This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.
Today the navigation systems are growing with a swift speed and have become a very crucial technology in the multimedia context. This technology is needed by almost
every one of us to explore and navigate around the globe. At present developers are
trying to make it as much portable and robust as they can. This research paper
explains the assimilation of an enhanced camera pose retrieval system into a
landmarked map based navigation system for the mobile devices. The system's
capability of retrieving information from the environment can be very useful to those
who are visually challenged or blind people. In mobile devices the improved sensors
and cameras like GPS, Radio Signals, etc makes it more possible to provide an
accurate location. Geographical navigation using a mobile device is a three way Map-
Navigation system which includes the Joystick Navigation method, The Magic lens
method with external visual surroundings and the Peephole paradigm without a visual
surrounding. The joystick method is the simple method of navigation using the arrow
key in the device keeping it stable or static. The Magic lens process deals with an
external content with a geographical layout, providing visual context outside the
device display. The Peephole method is used by moving the device with fixed map
with respect to the external frame of reference, without providing any visual data
outside the device. After comparing all the three methods, the result demonstrates the
benefit of Peephole and Magic lens process over the joystick process in terms of
searching delays, speed and the accuracy in the degrees of navigation.
Geographical Maps, Mobile Sensors, Interaction techniques based on camera, 3D
Table of Contents:
2. Literature Reviewâ€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦.2
3. Research ...............â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦.â€¦â€¦...3
4. Conclusion â€¦â€¦â€¦â€¦â€¦â€¦â€¦.â€¦â€¦.â€¦â€¦â€¦â€¦â€¦....â€¦â€¦â€¦â€¦â€¦â€¦â€¦.â€¦..â€¦..7
5. Referencesâ€¦.....â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦â€¦... 8
2 | Faculty of Computing, Engineering & sciences S. Rahman: 11004728
Today, the mobile devices have become a basic requirement for maintaining the flow
of our daily lives. It is integrated with set of different applications and it is so
common because of its cheap price, portability, light weight, easy to handle nature,
etc properties. One among all of these applications is Geographical or Map navigation
which helps the users to understand and navigate their locations, especially to those
people who operate in the fields like tourists, archeologists, geologists, etc. Often they
use to carry a huge amount of paper sheets containing maps, plans, descriptions for
the places etc along with them. Mobile phones provide the users high portability and
with an advance computing technology. Apart from this it has sensors like GPS
systems, Radio Signals and the Digital Cameras which provide accurate navigating
information (Georg 2007). Mobile phones' universal connectivity give users access to
online sites, allowing them to both consume and produce geographically meaningful
data, such as to tag images for www.flickr.com or in social networking sites like
Facebook and Twitter. But after all these benefits a problem is that on these devices
users suffer from small display sizes and low resolutions which creates difficulties in
identifying the accurate landmarks on these maps, and so offering a limited support
while navigating a large scaled map on small screen.
For almost every city, public maps are universally available. They provide
information of the locations to all who are travelers and even to those who are locals
and want to explore unfamiliar places. These maps only provides them some general
information e.g. the street names and important places etc. The detailed information
like locations of banks, shops, clubs, hotels, etc are omitted as these will make the
map more complex and complicated to understand. It normally makes users confused
when more elements are used in a map (Roska 2009).
This research paper is an investigation on alternate navigation techniques based on the
Magic Lens method and the Dynamic Peephole comparisons. It will demonstrate the
implementation of Dynamic Peephole method using camera of the cell phone which
detects the position of the user and matches it with the location shown on the display
of the mobile device (Raubal 2007). The Magic Lens method of geographical
navigation is normally highly structured and uses the image investigation process to
track and detect the location of the device. It uses the mobile device as a see through
tool and the users can explore all dimensional (3D) geographical navigation on the
map (Michael, Johannes 2007).
2. LITERATURE REVIEWS
Today several types of information are available in the Internet. This navigation
system makes use of the database which contains photographs and metadata
information of all the locations. It uses the LSB- location Based Services and filters
the positions. Prof. Dr. Georg Granter (2007) and Karl Rehrl (2009) had provided
very useful information on the LSB system which helps the users to interact and
navigate maps on their mobile devices through simple interaction methods like
keypad, trackball and joystick navigations in the phone.
Achim Ebert (2004) had made research on human aspects of visualization and
provided an overview of sencemaking and construction of information using visual
3 | Faculty of Computing, Engineering & sciences S. Rahman: 11004728
information in the mobile devices. Today Google Maps on the smart phones and other
phones are live example of these navigation techniques.
