of buildings in ancient times or historical facts
and events. Complex queries can be formulated and stories can be told by
fictional avatars or historical persons.
Columbia’s Computer Graphics and User Interfaces Lab does
an outdoor demonstration of their mobile AR restaurant guide
at ISAR 2001, running on their Touring Machine [5] (see
Fig.8(f)). Pop-up information sheets for nearby restaurants are overlaid on
the user’s view, and linked to reviews, menus, photos, and restaurant URLs.
Kooper and MacIntyre create the RWWW Browser, a mobile
AR application that acts as an interface to the World Wide Web
[34] (see Fig.8(g)). It is the first AR browser. This early
system suffers from the cumbersome AR hardware of that time, requiring
a head mounted display and complicated tracking infrastructure. In 2008
Wikitude implements a similar idea on a mobile phone.
2002
Michael Kalkusch et al. present a mobile augmented reality
system to guide a user through an unfamiliar building to a
destination room [29] (see Fig. 9(a)). The system presents a
world-registered wire frame model of the building labeled with directional in-
formation in a see-through heads-up display, and a three-dimensional world-
in-miniature (WIM) map on a wrist-worn pad that also acts as an input
device. Tracking is done using a combination of wall-mounted ARToolkit
13
(a) (b)
(c) (d)
(e) (f) (g)
Figure 9: (a): Navigation system by Kalkusch et al. [29]. (b): ARPad by
Mogilev et al. [42]. (c): Human Pacman by Cheok et al. [8]. (d): iLamps
system by Raskar et al. [52]. (e): Indoor AR guidance system by Wagner
and Schmalstieg [76]. (f) Siemens SX1 AR game ”Mozzies”. (g): Mobile
Authoring system by Guven and Feiner [21].
14
markers observed by a head-mounted camera, and an inertial tracker.
Leonid Naimark and Eric Foxlin present a wearable low-
power hybrid visual and inertial tracker [45]. This
tracker, later to be known as InterSenses IS-1200, can be used
for tracking in large scale, such as a complete building. This is achieved by
tracking a newly designed 2-D barcode with thousands of different codes and
combining the result with an inertial sensor.
Mogilev et al. introduce the AR Pad, an ad-hoc mobile AR
device equipped with a spaceball controller [42] (see Fig 9(b)).
2003
Adrian David Cheok et al. present the Human Pacman [8]
(see Fig. 9(c)). Human Pacman is an interactive ubiquitous
and mobile entertainment system that is built upon position
and perspective sensing via Global Positioning System and inertia sensors;
and tangible human-computer interfacing with the use of Bluetooth and ca-
pacitive sensors. Pacmen and Ghosts are now real human players in the
real world experiencing mixed computer graphics fantasy-reality provided by
using wearable computers that are equipped with GPS and inertia sensors
for players’ position and perspective tracking. Virtual cookies and actual
tangible physical objects with Bluetooth devices and capacitive sensors are
incorporated into the game play to provide novel experiences of seamless
transitions between real and virtual worlds.
Ramesh Raskar et al. present iLamps [52] (see Fig. 9(d)). This
work created a first prototype for object augmentation with a
hand-held projector-camera system. An enhanced projector
can determine and respond to the geometry of the display surface, and can
be used in an ad-hoc cluster to create a self-configuring display. Furthermore
interaction techniques and co-operation between multiple units are discussed.
Daniel Wagner and Dieter Schmalstieg present an indoor AR
guidance system running autonomously on a PDA [76] (see
Fig. 9(e)). They exploit the wide availability of consumer
devices with a minimal need for infrastructure. The application provides the
user with a three-dimensional augmented view of the environment by using
a Windows Mobile port of ARToolKit for tracking and runs directly on the
PDA.
15
(a) (b)
(c)
(d)
Figure 10: (a): Tracking 3D markers by M¨ohring et al. [43]. (b): Visual
Codes by Rohs and Gfeller [58]. (c): OSGAR system by Coelho et al . [9].
(d): The Invisible Train [74].
The Siemens SX1 is released, coming with the first commercial
mobile phone AR camera game called Mozzies (also known as
Mosquito Hunt) (see Fig. 9(f)). The mosquitoes are superim-
posed on the live video feed from the camera. Aiming is done by moving the
phone around so that the cross hair points at the mosquitoes. Mozzies was
awarded the title of best mobile game in 2003.
Sinem Guven presents a mobile AR authoring system for creat-
ing and editing 3D hypermedia narratives that are interwoven with
a wearable computer user’s surrounding environment15 [21] (see Fig.
9(g)). Their system was designed for authors who are not programmers and
used a combination of 3D drag-and-drop for positioning media and a timeline
for synchronization. It allowed authors to preview their results on a desktop
workstation, as well as with a wearable AR or VR system.
15http://graphics.cs.columbia.edu/projects/mars/Authoring.html
16
2004
Mathias M¨ohring et al. present a system for tracking 3D
markers on a mobile phone [43] (see Fig.10(a)). This work
showed a first video see-through augmented reality system on
a consumer cell-phone. It supports the detection and differentiation of dif-
ferent 3D markers, and correct integration of rendered 3D graphics into the
live video stream.
Michael Rohs and Beat Gfeller present Visual Codes, a 2D
marker system for mobile phones [58] (see Fig.10(b)). These
codes can be attached to physical objects in order to retrieve
object-related information and functionality. They are also suitable for dis-
play on electronic screens.
Enylton Machado Coelho et al. presents OSGAR, a scene graph
with uncertain transformations [9] (see Fig.10(c)). In their work
they target the problem of registration error, which is especially im-
portant for mobile scenarios when high quality tracking is not available and
overlay graphics will not align perfectly with the real environment. OSGAR
dynamically adapts the display to mitigate the effects of registration errors.
The Invisible Train, is shown at SIGGRAPH 2004 Emerging
Technologies16 (see Fig.10(d)). The Invisible Train is the first
multi-user Augmented Reality application for handheld devices
[74].
2005
Anders Henrysson ports ARToolKit to Symbian [22] (see
Fig.11(a)). Based on this technology he presents the famous
AR-Tennis game, the first collaborative AR application run-
ning on a mobile phone. ARTennis was awarded the Indepdent Mobile Gam-
ing best game award for 2005, and the technical achievement award.
Project ULTRA shows how to use non-realtime natural fea-
ture tracking on PDAs to support people in multiple domains
such as the maintenance and support of complex machines,
construction and production, and edutainment and cultural heritage [39].
Furthermore an authoring environment is developed to create the AR scenes
for the maintenance tasks.
16The Invisible Train: http://studierstube.icg.tugraz.at/invisible_train/
17
(a) (b) (c)
Figure 11: (a): AR-Tennis by Henrysson et al. [22]. (b): Going Out by
Reitmayr and Drummond [55]. (c): Mara system by Nokia in 2006.
The first mobile phones equipped with three-axis accelerom-
eters were the Sharp V603SH and the Samsung SCH-S310 both sold
in Asia in 2005.
2006
Reitmayr and Drummond present a model-based hybrid track-
ing system for outdoor augmented reality in urban envi-
ronments enabling accurate, real-time overlays on a handheld
device [55] (see Fig.11(b)). The system combines an edge-based tracker for
accurate localization, gyroscope measurements to deal with fast motions,
measurements of gravity and magnetic field to avoid drift, and a back store
of reference frames with online frame selection to re-initialize automatically
after dynamic occlusions or failures.
