EVALUATION & TRENDS OF SURVEILLANCE SYSTEM NETWORK IN UBIQUITOUS COMPUTING ENVIRONMENT

Int. J. Advanced Networking and Applications
Volume: 6 Issue: 5 Pages: 2487-2494 (2015) ISSN: 0975-0290
2487
EVALUATION & TRENDS OF SURVEILLANCE
SYSTEM NETWORK IN UBIQUITOUS COMPUTING
ENVIRONMENT
Sunil Kr Singh
CSE Department, Bharati Vidyapeeth College of Engineering
Affiliated to GGSIP University, New Delhi, India
E-mail:drsunilsingh@acm.org
Anuj Aggarwal
ECE Department, Bharati Vidyapeeth College of Engineering
E-mail:anuj.aggarwal@acm.org
Kavneet Kaur
CSE Department, Bharati Vidyapeeth College of Engineering
E-mail: kavneetk@acm.org
---------------------------------------------------------- ABSRACT--------------------------------------------------------------------------
With the emergence of ubiquitous computing, whole scenario of computing has been changed. It affected many
inter disciplinary fields. This paper visions the impact of ubiquitous computing on video surveillance system. With
increase in population and highly specific security areas, intelligent monitoring is the major requirement of
modern world .The paper describes the evolution of surveillance system from analog to multi sensor ubiquitous
system. It mentions the demand of context based architectures. It draws the benefit of merging of cloud
computing to boost the surveillance system and at the same time reducing cost and maintenance. It analyzes some
surveillance system architectures which are made for ubiquitous deployment. It provides major challenges and
opportunities for the researchers to make surveillance system highly efficient and make them seamlessly embed
in our environments.
Keywords- surveillance systems, ubiquitous computing, cloud based surveillance, challenges in surveillance system,
generations of surveillance system
-------------------------------------------------------------------------------------------------------------------------------------------
Date of Submission: November 20, 2014 Date of Acceptance: January 13, 2015
-------------------------------------------------------------------------------------------------------------------------------------------
1. INTRODUCTION
A noteworthy quote by Mark Weiser “the most
profound technologies are those that disappear. They
weave themselves into the fabric of everyday life until
they are indistinguishable from it.” [11]. These words
gave emergence to the new third wave of computing
know as Ubiquitous computing (UbiComp), which
prospects the vision of making computing appear
everywhere and anywhere. It enhances and empowers
the human-computer interaction to a whole different
dimension in which the user is surrounded by a
complete smart environment with devices/sensors
communicating with each other and combining their
functionalities to provide an array of amalgamated
services. Fundamentally, it takes conventional
computing that deals with virtual world to modern
computing which deals with physical or real world with
unobtrusive human interactions.
A feel of the ubiquitous computing concept can be
perceived by a simple example. Let us consider that a
person in sitting in a room and his clothes are fabricated
with invisible biometric monitors. It records his
behavior and movement i.e. sleepy, excited, and
reading, etc. With the calculation of readings, smart
environment controls light brightness, air conditioner
temperature and noise, thus making one’s life more
comfortable.
1.1 SMART DEI MODEL
A three dimensional approach called the smart DEI
model has been proposed to analyze and design a
comprehensive framework for ubiquitous computing
[17]. The dimensions are included in the acronym DEI,
in which ‘D’ stands for Device, ‘E’ for Environment
and ‘I’ for Interaction. Model consists of architecture
design, internal model and interaction with physical
environment.
1.1.1 Architecture Design
The architecture design of smart DEI model is a sack of
three types: Smart Device, Smart Environment, and
Smart, Smart interaction. ‘Smart’ means that the entity
is active, digitally networked, can operate to some
extent autonomously, is reconfigurable and has local