M.Welsman-Dinelle (2005) used maps enabled with an array of Radio Frequency
Identification - RFID in which joystick is used to point and tag to realize the physical
hyperlinks. The demerits of this process was it was very costly and also the resolution
of the map on the small screens of mobile was not that much good because of the
additional tags of RFID. Later on by R.Argue (2007) this method was improved for
achieving 3Dimensional resolutions. The Magic Lens method of navigation using
mobile devices is the process of enlarging or amplifying a physical background just
like a magnifying glass. It uses the mobile camera to enhance the users' views for
geographical navigation (Rekimoto 1995). The Magic Lens application is used by the
camera of the mobile phones. It records the map's characteristics and features with the
help of the camera when it is positioned on a map over a predefined distance. These
data can be identified on the screen in real time. It uses the classical video see through
which are enlarged 3Dimensional geographic information that are then overlaid over
the map (Roska 2009).
P.Werkhoven (2006) had compared the static and the dynamic peephole method
images and provided evidence that dynamic method is more reliable and advance for
tasks like 3D navigation and suitable display resolution issues.
Johannes Schöning (2006) used a marker based approach with the help of Ultra
Mobile PC-UMPC and a Symbain Smartphone. This method had lots of demerits like
when they had used the markers it had scattered the map and divided it into several
parts, disturbing the usability. The main use of this approach was it had the ability of
geospatial or 3D conversions.
Kristóf Karacs (2009) developed their experimental system for mobile visual
navigation system in which they used a mobile phone and a compact cellular visual
computer which was capable of detecting objects, and even it can read and understand
activities in its surroundings with some predefined instructions which help the blind
people in their daily lives. They had developed the system using algorithms of
detecting pedestrian crosswalk.
William Plant and Gerald Schaefer (2009) proposed the algorithms which distinguish
between horizontal browsing and vertical browsing with mobile devices. Horizontal
browsing works on images of a visual plane, which also included functions like zoom,
magnify, scale etc. The vertical browsing allows navigation to different aspects of
hierarchically arranged visualizations. They had used the time stamp information for
Yuntai Kyong (2007) developed a mobile robot with a memory which stores a
sequence of global locations. They used the self localization algorithms for a
geographical map based navigation process. As the maps are in 2D projection, the
idea was to transform 3D geographical locations into 2D map based navigation with
the help of camera based tables.
The previous done work and research as stated in the literature review section had
provided a very good support for this research to be implemented. The research
4 | Faculty of Computing, Engineering & sciences S. Rahman: 11004728
includes three methods of navigation techniques which are (1) The static peephole
Method of Navigation, (2) The Dynamic Peephole navigation method without visual
surrounding and (3) The Magic lens Navigation method with a visual surrounding
(Rohs & Schöning 2007). These three methods are explained as follow:
Figure 1: The three methods of navigation: Static Navigation (left), the dynamic peephole
method without visual surrounding (middle) and the Magic lens method with visual
surrounding (Left). (Krüger 2007)
(1) The Static peephole Method of Navigation: This is a simple navigation process in
which scrolling keys are used for the navigation of maps. Nowadays instead of
joystick touch screen phones are available which makes it more convenient to be
used. These techniques are called static peephole methods of interface i.e. the window
is kept still but the subject moves behind it. A mobile is used with a standard
multidirectional navigation joystick as shown in figure 1 (left). In this experiment the
button of the joystick was held in one direction and accordingly parameters are
recorded. The timer starts generating updates @10Hz in which the content of the
display moves by about 35 pixels in the corresponding direction, which is equal to
50mm/sec in the testing mobile device. The map size was about 1800 X 1200 pixels
and has a limited scrolling area; it took about 3 seconds to scroll from left to right and
about 2 seconds to navigate from top to bottom. In this process the display was almost
clear and readable when navigated in an average speed but if it is scrolled fast it is
hard to read the contents of the display.
(2) The Dynamic peephole method of navigation without visual surrounding: This is a
grid based technique in which the mobile's camera is tracked in a marked base area.
There is no visual surrounding for this process. Basically the grids contain frames of
references for a virtual map on the display. In this method the mobile device is moved
and the map is fixed in space. The user needs to assimilate the layout of the map for a
while. The grid are defined with (x, y, z) coordinates which are computable by the
mobile device. This coordinating system helps in tracking large area of navigation.
The camera lens is kept at about 10 cms above the grid by which the Z coordinate is
5 | Faculty of Computing, Engineering & sciences S. Rahman: 11004728
omitted. Digital zooming techniques are used which are adjusted such that the marker
appears at a suitable position. In this method the reorganization range is increased up
to 45 cms. The grid border scales the map on the display with respect to the distance
between the camera and the image map. The zoom out and zoom in stands for moving
far and moving closer to the grid respectively.