Nokia presents Mara, a multi-sensor mobile phone AR
guidance application for mobile phones17. The prototype ap-
plication overlays the continuous viewfinder image stream cap-
tured by the camera with graphics and text in real time, annotating the
user’s surroundings (see Fig.11(c)).
17Mara: http://research.nokia.com/page/219
18
(a) (b) (c)
(d) (e)
Figure 12: (a): PTAM by Klein and Murray [32]. (b): Groundcam by
DiVerdi and H¨ollerer [11]. (c): Map Navigation with mobile devices by Rohs
et al. [59]. (d): Apple iPhone 2G. (e): AR advertising app by HIT Lab NZ
and Saatchi.
2007
Klein and Murray present a system capable of robust real-
time tracking and mapping in parallel with a monocular
camera in small workspaces [32] (see Fig. 12(a)). It is an
adaption of a SLAM approach which processes the tracking and mapping
task on two separated threads.
DiVerdi and H¨ollerer present the GroundCam, a system com-
bining a camera and an orientation tracker [11] (see Fig. 12(b)).
The camera points at the ground behind the user and provides
2D tracking information. The method is similar to that of an optical desktop
mouse.
Rohs et al. compare the performance of the following naviga-
tion methods for map navigation on mobile devices: joystick
navigation, the dynamic peephole method without visual con-
text, and the magic lens paradigm using external visual context [59] (see Fig.
12(c)). In their user study they demonstrate the advantage of dynamic peep-
hole and magic lens interaction over joystick interaction in terms of search
time and degree of exploration of the search space.
19
(a) (b) (c)
Figure 13: (a): Real-time natural feature tracking on mobile phones by
Wagner et al. [75]. (b): Commercial AR museum guide by METAIO [41].
(c): Wikitude AR Browser.
The first multi-touch screen mobile phone, famously known as
iPhone sold by Apple, leverages a new way to interact on mobile
devices (see Fig. 12(d)).
HIT Lab NZ and Saatchi deliver the world’s first mobile phone
based AR advertising application for the Wellington Zoo [78] (see
Fig. 12(e)).
2008
Wagner et al. present the first 6DOF implementation of
natural feature tracking in real-time on mobile phones
achieving interactive frame rates of up to 20 Hz [75] (see Fig.
13(a)). They heavily modify the well known SIFT and Ferns methods in
order to gain more speed and reduce memory requirements.
METAIO presents a commercial mobile AR museum
guide using natural feature tracking or a six-month exhibi-
tion on Islamic art [41] (see Fig. 13(b)). In their paper they
describe the experiences made in this project.
With Augmented Reality 2.0, Schmalstieg et al. presented at the
Dagstuhl seminar in 2008 for the first time a concept that combined
ideas of the Web 2.0 such as social media, crowd sourcing through
public participation, and an open architecture for content markup and dis-
tribution, and applied it to mobile Augmented Reality to create a scalable
AR experience [62].
20
Mobilizy launches Wikitude18, an application that combines
GPS and compass data with Wikipedia entries. The Wikitude
World Browser overlays information on the real-time camera
view of an Android smartphone (see Fig. 13(c)).
2009
Morrison et al. present MapLens which is a mobile augmented
reality (AR) map using a magic lens over a paper map [44]
(see Fig. 14(a)). They conduct a broad user study in form
of an outdoor location-based game. Their main finding is that AR features
facilitate place-making by creating a constant need for referencing to the
physical. The field trials show that the main potential of AR maps lies in
their use as a collaborative tool.
Hagbi et al. presented an approach allowing to track the pose of
the mobile device by pointing it to fiducials [6] (see Fig. 14(b)).
Unlike existing systems the approach allows to track a wide set
of planar shapes while the user can teach the system new shapes at runtime
by showing them to the camera. The learned shapes are then maintained
by the system in a shape library enabling new AR application scenarios in
terms of interaction with the scene but also in terms of fiducial design.
Sean White introduces SiteLens (see Fig. 14(c)), a hand-held
mobile AR system for urban design and urban planning
site visits [77]. SiteLens creates ”situated visualizations” that
are related to and displayed in their environment. For example, represen-
tations of geocoded carbon monoxide concentration data are overlaid at the
sites at which the data was recorded.
SPRXmobile launches Layar19, an advanced variant of Wiki-
tude (see Fig. 14(d)). Layar uses the same registration mecha-
nism as Wikitude (GPS + compass), and incoperates this into
an open client-server platform. Content layers are the equivalent of web
pages in normal browsers. Existing layers include Wikipedia, Twitter and
Brightkite to local services like Yelp, Trulia, store locators, nearby bus stops,
mobile coupons, Mazda dealers and tourist, nature and cultural guides. On
August 17th Layar went global serving almost 100 content layers.
18Wikitude: http://www.mobilizy.com/wikitude.php?lang=en
19LayAR: http://layar.eu/
21
(a) (b)
(c) (d)
(e) (f)
Figure 14: (a): MapLens by Morrison et al . [44]. (b) Hagbi’s pose tracking
using shape [6]. (c): SiteLens by White and Feiner [77]. (d): LayAR AR
browser. (e): ARhrrrr! Zombie game by Spreen et al. from Georgia Tech.
(f): Klein’s PTAM system running on an iPhone [33].
22
Kimberly Spreen et al. develop ARhrrrr!, the first mobile AR
game with high quality content at the level of commercial
games[79] (see Fig. 14(e)). They use an NVIDIA Tegra devel-
oper kit (”Concorde”) with a fast GPU. All processing except for tracking
are running on the GPU, making the whole application run at high frame
rates on a mobile phone class device despite the highly detailed content and
natural feature tracking.
Georg Klein presents a video showing his SLAM system run-
ning in real-time on an iPhone [33] (see Fig. 14(f)) and
later presents this at ISMAR 2009 in Orlando, Florida. Even
though it has constrains in terms of working area it is the first time a 6DoF
SLAM system is known to run on mobile phones in sufficient speed.
Update April 2015: The following parts of the document until beginning
of 2015 cover the years since the last homepage update, following the same
categorization and scheme as before.
From end of 2009 onwards, AR research and development is generally driven
by high expectations and huge investments from world-leading companies
such as Microsoft, Google, Facebook, Qualcomm and others. At the same
time, the landscape of mobile phone manufacturers started to change radi-
cally.
In general the advances in mobile device capabilities introduce a strong drive
towards mobile computing, and the availability of cloud processing further
supports the proposal and development of server-client solutions for AR pur-
poses. One major trend starting around 2010, originating by the work of
Davison in 2003 [10] and later further explored by Klein and Murray [32,33],
is the heavy use of SLAM in AR, which still continues to dominate a major
part of AR research and development as of beginning of 2015.
Microsoft presents ”Project Natal” at the game exhibition E3. It
is the first version of a new hardware interface, consisting of
motion detection technology, microphone, color camera and software,
to be integrated into the game console Xbox 360.
At ISMAR 2009, Clemens Arth et al. present a system for
large-scale localization and subsequent 6DOF tracking
on mobile phones [2]. The system uses sparse point clouds of
city areas and FAST corners and SURF-like descriptors that can be used on
memory-limited devices (see Fig. 15(a)).
23
Qualcomm Inc. acquires the mobile AR IP from Imagination
Computer Services GmbH., Vienna, and takes over the funding of
the Christian Doppler Laboratory for Handheld AR at Graz Univer-
sity of Technology. A research center to focus on AR is opened later in 2010
in Vienna [80].