2488
control of resources (such as energy, data storage,
etc.) It needs. Smart devices mostly focus on
interaction within a virtual world and are less context
aware of the physical world compared to smart
environment devices. They are manually activated
devices. Smart devices consist of three basic forms with
three additional forms. The three basic forms consist of
Tabs (wearable centimeter sized devices), Pads (hand
held decimeter sized devices), and Boards (meter sized
interactive display devices) .We have seen these basic
forms in our environments but three additional forms
which are about to blow up in future, giving future
technologies a new shape, are Dust, Skin, and Clay.
Dust is a miniature sized device that can be without
visual output displays. It senses every minute thing and
can be fabricated ubiquitously such as on buildings,
streets, traffic. This device may be proven highly
effective in military. Skin is the fabrics based upon light
emitting and conductive polymers. Organic computer
devices can be formed into more flexible non planar
display surfaces and products such as clothes and
curtains. Example – OLED (Organic Light Emitting
Diode) technology can be used as an efficient way of
lightning and it is flexible too. Thus may be employed
in future PDA’s. Clay ensembles MEMS (Micro Electro
Mechanical System) that can be formed into an
arbitrary 3D shapes as artifacts resembling many
different physical objects. One such example is tangible
user interface.
Smart Environment consists of a set of networked
devices that have some connection with the physical
world. These devices are strongly context aware of their
physical environment. Example- automated door
opens/closes.
Smart Interaction consists of components that
dynamically organize and interact to achieve goals. This
organization may occur internally without external
influence, thus making it a self organized system.
1.1.2 Internal model
Internal model of smart DEI is based upon five
fundamental properties, namely, Distributed ICT, iHCI,
Context Awareness, Autonomy, and Artificial
Intelligence. Distributed ICT systems are in layers, in
which bottom layer forms hardware, middle layer forms
operating system, and top layer forms human computer
interaction (HCI). Implicit human computer interactions
(iHCI) consist of calm computing and systems that
interact autonomously with human actions without
being noticed. Context awareness is to make a system
dedicated to a particular task or context rather than
supporting all. Autonomy can be defined as a system
which is self governing and capable of own independent
decisions and actions. Artificial Intelligence refers to
intelligence or decision making by machines pertaining
to some algorithm or sensors.
1.1.3 Interaction with external environment
This model consists of three types of interactions
 The interaction with virtual environment
(conventional C2C computing like mobile
phones).
 HCI interaction between human and computer.
 CPI interaction between computer and physical
world
1.1.4 Aims of smart DEI model
The ubiquitous systems should be as small as possible
so that they can be hidden in our environment. They
should be inexpensive so that they can reach to every
human in every corner of the world. They should form a
robust network which do not fails as once they enter our
lives; we shall be highly reliable on them. The
ubiquitous systems should be mobile to provide
nomadic computing. The user interfaces on the systems
should be as simple as possible. They should form a
reliable system.
1.2 APPLICATIONS
Ubiquitous computing has wide range of applications as
it can be embedded everywhere in our environment. It
can be applied on devices forming smart PDAs, smart
classrooms, medicinal purpose, home environment,
high performance systems, and surveillance system.
Significantly, the vision of surveillance will be
revolutionized with the advent of ubiquitous computing.
It will not be very hard to figure out that you are being
watched, due to video surveillance cameras protruding
from every building. But with the internet of things, the
surveillance grid will unite seamlessly and invisibly into
the entire environment. In a ubiquitous environment,
every object, as well as person who wears RFID tagged
clothes or are using electronic devices, would be
“readable” by a computer or wireless network. The
node’s details, exact location and other information can
be obtained electronically by invisible sensors in
sidewalks, roads, or doorways. Ubiquitous vision
system utilizes redundant visual information for robust
monitoring tasks in large scene area. Several vision
sensors observe a common area and provide redundant
information [6]. This redundant observation contains
rich information for robust vision functions. Key
specifications of the ubiquitous vision system are
summarized as follows:
 Covers large scene area to observe dynamic events
happening in the environment.
 Tracks dynamic event in real time.
 Synthesizes views for visualization at arbitrary
viewpoint.
 Enables us to develop integrated information
framework that can access both real world and virtual
world through computer network.

