https://www.irjet.net/archives/V7/i1/IRJET-V7I1120.pdf

1. International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 07 Issue: 01 | Jan 2020 www.irjet.net p-ISSN: 2395-0072
© 2020, IRJET | Impact Factor value: 7.34 | ISO 9001:2008 Certified Journal | Page 737
Customized campus surveillance system
Sarikonda Poojitha1, Dr. Sunil Bhutada2
1M.Tech Student, Information Technology Department, Srinidhi college of science and technology, Hyderabad,
Telangana, India
2Professor, Information Technology Department, Srinidhi college of science and technology, Hyderabad,
Telangana, India
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - Colleges these days are trying to augment the
quality of being in the campus by integrating the finest
practices of safety along with security with the help of
technology. A significant element of a comprehensive security
system is the deployment of a surveillance camera scheme.
This surveillance in the college is intended to identify the
authorization of the people who try to enter into the college
and aid in protecting the property of the College in
conjunction with student`s safety. These guidelines
concentrate on the College’s safety and security needs while
respecting individual privacy of those attending, working or
visiting the College. This ensures monitoring the presence of
the student in the campus by the parents and the college.
Key Words: surveillance camera scheme, authorization,
monitoring, privacy.
1. INTRODUCTION
Surveillance system is generally significant in the
field of security. Presetdays,videosurveillanceframework is
a significant security resource for business, law
implementation and military applications. Straightforward
frameworks recognize movement in a camera's field of
view. Surveillance cameras are camcordersutilizedtowatch
a region. They are regularly associated with a chronicle
gadget or IP organize, and might be viewed by a security
gatekeeper or law authorization official. Cameras and
recording gear used to be moderately costly and required
human faculty to screen camera film.
Identification and authorization has become an
important subject. This framework provides the identifying
and authorizing a person fromhisentryintothecampus. The
visual Surveillance captures the images of the persons and
identifies them to the already existing database. This
includes installation of a cctv camera to capture the images
of the persons. The fundamental issues when structuring an
insightful CCTV framework are identified with the choice of
the correct focal points and sensors for camcorders utilized,
as these cameras ought to have the option to give 100%
inclusion to the zone being under observation. It plans to
consequently distinguish, perceiveandtrack individualsand
items from picture arrangements so as to comprehend and
portray elements and cooperation among them.
Entryway security or passage entryway security is
important to anticipate the authorization in the controlled
region. A customary checking gadget makes utilization of
costly hardware, includes high power utilization and
requires consistent plate get to which clearly requires
colossal space. The installed arrangements beat these
deficiencies and give live encourages from webcamworking
on cloud without the requirement of the PC. Calculationsare
added to the inserted frameworks to increase their
effectiveness, by making utilization of the open source
libraries, the development location calculations permit
Aurdino Uno to distinguish development, the protest
discovery calculations identify true elements like face of a
person.
The new TensorCam is the security camera with
built-in Artificial Intelligence that has been specifically
designed for capturing images of livingand nonlivingthings.
Here the AI is built on Google’s TensorFlow, Keras, and
Caffe2, and one can setup their own recognition tasks using
the Deep Cognition Deep Learning Studio. We can use the
deep learning capabilities to setup specific tasks, like
detecting certain objects, specific faces, or even particular
kinds of actions. The camera itself has full HD 1080p
resolution and night vision, and will live stream the feed so
one can check it whenever required.
Facial acknowledgment CCTV innovation yields
phenomenal execution in spite of halfway impediments of
the face, the utilization of glasses, scarves or tops,changesof
outward appearance, and moderatepivotsoftheface.Also,it
doesn't enable clients to be imitated utilizing photos.Tensor
biometric innovation is the ideal answer for control the
entrance of staff to confined security zones.
2. LITERATURE SURVEY:
Ching-Kai Huang and Tsuhan Chen [1] proposed a
technique by recording just video that has significant data,
i.e., video that has movement in the scene. This can be
accomplished with a computerized camcorder furthermore,
a DSP calculation that identifies movement.