2010
Areal-time panoramic mapping and tracking system
for mobile phones is presented by Wagner et al . at VR, which
performs 3DOF tracking in cylindric space and supports the use
of panoramic imagery for improved usability and experience in AR [73] (see
Fig. 15(b)).
KHARMA is a lightweight and open architecture for referencing
and delivering content explicitly aiming for mobile AR applica-
tions running on a global scale. It uses KML for describing the
geospatial or relative relation of content while utilizing on HTML, JavaScript
and CSS technologies for content development and delivery [24].
Microsoft announces a close cooperation with Primesense [81],
an Israeli company working on structured-light based 3D sen-
sors, to supply their technology to ”Project Natal”, now coined Kinect. The
Kinect becomes commercially available in November 2010.
Apple releases the iPad20 on April 2010, which becomes the first
tablet computer to be adopted by the large public. The iPad
featured an assisted GPS, accelerometers, magnetometers, advanced graphics
chipset (PowerVR SGX535), enabling the possibilities to create efficient AR
application on tablet computer (see Fig. 15(c)).
At ISMAR Lukas Gruber et al . present the ”City of Sights”, a col-
lection of datasets and paperboard models21 to evaluate the tracking
and reconstruction performance of algorithms used in AR [19] (see
Fig. 15(d)).
After several delays, Microsoft releases Windows Phone in Oc-
tober 2010, to become the third major mobile phone operating sys-
tem to challenge iOS and Android.
20Wikipedia: http://en.wikipedia.org/wiki/IPad
21http://studierstube.icg.tugraz.at/handheld_ar/cityofsights.php
24
(a)
(b)
(c) (d) (e)
Figure 15: (a): City reconstruction as used by Arth et al . [2]. (b): Panoramic
image captured on a mobile phone using the approach of Wagner et al. [73].
(c) Apple iPad. (d): City-of-Sights paperboard models by Gruber et al . [19].
(e) In-situ information creation by Langlotz et al. [36].
25
(a) (b)
Figure 16: (a): KinectFusion system presented by Newcombe et al. at
ISMAR 2011 [46]. (b): Mobile phone scene reconstruction by Pan et al. [49].
Existing mobile AR applications where exclusively used to
browser and consume digital information. Langlotz et al. pre-
sented an new approach aiming for AR browsers that also sup-
ported creation of digital information in-situ. The information is registered
with pixel-precision by utilizing a panorama of the environment that is cre-
ated in the background [36] (see Fig. 15(e)).
2011
Qualcomm announces the release of its AR platform SDK
to the public in April. At that time it is called QCAR [82],
which will later be called Vuforia.
In August, Google announces the acquisition of Motorola
Mobility for about $12.5 million [83]. A major asset of Motorola
is a large patent portfolio, which Google needs to secure the further
Android platform development.
At ICCV 2011, Newcombe presents DTAM, a dense real-time
tracking and mapping algorithm [47]. Later at ISMAR
2011, Richard Newcombe presents the KinectFusion work
[46], in which depth images from the Kinect sensor are fused to create a single
implicit surface model. KinectFusion becomes publicly available within the
Kinect SDK later [84] (see Fig. 16(a)).
Qi Pan presents his work on reconstructing scenes on mo-
bile phones using panoramic images. By using FAST corners
and a SURF-like descriptor, multiple panoramas are registered
26
and a triangulated model is created after voxel carving [49] (see Fig. 16(b)).
Following the still challenging problem of running SLAM in
real-time on mobiles, Pirchheim presents an approach using
planarity assumptions, and demonstrates his approach on a
Nokia N900 smartphone [50].
Grubert et al . publish a technical report about the plausibility
of using AR browsers [20], which becomes a survey about the
pros and cons of AR browser technology at that point in time.
2012
Smart watches are broadly introduced as a new generation of mo-
bile wearables. Pebble and the Sony SmartWatch are built to con-
nect to a personal smartphone and to provide simple functionality, such as
notifications or call answering.
Google Glass (also known as Google Project Glass) is firstly pre-
sented to the public22 (see Fig.17(b)). Goggle Glass is is an optical
HMD that can be controlled with an integrated touch-sensitive sensor or nat-
ural language commands. After it’s public announcement Google Glass had
a major impact on research but even more on the public perception of mixed
reality technology.
NVidia is demonstrating at Siggraph Emerging Technologies their
prototype of a head mounted display supporting accurate accommo-
dation, convergence, and binocular-disparity depth cues (see Fig. 17(c)).
The prototype introduces a light-field-based approach to near-eye displays
and can be seen as a next generation wearable display technology for AR as
existing hardware can’t provide accurate acommodation [85].
13th lab released the first commercial mobile SLAM (Simultaneous
localization and mapping) system coined Pointcloud23 to the public,
marking a major milestone for app developers who want to integrate SLAM-
based tracking into their application24.
PrimeSense, the creator of the Microsoft Kinect, introduced a smaller
version of a 3D sensing device called Capri [86] that is small enough
22Google Glass project page on Google+: https://plus.google.com/+GoogleGlass
23Pointcloud homepage: http://pointcloud.io/
24Pointcloud video: http://www.youtube.com/watch?v=K5OKaK3Ay8U
27
(a) (b) (c)
Figure 17: (a): Oculus Rift developer edition. (b): Google Glass. (c):
Near-eye light field project by NVidia.
to be integrated into mobile devices such as tablets or smartphones25.
At ISMAR 2012, Steffen Gauglitz et al . present their ap-
proach on tracking and mapping from both general and
rotation-only camera motion [18].
In August, Oculus VR announces the Oculus Rift dev kit, a virtual
reality head-mounted display. This initiated a new hype in Virtual
Reality and in the development of more head-mounted displays for gaming
purposes mainly (see Fig.17(a)).
2013
As opposed to previous work from Gauglitz et al., Pirchheim
et al. present an approach to handle pure camera rotation
running on a mobile phone at ISMAR [51].
Google Glass, which was already announced as Project Glass in
2012, becomes available through the explorer program in late 2013.
and raises positive and negative attention, as well as concerns about privacy
and ethical aspects (see Fig.17(b)).
At ICRA, Li et al. present an amazing approach for mo-
tion tracking with inertial sensors and a rolling-shutter
camera running in real-time on a mobile phone [37].
Tan et al. propose an approach to SLAM working in dynamic
environments, allowing parts in the scene to be dynamic with-
out breaking the mapping and tracking [67].
25Capri Video: http://www.youtube.com/watch?v=ELTETXO02zE
28
(a) (b)
Figure 18: (a): SLAM map localization by Ventura et al. [71]. (b): LSD-
SLAM reconstruction by Engel et al. [12].
On November 24, 2013, Apple Inc. confirms the purchase of
PrimeSense for about $350 million [87]. Primesense was working
on shrinking their sensors to fit into mobiles at that point in time.
Taskanen et al. propose an approach to perform full 3D re-
construction on a mobile monocular smartphone and
creating a dense 3D model with known absolute scale [68].
2014
Three years after the acquisition, in January Google sells Mo-
torola Mobility to Lenovo for $2.91 million, however, keeping most
of the patent portfolio [88].
Also in January, Qualcomm acquires Kooaba [89], a Swiss ETH-
spin-off founded in 2007, built around image recognition using SURF
features. Kooaba’s technology is integrated into the services pro-
vided through the Vuforia SDK.