2489
2. SURVEILLANCE SYSTEMS
In domains of computer vision, the needs of intelligent
surveillance systems are rising in huge demand. The
deployment of distributed surveillance systems in
commercial, government monitoring and military
operations plays a vital role. Recent terrorist attacks and
criminal offences made it necessary to embed efficient
surveillance systems in the environment. Networks of
large number of cameras are required to provide wide
scene coverage for many surveillance tasks. In
designing such networked camera systems,
considerations of practical aspects matters such as cost,
complexity and robustness. Automated surveillance
systems deals with real time monitoring of persistent
and transient objects within a specific environment. The
primary aim of these systems is to provide an automatic
interpretation of scenes, to understand and predict the
actions along with interactions of the observed objects
based on the information acquired by sensors. The main
stages of processing in an intelligent visual surveillance
system are: moving object detection and recognition,
tracking, behavioral analysis, and retrieval.
Figure 1: Generalized architecture of visual surveillance
system [18]
2.1 FIRST GENERATION
The intelligent surveillance systems are evolved from
analog CCTV systems. [1] In that systems multiple
cameras are placed in remote locations and were
connected to set of monitors kept in a monitoring room.
It needs a person to monitor all the events. They were
mainly used to store visual recordings and use the
recordings, if any mishappening takes place. It uses
rectilinear images for stereo matching and their cameras
were arranged densely with short baseline. The scene,
where they were looking at, was restrained to small
area. Therefore, the major problems of such systems are
less intelligent monitoring, and data degradation due to
digital data recorded being converted to analog for
transmission to monitors leading to attenuation in
signals.
2.2 SECOND GENERATION
After first generation, around 1980 the second
generation of computer came, that uses algorithms in
digital video processing for automated data extraction
from CCTV systems and alert the concerned person or
authority if something unusual or illegal takes place. It
also helps the government authorities to track specific
persons and vehicles in case of breaching of laws.
2.3 THIRD GENERATION
The third generation is now taking its shape with the
help of various kinds of sensors which are being
embedded ubiquitously in environment. It uses systems
which are very efficient and provides large number of
information about an environment from recordings, in
real time by offering omni-directional view. The
agenda is to make an extensively distributed multi-
sensor wiretap system, possessing concentrated, and
time authenticated computer algorithms which enable
execution on several applications employing minimum
manual reconfiguration. Such devices must be
compatible to the extent that, they adopt and combat
variations (physical, natural or geographical) in the
environment. The merger of radio communication
technologies and algorithms, for calculating locations,
constitutes smart video security surveillance, which in
turn is based on pervasive sensor network technology.
Wireless Fidelity (Wi-Fi), Radio Frequency
Identification (RFID), ZigBee, Ultra Wide Band
(UWB), and etc., are examples of representatives’ radio
communication technologies. Angle of Arriva (AoA),
Time of Arriva (ToA), Received Signal Strength
Indication (RSSI), Chirp Spread Spectrum (CSS), and
etc., are some of the algorithms employed for
computing locations. CSS method, which is known for
its accuracy amidst location identification methods in
radio technology, is engaged for intelligent video
security surveillance system based on ubiquitous sensor
network technology.

2490
Table 1: Phase Wise Evolution of Surveillance System
3. BASIS OF FUTURE SURVEILLANCE
SYSTEMS
The foundation of future surveillance systems, based on
ubiquitous computing, will require interplay/integration
of various interdisciplinary fields. In this paper we are
concentrating on essential factors such as cloud
computing and context awareness that will contribute to
surveillance systems.
3.1 CLOUD BASED MULTIMEDIA
SURVEILLANCE SYSTEMS
With the advent of cloud based computing, surveillance
systems will be enormously benefited as video
recording can be stored on cloud (on a remote server
with access of data via internet). This will lead to
establishment of cheaper surveillance systems as there
will be no need for the user to buy and maintain
physical storage devices. This will also lead to lesser
manpower requirement to maintain and control the
surveillance system as software based surveillance
systems will be put in place [16]. The main benefit of
the cloud based surveillance system is automated
backup of data on the cloud which can be accessed from
anywhere in the world. As the cost and maintenance
overhead will be minimal, small organizations and
companies can also establish their own surveillance
system.
3.2 CONTEXT BASED SURVEILLANCE
SYSTEMS
In ubiquitous computing environment, multiple sensors
are embedded to provide accurate information and data
recording. As military surveillance, home surveillance,
city market surveillance have its own requirements,
there is a need to develop target-oriented architecture.
Therefore, global ubiquity of a particular architecture or
model of surveillance systems cannot be achieved.
Hence, surveillance systems must be context aware so
that the architecture may be fabricated as per user
requirement in different locations.
4. ARCHITECTURAL ANALYSIS OF MODERN
SURVEILLANCE SYSTEMS
There are various existing architectures for ubiquitous
surveillance systems having their own merits and
demerits. In this paper we are analyzing some renowned
architecture of surveillance systems which are proposed
for ubiquitous deployment. Some of the highly
successful modern surveillance systems are DETER,
PRISMATICA, VIGILANT
4.1 DETER
DETER (Detection of Events for Threat Evaluation and
Recognition) is a surveillance system for commercial
outdoor environments such as parking location.[19] It is
used to monitor vehicles, objects and pedestrians. It
visions the bridging of gap between current systems that