(3) The Magic Lens method of navigation with visual surrounding: This process is
similar to that of Dynamic peephole method but instead of using a grid based surface
a camera phone is tracked over a fixed map. The user can read all the information
available on the map. The device displays dynamic information to the user which is
used along with the static overview to gain a better navigation experience or facilities
as shown in figure 1(right). It used the tracking algorithm which computes the
position of the camera and then generates accurate graphics and pixels on the display.
The steps of the tracking algorithm are as follows:
i) It finds the map dot candidates in which some particular size and axis ratios are
ii) It finds the edges for an efficient lookup.
iii) It finds the patches in which the dots are subdivided into square areas.
iv) It samples each of the patches into 12 X 12 pixels.
v) It observes the connections between each patch.
vi) It calculates the maximum corresponding indices.
vii) It finds some dependable patch using voting techniques.
viii) It sets up the dot coordinates.
ix) Finally it calculates the maximum wrapper in which the graphical overlays are
defined according to the closest dots of the image map.
In a user study conducted by Rohs (2007) for the comparing all of these three
methods of map navigation, in the starting it was assumed that the Dynamic peephole
method of geographical navigation will beat up the performance of Static Joystick
method of navigation and the Magic Lens i.e. the map based method of navigation
will beat up the performance of the Dynamic peephole i.e. the grid based method of
navigation. Rohs and his team had performed the experiments such that it will join
and combine the static long term data present in the Maps with an online dynamic
database with the help of a mobile GPRS system.
They amplified the maps using different symbols like marking the shopping malls,
parking lots, petrol bunks, etc, with some hourly updates and ratings e.g. number of
vehicles parked per hour in a parking lot, rating of shopping malls according to their
cheapness, etc. After doing this they choose a task to search the cheapest Shopping
mall in the map. The task can be easily performed by an automatic device but there
are many tasks where the users have lack of knowledge in the starting. This task
resulted in marking all the shopping malls in the map according to their rates and also
during this process several numbers of tasks are automatically generated.
These three methods are further investigated and the time average and error rates are
mapped out in graphical representations as shown in the figure (2) and figure (3)
6 | Faculty of Computing, Engineering & sciences S. Rahman: 11004728
Figure 2: Mean search time (left) and mean Error rate (right) of the three methods
Figure 3: Rate of omission of targets (left) and Rate of viewing per target in seconds (right)
Comparing the performance of all the three methods as in the graphs it is clear that the
dynamic peephole method and the magic lens method of navigation is better than that
of static navigation method in terms of time and accuracy. But when comparing the
7 | Faculty of Computing, Engineering & sciences S. Rahman: 11004728
results for magic lens method and dynamic method both are not that much different.
The map method navigation is not that much faster than that of the grid method but it
provides a fixed visual surrounding. The reason might be the system performance or
the human visual readings. Technically the implementation of map method has slow
update and detection rates when compared to the grid method which is quite fast
because of the digital zooming process. The map has a limited range for the distance
and if it crosses the boundary limit the locations are not properly displayed. But in
map based navigation the display quality does not depends upon the quality of the
camera of the device.
4. CONCLUSION AND FURTHER WORKS
This research paper is based on three methods of geographical navigation techniques.
The first method is the static peephole navigation process, which is a simple joystick
based navigation technique based on visual tracking. The second method is the
dynamic peephole technique of navigation without a visual surrounding, which is
implemented by using the camera of a mobile device where it tracks the grids which
are encoded with maps. The third method of navigation is the magic lens technique
where the mobile device camera is used to track a visual map. All the three methods
are being compared with each other and the results concludes that the navigation is
experienced better when the device is moved over a fixed map, rather than that of a
moving virtual map. The device can track and explore the maps easily. Although the
methods of navigation had been improved and present techniques are very useful but
there are still areas which need to be improved. In these techniques it is very
important to choose the camera positions for getting better results. The navigation
performance is better experienced while moving the mobile device over a fixed map
than that of moving a virtual map behind a mobile device keeping the device in a
fixed position. In device tracking condition almost all the parts of the map is covered.
It is somewhat expensive using different techniques. In future experiments will be
done for a better result. Balancing the device in a desired view by the user according
to his/her convenience still needs further research. The given experiment includes
simple 3 Dimensional explorations and inspecting the map using the mobile device
display covering all the objects on the map. The magic lens approach can be further
improved by extracting more valuable informations from the fixed map. In future the
magic lens and the grid method process can be combined together for a better
navigation process. Several works are being done and are to be implemented in this
field regarding the investigation of the process and new technologies are to be
8 | Faculty of Computing, Engineering & sciences S. Rahman: 11004728