In February, Google announces Project Tango [90], which is an
Android smartphone equipped with a full Kinect-like 3D sensor and
hands out a few hundred units to developers and companies.
In March, Facebook acquires Oculus VR for $2 billion, although
Oculus does not make any consumer products at that point in time
yet [91]. This strengthens the hype in upcoming VR interfaces.
At VR, Ventura et al. present an approach to localize SLAM
maps built on a mobile phone accurately wrt. a sparse 3D
reconstruction of urban environments [71] (see Fig.18(a)).
29
In April, Microsoft announces the acquisition of Nokia’s De-
vices and Services unit for $7.2 billion [92], as Nokia is the primary
vendor for Windows devices devices, especially the Lumia phones.
Following up on previous work at ICCV 2013[13], at
ECCV Engel et al. present LSD-SLAM, a feature-less
monocular SLAM algorithm using keyframes and semi-
dense depth maps, and release the code to the public [12] (see Fig.18(b)).
At ISMAR, a mobile version is presented as well [63].
At 3DV, Herrera et al. present DT-SLAM [23]. The key idea
behind the approach is to defer the triangulation step of 2D
features matched across keyframes until those have undergone
a certain baseline, improving the overall robustness of SLAM.
At ISMAR, Salas-Moreno et al. present Dense Planar
SLAM, leveraging the assumption that many man-made
surfaces are planar [60].
2015
In January, Microsoft announces the Hololens, a headset to fuse AR
and VR [93] to be made available later in 2015. The device is a
complete computer with a see-through display and several sensors.
Acknowledgements
Thanks go to the ISMAR09 mobile committee and all others for their valuable
suggestions.
References
[1] Radio Telephone System, U.S. Patent Application U.S. 3,906,166,
September 16, 1975. 3
[2] C. Arth, D. Wagner, M. Klopschitz, A. Irschara, and D. Schmalstieg,
“Wide area localization on mobile phones,” in IEEE International Sym-
posium on Mixed and Augmented Reality (ISMAR), pp. 73–82, IEEE,
2009. 23,25
[3] R. T. Azuma, “A survey of augmented reality,” Presence: Teleoperators
and Virtual Environments, vol. 6, pp. 355–385, August 1997. 6
30
[4] B. B. Bederson, “Audio augmented reality: A prototype automated
tour guide,” in Conference Companion on Human Factors in Computing
Systems, CHI ’95, (New York, NY, USA), pp. 210–211, ACM, 1995. 6
[5] B. Bell, S. Feiner, and T. H¨ollerer, “View management for virtual and
augmented reality,” in Proceedings of the 14th Annual ACM Symposium
on User Interface Software and Technology, UIST ’01, (New York, NY,
USA), pp. 101–110, ACM, 2001. 12,13
[6] O. Bergig, N. Hagbi, J. El-Sana, and M. Billinghurst, “In-place 3d
sketching for authoring and augmenting mechanical systems,” in IEEE
International Symposium on Mixed and Augmented Reality (ISMAR),
pp. 87–94, 2009. 21,22
[7] T. Caudell and D. Mizell, “Augmented reality: an application of heads-
up display technology to manual manufacturing processes,” in System
Sciences, 1992. Proceedings of the Twenty-Fifth Hawaii International
Conference on, vol. ii, pp. 659–669 vol.2, Jan 1992. 2,3
[8] A. D. Cheok, S. W. Fong, K. H. Goh, X. Yang, W. Liu, and F. Farzbiz,
“Human pacman: A sensing-based mobile entertainment system with
ubiquitous computing and tangible interaction,” in Proceedings of the
2Nd Workshop on Network and System Support for Games, NetGames
’03, (New York, NY, USA), pp. 106–117, ACM, 2003. 14,15
[9] E. Coelho, S. Julier, and B. Maclntyre, “Osgar: a scene graph with
uncertain transformations,” in IEEE International Symposium on Mixed
and Augmented Reality (ISMAR), pp. 6–15, Nov 2004. 16,17
[10] A. J. Davison, “Real-time simultaneous localisation and mapping with
a single camera,” in 9th IEEE International Conference on Computer
Vision (ICCV 2003), 14-17 October 2003, Nice, France, pp. 1403–1410,
2003. 23
[11] S. DiVerdi and T. Hollerer, “Groundcam: A tracking modality for mobile
mixed reality,” in IEEE Virtual Reality Conference (VR), pp. 75–82,
March 2007. 19
[12] J. Engel, T. Sch¨ops, and D. Cremers, “LSD-SLAM: Large-scale di-
rect monocular SLAM,” in European Conference on Computer Vision
(ECCV), September 2014. 29,30
31
[13] J. Engel, J. Sturm, and D. Cremers, “Semi-dense visual odometry for
a monocular camera,” in IEEE International Conference on Computer
Vision (ICCV), (Sydney, Australia), December 2013. 30
[14] S. Feiner, B. MacIntyre, T. H¨ollerer, and A. Webster, “A touring ma-
chine: prototyping 3d mobile augmented reality systems for exploring
the urban environment,” in IEEE International Symposium on Wearable
Computers, pp. 74–81, Oct 1997. 5,7
[15] G. W. Fitzmaurice, “Situated information spaces and spatially aware
palmtop computers,” Commun. ACM, vol. 36, pp. 39–49, July 1993. 4
[16] D. Fritsch, D. Klinec, and S. Volz, “Nexus positioning and data man-
agement concepts for location-aware applications,” Computers, Envi-
ronment and Urban Systems, vol. 25, no. 3, pp. 279 – 291, 2001. 10
[17] J. Fruend, C. Geiger, M. Grafe, and B. Kleinjohann, “The augmented
reality personal digital assistant,” in IEEE and ACM International Sym-
posium on Augmented Reality (ISAR), 2001. 11,12
[18] S. Gauglitz, C. Sweeney, J. Ventura, M. Turk, and T. Hollerer, “Live
tracking and mapping from both general and rotation-only camera mo-
tion,” in IEEE International Symposium on Mixed and Augmented Re-
ality (ISMAR), pp. 13–22, Nov 2012. 28
[19] L. Gruber, S. Gauglitz, J. Ventura, S. Zollmann, M. Huber, M. Schlegel,
G. Klinker, D. Schmalstieg, and T. H¨ollerer, “The city of sights: Design,
construction, and measurement of an augmented reality stage set,” in
IEEE International Symposium on Mixed and Augmented Reality (IS-
MAR), pp. 157–163, 2010. 24,25
[20] J. Grubert, T. Langlotz, and R. Grasset, “Augmented reality browser
survey,” Tech. Rep. 1101, Institute for Computer Graphics and Vision,
Graz University of Technology, Graz, Austria, 2011. 27
[21] S. Guven and S. Feiner, “Authoring 3d hypermedia for wearable aug-
mented and virtual reality,” in IEEE International Symposium on Wear-
able Computers, pp. 118–126, Oct 2003. 14,16
[22] A. Henrysson, M. Billinghurst, and M. Ollila, “Face to face collaborative
ar on mobile phones,” in IEEE International Symposium on Mixed and
Augmented Reality (ISMAR), pp. 80–89, Oct 2005. 17,18
32
[23] D. Herrera C, K. Kim, J. Kannala, K. Pulli, and J. Heikkila, “Dt-slam:
Deferred triangulation for robust slam,” in 3D Vision (3DV), 2014 2nd
International Conference on, vol. 1, pp. 609–616, Dec 2014. 30
[24] A. Hill, B. MacIntyre, M. Gandy, B. Davidson, and H. Rouzati,
“Kharma: An open kml/html architecture for mobile augmented real-
ity applications,” in Mixed and Augmented Reality (ISMAR), 2010 9th
IEEE International Symposium on, pp. 233–234, Oct 2010. 24
[25] T. H¨ollerer, S. Feiner, and J. Pavlik, “Situated documentaries: embed-
ding multimedia presentations in the real world,” in IEEE International
Symposium on Wearable Computers, pp. 79–86, Oct 1999. 8,9
[26] T. H¨ollerer, S. Feiner, T. Terauchi, G. Rashid, and D. Hallaway, “Ex-
ploring MARS: developing indoor and outdoor user interfaces to a mo-
bile augmented reality system,” Computers & Graphics, vol. 23, no. 6,
pp. 779–785, 1999. 8,9
[27] S. Julier, Y. Baillot, D. Brown, and M. Lanzagorta, “Information filter-
ing for mobile augmented reality,” Computer Graphics and Applications,
IEEE, vol. 22, pp. 12–15, Sep 2002. 11
[28] S. Julier, Y. Baillot, M. Lanzagorta, D. Brown, and L. Rosenblum,
“Bars: Battlefield augmented reality system,” in In NATO Symposium
on Information Processing Techniques for Military Systems, pp. 9–11,
2000. 10
[29] M. Kalkusch, T. Lidy, M. Knapp, G. Reitmayr, H. Kaufmann, and
D. Schmalstieg, “Structured visual markers for indoor pathfinding,” in
Augmented Reality Toolkit, The First IEEE International Workshop,
pp. 8 pp.–, 2002. 13,14
[30] H. Kato and M. Billinghurst, “Marker tracking and hmd calibration for a
video-based augmented reality conferencing system,” in Augmented Re-
ality, 1999. (IWAR ’99) Proceedings. 2nd IEEE and ACM International
Workshop on, pp. 85–94, 1999. 7,8
[31] A. C. Kay, “A personal computer for children of all ages,” August 1972.