2491
use to notify isolated events and automated systems that
can notify systems without any human user. The
technique DETER used is fusion of overlapping field of
vision of various cameras. The threat analyzer is
assisted by off line thread model that analyses in real
time whether specific activity is threat or not. It reports
abnormal moving patterns of pedestrians and vehicles in
outdoor environment.
Figure 2: Architecture of DETER surveillance system
[19]
4.2 PRISMATICA
PRISMATICA was an EU funded project that was
made to monitor tramps and buses, and become part of
its surveillance system. The agenda of the project is the
fusion of visual and audio information, using crowd and
object detection algorithms to perform analyzing
process within the subsystem and transferring of high
level data so that no loss of clarity and misidentification
happens [20]. Basically in PRISMATICA system, the
tasks and processes are performed locally in each
computer forming a distributed system in which each
device is carrying out their own standalone processes
which are then connected and synchronized with the
help of COBRA (Common Object Request Broker
Architecture) and it communicates only high level
information to the monitor room.
4.3 VIGILANT
VIGILANT is a multi-camera surveillance system that
understands the scene with query driven search
algorithm and can generate retrospective video report
from previous events. [21] It is utilized to monitor the
pedestrians walking in a parking lot. VIGILANT
system tracks people across multi-cameras using
software agents. An agent is made for each camera for
each detected person and those agents from each
camera communicates to obtain a combine decision to
analyze that each agent is tracking the same person or
not by using trajectory geometry.
Figure 3: Architecture of PRISMATICA surveillance
system [20]
Figure 4: Architecture of VIGILANT surveillance
system [21]
5. Qualitative comparisons of results obtained from
various techniques used in DETER, PRISMATICA
and VIGILANT surveillance systems
We would like to highlight some results of various
techniques used in three models i.e. DETER,
PRISMATICA and VIGILANT, so a researcher
following the particular architecture can look and
modify the architecture for increased efficiency. The
flaws and effectiveness of all surveillance systems are
clearly mentioned to have a proper understanding of the
usefulness of these models in environment.
· DETER- it uses the technique of overlapping field of
vision of different cameras and a threat analyzer to
analyze the event. As on the 16 hour testing experiment
conducted in Honeywell Laboratories it was found that
the system successfully classified and identified most of
the vehicles and people but on detection of paths it
called 32 times false alarms and several times it also
missed tracks of people under surveillance. [23]
PRISMATICA- it uses the technique of fusion of audio
and video sensors, and it performs most of the
computation at local computers and sends only the high
level information to the monitor room. It is a highly
successful project and was fabricated in many buses and
tramps in London, Paris and New castle which lead to