2
[32] G. Klein and D. Murray, “Parallel tracking and mapping for small ar
workspaces,” in IEEE International Symposium on Mixed and Aug-
mented Reality (ISMAR), (Washington, DC, USA), pp. 1–10, IEEE
Computer Society, 2007. 19,23
33
[33] G. Klein and D. Murray, “Parallel tracking and mapping on a camera
phone,” in IEEE International Symposium on Mixed and Augmented
Reality (ISMAR), (Orlando), IEEE Computer Society, October 2009.
22,23
[34] R. Kooper and B. MacIntyre, “Browsing the real-world wide web: Main-
taining awareness of virtual information in an ar information space.,”
Int. J. Hum. Comput. Interaction, vol. 16, no. 3, pp. 425–446, 2003. 12,
13
[35] U. Kretschmer, V. Coors, U. Spierling, D. Grasbon, K. Schneider, I. Ro-
jas, and R. Malaka, “Meeting the spirit of history,” in Proceedings of the
2001 Conference on Virtual Reality, Archeology, and Cultural Heritage,
VAST ’01, (New York, NY, USA), pp. 141–152, ACM, 2001. 13
[36] T. Langlotz, D. Wagner, A. Mulloni, and D. Schmalstieg, “Online cre-
ation of panoramic augmented reality annotations on mobile phones,”
Pervasive Computing, IEEE, vol. 11, pp. 56–63, Feb 2012. 25,26
[37] M. Li, B. H. Kim, and A. Mourikis, “Real-time motion tracking on
a cellphone using inertial sensing and a rolling-shutter camera,” in
Robotics and Automation (ICRA), 2013 IEEE International Conference
on, pp. 4712–4719, May 2013. 28
[38] J. M. Loomis, R. G. Golledge, and R. L. Klatzky, “Personal guidance
system for the visually impaired using gps, gis, and vr technologies,”
in Proceedings of the First Annual International Conference, Virtual
Reality and Persons with Disabilities, pp. 71–74, June 1993. 4
[39] A. Makri, D. Arsenijevic, J. Weidenhausen, P. Eschler, D. Stricker,
O. Machui, C. Fernandes, S. Maria, G. Voss, and N. Ioannidis, “Ul-
tra: An augmented reality system for handheld platforms, targeting in-
dustrial maintenance applications,” in Proceedings of 11th International
Conference on Virtual Systems and Multimedia (VSMM’05), 2005. 17
[40] P. Milgram and F. Kishino, “Taxonomy of mixed reality visual displays,”
IEICE Transactions on Information and Systems, vol. E77-D, no. 12,
pp. 1321 – 1329, 1994. 5
[41] T. Miyashita, P. Meier, T. Tachikawa, S. Orlic, T. Eble, V. Scholz,
A. Gapel, O. Gerl, S. Arnaudov, and S. Lieberknecht, “An augmented
reality museum guide,” in IEEE International Symposium on Mixed and
Augmented Reality (ISMAR), pp. 103–106, Sept 2008. 20
34
[42] D. Mogilev, K. Kiyokawa, and M. Billinghurst, “Ar pad: An interface for
face-to-face ar collaboration,” in In CHI 2002 Conference Proceedings,
pp. 654–655, 2002. 14,15
[43] M. M¨ohring, C. Lessig, and O. Bimber, “Video see-through ar on con-
sumer cell-phones,” in IEEE International Symposium on Mixed and
Augmented Reality (ISMAR), pp. 252–253, Nov 2004. 16,17
[44] A. Morrison, A. Oulasvirta, P. Peltonen, S. Lemmela, G. Jacucci, G. Re-
itmayr, J. N¨as¨anen, and A. Juustila, “Like bees around the hive: A
comparative study of a mobile augmented reality map,” in Proceedings
of the SIGCHI Conference on Human Factors in Computing Systems,
CHI ’09, (New York, NY, USA), pp. 1889–1898, ACM, 2009. 21,22
[45] L. Naimark and E. Foxlin, “Circular data matrix fiducial system and
robust image processing for a wearable vision-inertial self-tracker,” in
IEEE International Symposium on Mixed and Augmented Reality (IS-
MAR), pp. 27–36, 2002. 15
[46] R. A. Newcombe, S. Izadi, O. Hilliges, D. Molyneaux, D. Kim, A. J.