2492
the decrease of robberies, thefts and pick-pocketing by
32% and decrease of breaches by 52% and the overall
decrease of threat activities by 23% evaluated over
three months. [24]
VIGILANT- it uses the technique of query based
search and focuses on creation of user friendly GUI for
monitoring and classification of objects and people. The
VIGILANT surveillance setup was installed in one of
the university car parks to test the efficiency. To
monitor wide field of view of car park entrance and
exit, various cameras such as pan, tilt, zoom were pre-
set. The experimental setup was used to classify objects
on the basis of height to width and normalized velocity.
To get an account of real monitoring efficiency, a large
data consisting of 320,000 frames was captured during
busy periods over four days in which approximately
200 vehicles and 400 people came into the periphery of
surveillance system. It detected a vehicle event
correctly by 89% and incorrectly marked it as person by
6% and detected a person event correctly by 79% and
incorrectly marked it as vehicle by 19% and others by
5% [21].
6. KEY APPLICATIONS OF SURVEILLANCE
SYSTEMS
The deployment of surveillance systems in ubiquitous
environment has enormous number of applications. In
this paper we are highlighting the major areas which are
listed as:
1) Prime application of the surveillance system is
abnormality detection and warning. It is necessary to
scrutinize the etiquettes of people and vehicles and
determining them as normal or abnormal in certain
situations. Usually there are two methodologies of
warning: One to automatically make a recorded public
announcement whenever any abnormal behavior is
detected and the other is to contact the police
automatically.
2) Identification of specific persons can aid police very
much. Police can build database with biometric details
of the suspect and establish visual surveillance in
certain public areas such as bus stops, markets.
Whenever a surveillance system recognizes the suspect,
immediately location of the suspect will be updated to
the police.
3) Surveillance system can also be used for the
statistical overview of crowd flux and analysis of
congestion in certain public areas such as markets,
intersection of major roads, local rally, football grounds
and provide the information to police and government
authorities to adequate action to control and manage
people.
4) Control of access for people in some security-
sensitive areas such as military bases, hospitals,
significant government authorities, and units in which
special identification such as biometric identification is
used. A database is made in prior to the surveillance
system to automatically recognize a person through his
characters such as height, walking gait in real time.
7. FUTURE CHALLENGES AND OPPORTUNITIES
As we began research in surveillance systems, we
analyzed various challenges that will come before the
researchers to establish surveillance system in
ubiquitous environment. We have tried to propose
major challenges in this paper which is an opportunity
for researchers to solve and provide the world with
highly efficient and secure surveillance system.
7.1 Transmission of data to mobile agents
The clusters of cameras in environment are connected
to local processing proxy server (PPS) which is located
in monitoring room. There are several base stations for
PPS controlling and retrieving useful information from
sets of cameras. Consider a situation, a person jumps a
red light and traffic cop is tracking him. As the
environment is mobile, the PPS changes time to time,
inhibiting synchronization. [4] Thus it demands
efficient automated synchronization of PPS for the
transmission of data to mobile agents.
7.2 Scalability
As real-time surveillance systems grows in
sophistication with hundreds of cameras and sensors
forming a large network, the scalability of its modeling
is greatly challenged.[13] The intensity of computing
interactions will increase many folds leading to
complex algorithms, bandwidth problems, network
failures. New improved technology will be needed to
accommodate such a large network of surveillance
systems in ubiquitous environments. Some architecture
has been proposed for making it scalable but none of
them is able to fulfill the problem properly.
7.3 Privacy
Device used for providing security is itself security
prone. With the evolution in ubiquitous computing,
surveillance is interfering with activities that used to be
considered as private.[5] As phones, credit cards,
CCTV, e-mail, social media, telecommunications, ,
digital documents, and health records already tapped, it
becomes increasingly difficult to find a space where
data are not collected, indexed, distributed, searched,
and inferred.[9][10] . There is need of significant
protocols [8] and security models to safeguard the
privacy of people. [7]
7.4 Energy efficiency
As in ubiquitous environment, the number of distributed
networked video surveillance cameras and sensors will
be large and deployed everywhere to keep track of
activities. Energy efficiency has to put in consideration
for the long term effect of surveillance systems.
Though, lesser research is oriented towards energy
efficiency in surveillance systems. It poses a good
opportunity to researchers to make surveillance system
using minimum energy but still providing a high