Davison, P. Kohli, J. Shotton, S. Hodges, and A. W. Fitzgibbon,
“Kinectfusion: Real-time dense surface mapping and tracking,” in IEEE
International Symposium on Mixed and Augmented Reality (ISMAR),
pp. 127–136, 2011. 26
[47] R. A. Newcombe, S. Lovegrove, and A. J. Davison, “DTAM: dense track-
ing and mapping in real-time,” in IEEE International Conference on
Computer Vision (ICCV), pp. 2320–2327, 2011. 26
[48] J. Newman, D. Ingram, and A. Hopper, “Augmented reality in a wide
area sentient environment,” in IEEE and ACM International Symposium
on Augmented Reality (ISAR), pp. 77–86, 2001. 11,12
[49] Q. Pan, C. Arth, E. Rosten, G. Reitmayr, and T. Drummond, “Rapid
scene reconstruction on mobile phones from panoramic images,” in IEEE
International Symposium on Mixed and Augmented Reality (ISMAR),
pp. 55–64, 2011. 26,27
[50] C. Pirchheim and G. Reitmayr, “Homography-based planar mapping
and tracking for mobile phones,” in IEEE International Symposium on
Mixed and Augmented Reality (ISMAR), pp. 27–36, 2011. 27
35
[51] C. Pirchheim, D. Schmalstieg, and G. Reitmayr, “Handling pure camera
rotation in keyframe-based SLAM,” in IEEE International Symposium
on Mixed and Augmented Reality (ISMAR), pp. 229–238, 2013. 28
[52] R. Raskar, J. van Baar, P. Beardsley, T. Willwacher, S. Rao, and C. For-
lines, “ilamps: Geometrically aware and self-configuring projectors,”
in ACM SIGGRAPH 2003 Papers, SIGGRAPH ’03, (New York, NY,
USA), pp. 809–818, ACM, 2003. 14,15
[53] H. Regenbrecht and R. Specht, “A mobile passive augmented reality
device - mpard,” in IEEE and ACM International Symposium on Aug-
mented Reality (ISAR), pp. 81–84, 2000. 10
[54] G. Reitmayr and D. Schmalstieg, “Mobile collaborative augmented real-
ity,” in IEEE and ACM International Symposium on Augmented Reality
(ISAR), pp. 114–123, 2001. 11,12
[55] G. Reitmayr and T. Drummond, “Going out: robust model-based track-
ing for outdoor augmented reality,” in IEEE International Symposium
on Mixed and Augmented Reality (ISMAR), pp. 109–118, Oct 2006. 18
[56] J. Rekimoto, “Augmented reality using the 2d matrix code,” in Proceed-
ings of the Workshop on Interactive Systems and Software (WISS’96),
1996. 5,6
[57] J. Rekimoto and K. Nagao, “The world through the computer: Com-
puter augmented interaction with real world environments,” in Proceed-
ings of the 8th Annual ACM Symposium on User Interface and Software
Technology, UIST ’95, (New York, NY, USA), pp. 29–36, ACM, 1995.
5,6
[58] M. Rohs and B. Gfeller, “Using camera-equipped mobile phones for in-
teracting with real-world objects,” in Advances in Pervasive Computing,
pp. 265–271, 2004. 16,17
[59] M. Rohs, J. Sch¨oning, M. Raubal, G. Essl, and A. Kr¨uger, “Map navi-
gation with mobile devices: Virtual versus physical movement with and
without visual context,” in Proceedings of the 9th International Con-
ference on Multimodal Interfaces, ICMI ’07, (New York, NY, USA),
pp. 146–153, ACM, 2007. 19
[60] R. Salas-Moreno, B. Glocken, P. Kelly, and A. Davison, “Dense pla-
nar slam,” in IEEE International Symposium on Mixed and Augmented
Reality (ISMAR), pp. 157–164, Sept 2014. 30
36
[61] K. Satoh, M. Anabuki, H. Yamamoto, and H. Tamura, “A hybrid reg-
istration method for outdoor augmented reality,” in IEEE and ACM
International Symposium on Augmented Reality (ISAR), vol. 29-30,
pp. 67–76, Oct. 2001. 11,12
[62] D. Schmalstieg, T. Langlotz, and M. Billinghurst, “Augmented reality
2.0,” in Virtual Realities (G. Brunnett, S. Coquillart, and G. Welch,
eds.), pp. 13–37, Springer Vienna, 2011. 20
[63] T. Sch¨ops, J. Engel, and D. Cremers, “Semi-dense visual odometry for
AR on a smartphone,” in IEEE International Symposium on Mixed and
Augmented Reality (ISMAR), September 2014. 30
[64] J. Spohrer, “Information in places,” IBM Systems Journal, vol. 38, no. 4,
pp. 602–628, 1999. 8
[65] T. Starner, S. Mann, B. J. Rhodes, J. Levine, J. Healey, D. Kirsch,
R. W. Picard, and A. Pentland, “Augmented reality through wearable
computing,” Presence, vol. 6, no. 4, pp. 386–398, 1997. 7
[66] I. E. Sutherland, “A head-mounted three dimensional display,” in Pro-
ceedings of the December 9-11, 1968, Fall Joint Computer Conference,
Part I, AFIPS ’68 (Fall, part I), (New York, NY, USA), pp. 757–764,
ACM, 1968. 2
[67] W. Tan, H. Liu, Z. Dong, G. Zhang, and H. Bao, “Robust monocular
slam in dynamic environments,” in IEEE International Symposium on
Mixed and Augmented Reality (ISMAR), pp. 209–218, Oct 2013. 28
[68] P. Tanskanen, K. Kolev, L. Meier, F. Camposeco, O. Saurer, and
M. Pollefeys, “Live metric 3d reconstruction on mobile phones,” in IEEE
International Conference on Computer Vision (ICCV), pp. 65–72, Dec
2013. 29
[69] B. Thomas, B. Close, J. Donoghue, J. Squires, P. De Bondi, M. Mor-
ris, and W. Piekarski, “Arquake: an outdoor/indoor augmented reality
first person application,” in IEEE International Symposium on Wearable
Computers, pp. 139–146, Oct 2000. 9,10
[70] B. Thomas, V. Demczuk, W. Piekarski, D. Hepworth, and B. Gun-
ther, “A wearable computer system with augmented reality to support
terrestrial navigation,” in IEEE International Symposium on Wearable
Computers, pp. 168–171, Oct 1998. 7,8
37
[71] J. Ventura, C. Arth, G. Reitmayr, and D. Schmalstieg, “Global localiza-
tion from monocular slam on a mobile phone,” in IEEE Virtual Reality,
03/2014 2014. 29
[72] V. Vlahakis, J. Karigiannis, M. Tsotros, M. Gounaris, L. Almeida,
D. Stricker, T. Gleue, I. T. Christou, R. Carlucci, and N. Ioannidis,
“Archeoguide: First results of an augmented reality, mobile computing
system in cultural heritage sites,” in Proceedings of the 2001 Conference
on Virtual Reality, Archeology, and Cultural Heritage, VAST ’01, (New
York, NY, USA), pp. 131–140, ACM, 2001. 12,13
[73] D. Wagner, A. Mulloni, T. Langlotz, and D. Schmalstieg, “Real-time
panoramic mapping and tracking on mobile phones,” in IEEE Virtual
Reality Conference (VR), pp. 211–218, March 2010. 24,25
[74] D. Wagner, T. Pintaric, F. Ledermann, and D. Schmalstieg, “Towards
massively multi-user augmented reality on handheld devices,” in Per-
vasive Computing (H.-W. Gellersen, R. Want, and A. Schmidt, eds.),
vol. 3468 of Lecture Notes in Computer Science, pp. 208–219, Springer
Berlin Heidelberg, 2005. 16,17
[75] D. Wagner, G. Reitmayr, A. Mulloni, T. Drummond, and D. Schmal-
stieg, “Pose tracking from natural features on mobile phones,” in IEEE
International Symposium on Mixed and Augmented Reality (ISMAR),
pp. 125–134, Sept 2008. 20
[76] D. Wagner and D. Schmalstieg, “First steps towards handheld aug-
mented reality,” in IEEE International Symposium on Wearable Com-
puters, pp. 127–135, Oct 2003. 14,15
[77] S. White and S. Feiner, “Sitelens: Situated visualization techniques for
urban site visits,” in Proceedings of the SIGCHI Conference on Hu-
man Factors in Computing Systems, CHI ’09, (New York, NY, USA),
pp. 1117–1120, ACM, 2009. 21,22
Weblinks
[78] Inspiration Room Article: http://theinspirationroom.com/daily/
2007/augmented-reality-at-wellington-zoo/.