2493
performance. Diffusion tree creation and dynamic time
synchronization can be used to minimize the high
energy demands for surveillance system. [3] Reduced
clock speed can also be applied to save energy as
formulated at XEROX PARC. [14]
7.5 Bandwidth
The one problem that is faced in every ubiquitous
environment is a limiting bandwidth. As surveillance
system will make the notion “walls have eyes” real but
as they are all in a distributed network forming an ICT
system demand of bandwidth will be more than it can
be fulfilled. Optical fibers can used in some locations
but majority of system will be wireless.
7.6 Occlusion Handling
Occlusion handling is a challenging problem in visual
surveillance. [2] At the time of occlusion, only fragment
of each object is visible and often at very low
resolution. This problem is usually hard to control and
segmentation of motion based on techniques such as
background subtraction may become undependable. To
decrease vagueness from occlusion, better models are
needed to evolve to handle compatibility and correlate
different characters from multiple cameras and sensors,
and hence eradicate compatibility errors that develop in
tracking of multiple objects. A bit of resolution is
feasible through motion region analysis and partial
matching if objects are occluded by immovable objects
just as buildings, sign boards, street lamps. However,
when multiple moving objects occlude each other,
particularly when their speeds, directions and shapes are
very similar, their motion regions fuses with each other,
which makes the identification and tracking of objects
very difficult.
7.7 Fusion of Data from Multiple Sensors
It is evident that the surveillance systems in the future
will comprise of multiple sensors and it will
revolutionize the visual surveillance information
accuracy. The effectiveness of collaboration between
various cameras and sensors will depend heavily on
fusion of data. Fusion of data is not based on image
processing or decision making but it is based on
characteristic level. [15] The major problem involves
fusion of various types of character such as color, audio,
geometry of objects into a single group to track and
identify objects and therefore understand their
behaviors. Another problem is fusion of characters
obtained from multiple cameras having different
viewpoints and exchange of data about the same object
among themselves.
7.8 Cost effective
The surveillance systems comprising of multiple
sensors must be of low cost as they have to be
fabricated everywhere and if fabrication is performed
by using reconfigurable devices [22] then it will give
more cost benefit. Today, better surveillance systems
are developing such as AURGUS-IS, LIDAR but they
are very expensive. For surveillance system to become
ubiquitous, fabrication of cheap but effective systems is
necessary so that it can be deployed in every location.
8. CONCLUSION
We conclude that the emergence of ubiquitous
computing will affect surveillance system greatly. We
studied the evolution of surveillance systems from
analog CCTV to multi sensor systems. We proposed
that the use of cloud computing will further make the
surveillance systems highly effective and cheap. As
surveillance systems have to be ubiquitous their
architectures must be based on context of user and
location. With the analysis of some surveillance system
architecture, we provided researchers with an overview
of current development so that they can observe
benefits and weak points and create a robust
surveillance system. The significant challenges of
surveillance systems such as occlusion, fusion of data
and others are mentioned in the paper which researchers
should try to eliminate when developing a robust
architecture.
REFERNCES
[1] M. Valera, S.A Velastin, “Intelligent distributed
surveillance systems: a review”, IEEE Vision, Image
and Signal Processing, Vol. 152, No. 2, pp. 192-204,
2005.
[2] Weiru Liu, Paul Miller, Jianbing Ma, Weiqi Yan,
“Challenges of distributed intelligent surveillance
system with heterogeneous information”, Proceedings
of the workshop on Quantitative Risk Analysis for
Security Applications (QRASA'09), pp 69-74, 2009,
[3] Tian He, Sudha Krishnamurthy, John A. Stankovic,
Tarek Abdelzaher, Liqian Luo, Radu Stoleru, Ting Yan,
Lin Gu, “Energy-Efficient Surveillance System Using
Wireless Sensor Networks”, Proceedings of the 2nd
international conference on Mobile systems,
applications, and services (MobiSys '04), pp 270-283 ,
2004.
[4] P Kumar, A Pande, A Mittal, A Mudgal, P
Mohapatra, “Distributed Video Surveillance Using
Mobile Agents”, IEEE International Conference on
Digital Convergence (ICDC 2011), pp 199-204, 2011.
[5] M. V. Venkatesh, S.-C. S. Cheung, J. K. Paruchuri,
J. Zhao and T. Nguyen, “Protecting and Managing
Privacy Information in Video Surveillance Systems”,
protecting privacy in video surveillance, Springer, pp
11-33, 2009.
[6] Kim Ng, Hiroshi Ishiguro, Mohan Trivedi, and
Takushi Sogo, “Monitoring Dynamically Changing
Environments by Ubiquitous Vision System”, IEEE