[79] ARhrrrr! Mobile Game: http://www.augmentedenvironments.org/
lab/research/handheld-ar/arhrrrr/.
38
[80] Press Release: https://www.qualcomm.com/news/releases/2010/03/
23/qualcomm-opens-austria-research-center-focus-augmented-reality.
[81] Press Release: http://news.microsoft.com/2010/03/31/
primesense-supplies-3-d-sensing-technology-to-project-natal-for-xbox-360/.
[82] CNXSoft article: http://www.cnx-software.com/2011/04/28/
qualcomm-officially-release-its-augmented-reality-sdk/.
[83] Tech Crunch: http://techcrunch.com/2011/08/15/
breaking-google-buys-motorola-for-12-5-billion/.
[84] Kinect Fusion project page: https://msdn.microsoft.com/en-us/
library/dn188670.aspx.
[85] Near-eye light field project page: https://research.nvidia.com/
publication/near-eye-light-field-displays.
[86] Capri announcement: http://www.primesense.com/news/
primesense-unveils-capri/.
[87] Tech Crunch: http://techcrunch.com/2013/11/24/
apple-primesense-acquisition-confirmed/.
[88] The Verge: http://www.theverge.com/2014/1/29/5358620/
lenovo-reportedly-buying-motorola-mobility-from-google.
[89] Mobile Marketing: http://mobilemarketingmagazine.com/
qualcomm-acquires-kooaba-visual-recognition-company/.
[90] The Verge: http://www.theverge.com/2014/2/20/5430784/
project-tango-google-prototype-smartphone-announced.
[91] Business Insider: http://www.businessinsider.com/
facebook-to-buy-oculus-rift-for-2-billion-2014-3?IR=T.
[92] Tech Crunch: http://techcrunch.com/2014/04/25/
microsofts-7-2bn-acquisition-of-nokias-devices-business-is-now-complete/.
[93] The Verge: http://www.theverge.com/2015/1/21/7867593/
microsoft-announces-windows-holographic.
39
... Among the different technologies that are open source to support training processes is augmented reality (AR). This technology (data) finds its beginnings in the mid-20th century (Grubert et al., 2017), although that name was not coined until 1992, when researchers Tom Caudell and David Mizell (Arth et al., 2015) proposed it to Boeing as a system to help their customers used to assemble complex wire harnesses; They defined it as the superposition of virtual materials made by computer and images of the real world (Arth et al., 2015). In this sense, AR can be defined as a technique that links digital information with the real world by placing digital content such as text, images, audio and video generated by computer in the image of the real world that individuals see around them (Köse & Güner-Yildiz, 2021). ...
... Among the different technologies that are open source to support training processes is augmented reality (AR). This technology (data) finds its beginnings in the mid-20th century (Grubert et al., 2017), although that name was not coined until 1992, when researchers Tom Caudell and David Mizell (Arth et al., 2015) proposed it to Boeing as a system to help their customers used to assemble complex wire harnesses; They defined it as the superposition of virtual materials made by computer and images of the real world (Arth et al., 2015). In this sense, AR can be defined as a technique that links digital information with the real world by placing digital content such as text, images, audio and video generated by computer in the image of the real world that individuals see around them (Köse & Güner-Yildiz, 2021). ...
Article
Full-text available
The attitude of the teaching staff is positioned as a fundamental aspect for the development of good training practices. These good practices are essential when applied within an innovative techno-pedagogical methodology: augmented reality in education. The objectives of this study are to analyze the development of good teaching practices with augmented reality and to discover the factors that influence their quality. A descriptive and correlational design has been carried out. A total of 1490 Spanish Secondary Education teachers have participated. The instrument used was the adaptation to the Spanish context of the questionnaire of the Attitude Scale of Augmented Reality Applications. The results reveal that teachers show a positive attitude towards the use of augmented reality. As for the aspects that influence the good attitude of teachers are age, the number of devices teachers use, the time they dedicate to technological resources and teaching experience. However, ICT training is what determines a direct influence on the attitude of teachers, as well as satisfaction with reliability.
... The advancement of mobile and wearable devices has opened numerous possibilities for creative application of extended reality in built environments (Arth et al., 2015; Lopez Rodriguez and Pantic, 2023; Geopel et al., 2023). This research study delves into the convergence of user interaction design, gamification, and augmented reality technologies within the context of an exhibition design project. ...
Conference Paper
The research discussed in this paper centers around the convergence of extended reality (XR) platforms, computational design, digital fabrication, and critical urban study practices. Its aim is to cultivate interdisciplinary and multi- scalar approaches within these domains. The research endeavor represents a collaborative effort between two primary disciplines: critical urban studies, which prioritize socio-environmental justice, and integrated digital design- to-production, which emphasize the realization of volumetric or voxel-based structural systems. Moreover, the exploration encompasses augmented reality to assess its utilization in both the assembly process of the structures and the integration of phygital (physical and digital) data with the physical environment. Within the context of these research scopes, this paper introduces FabriCity-XR as an interactive phygital installation. In addition to presenting an overview of the integrated research driven and performative design to production methodologies, the project showcases the practical implementation of web-based augmented reality trails, eliminating the requirement for external applications for interaction. This approach allows users to seamlessly navigate and engage with phygital content overlaid on physical objects using their personal smart devices. The result is a captivating and immersive user experience that effectively merges the physical and digital realms.
... By the year 2000, the first AR Game was launched with the name 'AR Quake' that used a head-mounted display and an attached computer as a backpack. The year 2005 was the debut year of AR apps to be used in mobile phones (Clemens A., et al., 2015). The first AR-related app on mobile was called 'AR Tennis' which was played on a Nokia phone by two people. ...
Chapter
Education 4.0 is an evolving paradigm that leverages technology to revolutionize teaching and learning methods, by giving augmented reality (AR) and virtual reality (VR) technologies that provide a more engaging, personalized, and immersive educational experience. This book chapter explores the integration of AR and VR in the transformation of teaching and learning pedagogy by bridging the gap between the physical and digital worlds within the framework of Education 4.0. Integrating AR and VR has led the traditional pedagogical approaches be augmented with interactive and dynamic elements, facilitating deeper knowledge retention and application. These technologies are particularly beneficial in STEM fields. It has proven successful in the early childhood, language education, and special needs education also. The chapter also explains a case study on how AR and VR can be used to promote learning in the subject of chemistry. AR and VR promise to redefine the future of education, preparing learners to thrive in an ever-changing global landscape.
... Research on AR has grown exponentially since the introduction of the smartphone. Competing tech companies heavily invested in improving mobile technology and the capacity of processing power as each new generation decreases in cost (Arth et al., 2015), resulting in high-powered tiny computers in the hands of more and more consumers. Now with almost half the world's population owning a smartphone, there is potential for this group of smartphone users to tap into the powerful computers in their pockets as a digital tool for building (Silver, 2019). ...
Conference Paper
Full-text available
This paper investigates how augmented reality (AR) can instruct and assist in assembling an earthen structure consisting of a limited set of geometrically different interlocking blocks. By adapting a visual-inertial object tracking software, to the assembly process of a mortarless, compressed earth block (CEB) dome, the construction site no longer needs physical templates and manuals. This enables the builders to have real-time tracking with visual feedback to actively adjust according to the optical guidance during the course of assembly. Two identical dome structures are built with the same set of earth blocks, one with AR and one without. The results show that using AR can significantly improve construction efficiency for complex, dry-stacked structures as it acts as assembly guidance and provides insight into the limits of the tracking tolerances. Further, this paper discusses the limitations and challenges and can provide an outlook for further research scaling up the production to construct a habitable dwelling. Starting with just a pile of dirt and a mobile phone, the demonstrator exhibits the compatibility of local, sustainable materials and digital, efficient processes.