2494
Workshop on Visual Surveillance (VS'99), pp 67-73,
1999.
[7] A. Longheu, V. Carchiolo, M. Malgeri, G.
Mangioni, “An Intelligent and Pervasive Surveillance
System for Home Security”, IJCC (ISSN: 1841-9836),
Vol.7, No. 2, pp. 312-324. 2012.
[8] Jamal Raiyn, "Detection of Objects in Motion: A
Survey of Video Surveillance," Advances in Internet of
Things, Vol. 3 No. 4, pp 73-78, 2013.
[9]Antti Oulasvirta, Aurora Pihlajamaa, Jukka Perkiö,
Debarshi Ray, Taneli Vähäkangas, Tero Hasu, Niklas
Vainio, Petri Myllymäki, “ Long-term Effects of
Ubiquitous Surveillance in the Home”, Proceedings of
the ACM Conference on Ubiquitous Computing
(UbiComp '12), pp 41-50, 2012.
[10] Jehan Wickramasuriya, Mohanned Alhazzazi,
Mahesh Datt, Sharad Mehrotra, Nalini
Venkatasubramanian, “ Privacy-Protecting Video
Surveillance”, SPIE International Symposium on
Electronic Imaging (Real-Time Imaging IX) San Jose,
CA, 2005.
[11] M. Weiser, “The Computer for the 21st Century”,
IEEE Pervasive Computing, pp 19-25, 2003.
[12] M. Satyanarayanan, “Pervasive computing: Vision
and challenges”, IEEE Personal Communications, Vol.
8, pp. 10-17, 2001.
[13] Tomi Räty, Lassi Lehikoinen, Francois
Bremond, “Scalable Video Transmission for a
Surveillance System”, In the proc. of Fifth
International Conference on Information Technology:
New Generations (ITNG 2008), pp 1011-1016,
2008.
[14] Mark Weiser, Brent Welch, Alan Demers, Scott
Shenker, “Scheduling for Reduced CPU Energy”,
proceedings of the 1st USENIX conference on
Operating Systems Design and Implementation (ODSI
’94), Article No. 2, 1994.
[15] Weiming Hu, Tieniu Tan, Liang Wang, and Steve
Maybank, “A Survey on Visual Surveillance of Object
Motion and Behaviors”, IEEE Transactions on Systems,
Man, and Cybernetics, Vol. 34, Issue 3, pp 334-352,
2004.
[16] M. Anwar Hossain, “Framework for a Cloud-
Based Multimedia Surveillance System”, International
Journal of Distributed Sensor Networks, Hindawi
Publishing Corporation, Volume 2014, Article ID
135257, pp 1-11, 2014.
[17] Stefan Poslad, “Ubiquitous Computing: Smart
Devices, Environment sand Interactions”, ISBN: 978-0-
470-03560-3, published by John Wiley& Sons, Ltd.,
2009
[18] F. Bremond, M. Thonnat, M. Zuniga, “Video
Understanding Framework for Automatic Behaviour
Recognition”, Behaviour Research Methods, Volume
38, No. 3, pp 416-426, 2006.
[19] I. Pavlidis, V. Morellas, P. Tsiamyrtzis, S. Harp,
“Urban surveillance systems: from the laboratory to the
commercial world”, IEEE, Vol. 89, Issue 10, pp 1478–
1495, 2001.
[20] Benny Ping Lai Lo,Jie Sun,Sergio A.Velastin,
“Fusing visual and audio information in a distributed
intelligent surveillance system for public transport
systems”, Acta Automatica Sinica, Vol. 29, Issue 3, pp.
393- 407, 2003.
[21] D. Greenhill, P. Remagnino, G.A Jones,
“VIGILANT: content querying of video surveillance
streams”, Kluwer Academic Publishers, Boston, USA,
pp. 193–205, 2002.
[22] Sunil Kr. Singh et.al, “Performance Evaluation of
Hybrid Reconfigurable Computing Architecture over
Symmetrical FPGA” International journal of Embedded
system & Application, doi:10.5121/ijesa.2012.2312,
Vol.2, No.3, pp 107-116, 2012
[23] V. Morellas, I. Pavlidis, P. Tsiamyrtzis, “DETER:
Detection of events for threat evaluation and
recognition”, Machine Vision and Applications,
Springer-Verlag, Vol. 15, pp 29–2003.
[24] Marcus Sanchez-Svenson, Sergio Velastin et.al,
“PRISMATICA, D13: Key findings and Results”,
version-v.1.0-29.04.2003, 2003.
--------------------

EVALUATION & TRENDS OF SURVEILLANCE SYSTEM NETWORK IN UBIQUITOUS COMPUTING ENVIRONMENT

More Related Content

What's hot (20)

Viewers also liked (19)

Similar to EVALUATION & TRENDS OF SURVEILLANCE SYSTEM NETWORK IN UBIQUITOUS COMPUTING ENVIRONMENT (20)

More from Eswar Publications (20)

Recently uploaded (20)

EVALUATION & TRENDS OF SURVEILLANCE SYSTEM NETWORK IN UBIQUITOUS COMPUTING ENVIRONMENT