... Past scholars (Dolas and Ulukavak, 2021) have referred to AR as a technology that can increase interaction with the real world and objects. Arth et al. (2015) defined augmented reality as a technology which enables computer-generated sound and images to be used in the real world. The concept of AR is also termed as a newer technological system in which virtual objects are added to the real world in real-time during the user's experience (Cipresso et al., 2018). ...
Article
Full-text available
This paper aims to explore the virtual world and television advertising of tourism destinations with insights from Africa. The continent of Africa experienced lock-downs and travel restrictions due to the Coronavirus disease 2019 (COVID-19) global pandemic which affected the tourism sector of many countries. Tanzania did not have a lock-down during the entire COVID-19 global pandemic hence the advantage over other African countries to the extent of being elevated to a middle-income country. Whilst there were positive effects of the pandemic, the negative effects tremendously reduced tourist numbers in destinations. However, existing technologies in the virtual world that can benefit the tourism industry such as Augmented Reality (AR) have received less attention in the Africa research agenda. To expand the scope of AR, this paper specifically explored AR technologies and television advertising of tourism destinations in the context of Tanzania. A literature review method with content analysis is used as methodology approach. The findings have shown that there are opportunities of AR in relation to television advertising of tourism destinations which include the ability to integrate AR technologies with television advertising of tourism destinations. The practical implication is for tourism stakeholders to consider the use of AR technologies to sustain television advertising of tourism destinations.
Article
Full-text available
As internet became an indispensable part of our lives, recent advancements in digital technologies, wide adaptation and usage of smartphones and tablets have changed visual media as well as impacting traditional children’s books. This article will examine the relation between children’s book illustrations and augmented reality applications. It will cover a brief history of existing examples of interactivity in children’s books and will conclude with an applied example of augmented reality project.
Conference Paper
Full-text available
In this paper, we propose a complete on-device 3D reconstruction pipeline for mobile monocular hand-held devices, which generates dense 3D models with absolute scale on-site while simultaneously supplying the user with real-time interactive feedback. The method fills a gap in current cloud-based mobile reconstruction services as it ensures at capture time that the acquired image set fulfills desired quality and completeness criteria. In contrast to existing systems, the developed framework offers multiple innovative solutions. In particular, we investigate the usability of the available on-device inertial sensors to make the tracking and mapping process more resilient to rapid motions and to estimate the metric scale of the captured scene. Moreover, we propose an efficient and accurate scheme for dense stereo matching which allows to reduce the processing time to interactive speed. We demonstrate the performance of the reconstruction pipeline on multiple challenging indoor and outdoor scenes of different size and depth variability.
Conference Paper
Full-text available
All existing methods for vision-aided inertial navigation assume a camera with a global shutter, in which all the pixels in an image are captured simultaneously. However, the vast majority of consumer-grade cameras use rolling-shutter sensors, which capture each row of pixels at a slightly different time instant. The effects of the rolling shutter distortion when a camera is in motion can be very significant, and are not modelled by existing visual-inertial motion-tracking methods. In this paper we describe the first, to the best of our knowledge, method for vision-aided inertial navigation using rolling-shutter cameras. Specifically, we present an extended Kalman filter (EKF)-based method for visual-inertial odometry, which fuses the IMU measurements with observations of visual feature tracks provided by the camera. The key contribution of this work is a computationally tractable approach for taking into account the rolling-shutter effect, incurring only minimal approximations. The experimental results from the application of the method show that it is able to track, in real time, the position of a mobile phone moving in an unknown environment with an error accumulation of approximately 0.8% of the distance travelled, over hundreds of meters.
Conference Paper
We propose a direct (feature-less) monocular SLAM algorithm which, in contrast to current state-of-the-art regarding direct meth- ods, allows to build large-scale, consistent maps of the environment. Along with highly accurate pose estimation based on direct image alignment, the 3D environment is reconstructed in real-time as pose-graph of keyframes with associated semi-dense depth maps. These are obtained by filtering over a large number of pixelwise small-baseline stereo comparisons. The explicitly scale-drift aware formulation allows the approach to operate on challenging sequences including large variations in scene scale. Major enablers are two key novelties: (1) a novel direct tracking method which operates on sim(3), thereby explicitly detecting scale-drift, and (2) an elegant probabilistic solution to include the effect of noisy depth values into tracking. The resulting direct monocular SLAM system runs in real-time on a CPU.
Conference Paper
Obtaining a good baseline between different video frames is one of the key elements in vision-based monoc-ular SLAM systems. However, if the video frames contain only a few 2D feature correspondences with a good base-line, or the camera only rotates without sufficient translation in the beginning, tracking and mapping becomes unstable. We introduce a real-time visual SLAM system that incrementally tracks individual 2D features, and estimates camera pose by using matched 2D features, regardless of the length of the baseline. Triangulating 2D features into 3D points is deferred until keyframes with sufficient base-line for the features are available. Our method can also deal with pure rotational motions, and fuse the two types of measurements in a bundle adjustment step. Adaptive criteria for keyframe selection are also introduced for efficient optimization and dealing with multiple maps. We demonstrate that our SLAM system improves camera pose estimates and robustness, even with purely rotational motions.
Conference Paper
Ego-motion estimation for an agile single camera moving through general, unknown scenes becomes a much more challenging problem when real-time performance is required rather than under the off-line processing conditions under which most successful structure from motion work has been achieved. This task of estimating camera motion from measurements of a continuously expanding set of self-mapped visual features is one of a class of problems known as Simultaneous Localisation and Mapping (SLAM) in the robotics community, and we argue that such real-time mapping research, despite rarely being camera-based, is more relevant here than off-line structure from motion methods due to the more fundamental emphasis placed on propagation of uncertainty. We present a top-down Bayesian framework for single-camera localisation via mapping of a sparse set of natural features using motion modelling and an information-guided active measurement strategy, in particular addressing the difficult issue of real-time feature initialisation via a factored sampling approach. Real-time handling of uncertainty permits robust localisation via the creating and active measurement of a sparse map of landmarks such that regions can be re-visited after periods of neglect and localisation can continue through periods when few features are visible. Results are presented of real-time localisation for a hand-waved camera with very sparse prior scene knowledge and all processing carried out on a desktop PC.
Conference Paper
We propose a fundamentally novel approach to real-time visual odometry for a monocular camera. It allows to benefit from the simplicity and accuracy of dense tracking - which does not depend on visual features - while running in real-time on a CPU. The key idea is to continuously estimate a semi-dense inverse depth map for the current frame, which in turn is used to track the motion of the camera using dense image alignment. More specifically, we estimate the depth of all pixels which have a non-negligible image gradient. Each estimate is represented as a Gaussian probability distribution over the inverse depth. We propagate this information over time, and update it with new measurements as new images arrive. In terms of tracking accuracy and computational speed, the proposed method compares favorably to both state-of-the-art dense and feature-based visual odometry and SLAM algorithms. As our method runs in real-time on a CPU, it is of large practical value for robotics and augmented reality applications.