Wijnhoven et al [2] considered model-based
element location for traffic reconnaissance, going for object
arrangement. Inside distinguished locales of-intrigue(ROIs)
of moving articles in the scene, the direction of the item is
detected utilizing an angle bearing histogram. For the
purposeful direction,a 3Dwire-outlinemodel isappliedonto
the picture information and the best coordinating pixel-
position is determined inside the item's districts of-intrigue.

2. International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 07 Issue: 01 | Jan 2020 www.irjet.net p-ISSN: 2395-0072
© 2020, IRJET | Impact Factor value: 7.34 | ISO 9001:2008 Certified Journal | Page 738
Stauffer, C et al [3] examined about demonstrating
each pixel as a gathering of Gaussians and utilizations anon-
line estimate to restore the model. The Gaussian
appropriations of the versatile blend model are evaluatedto
figure out which are probably to result from a foundation
process. Every pixel is sorted dependent on the Gaussian
dispersion which speaks to it most productivelyisviewedas
a component of the foundation model..
Rama et al [4] tended to the issue of how to pick the
most good number of sensors and how to choose their area
in a given observed region for MSS.
Pradeep et al [5] suggested a novel strategy which
energetically processes the certainty level of new streams in
view of the reality whether it gives proof which agrees or
repudiates with the effectively confided in streams
Atrey et al [6] displayed a top-down occasion
discovery approach forsoundbasedoccasionrecognition for
reconnaissance. The proposed approachatfirstsortsa given
sound edge into non-vocal and vocal occasions, and next
convey outs advance grouping into energized and typical
occasions.
An ensuing system,W4S [7], incorporated ongoing
sound system to survive the complexities that W4 met with
unexpected brightening changes, shadow and impediment
which make following a lot harder in force pictures.
Wren et al. [8] built up The Pfinder framework
which is utilized to recover a 3-dimensional portrayal of an
individual in a tremendous zone.
Axel Baumann et al [9] gave a precise survey of
quantifies and assesses their adequacyfor explicithighlights
like division, occasion recognition and following.
Ismail Haritaoglu et al [10] proposed a minimal
effort PC based constant visual observation framework,
called W4, for following individuals and their pieces of the
body, and administering their exercises in surround sound
and monochromatic symbolism.Itworksongrayscalejustas
infrared video symbolism
3. PROPOSED SYSTEM:
The system`s architecture propounds the installation of
TensorCam at each entry level. The objective of this system
is the availability of the TensorCam is to capturing and the
comparing for the identification of the authorized and the
unauthorized persons upon the entry into the campus. A
database is connected to the TensorCam in which thedetails
of the persons (studying, working ,visiting the campus )
along with the photographs are stored. Tensorcam captures
each person`s face upon the entry and compares it with the
data present in the database. The architecture includes the
installation of end-end surveillance i.e., upon entering and
leaving the college. And also intermediate cameras are
installed at different levels to check the presence and
location of the respective person. Uponenteringthecampus,
each time the person has to be identified by the Cam i.e.,
whether he is already registered or not. If the person`s data
is already available at the database, he is allowed to enter. If
the person`s data is not available in the database the
computer that is connected to the cam asks for the
registration or creation of the new table. The person who is
identified is highlighted in the green color and the person
who is unidentified is highlighted in red color. Upon the
indication of the person`s face in red highlighted box, they
are entitled to register. It is easy to monitor who is entering
into the campus. Basically there are two main categories in
the database with the person`s details. They are PRIVATE
(Pr) and PUBLIC (Pu). Private consists of students,
employees (working and non- workingstaff).Publicisfor the
people who are visitors of the college. Each data sheet of the
student contains the name, roll no, admission no, class,
department, entry and exit data. Each data sheet of the
employee contains the name, employee id, designation and
the department entry and exit. Visitor`s data sheet contains
the details as name, person to be visited, mobile number,
entry and exit details. A log entry is enlarged upon
requirement. This also includesthefacilityoflogginginto the
website for monitoring the presence of the students by their
parents.
Figure: Architecture of the proposed system.
RESULTS:
Using the proposed system a databasehasbeencreated with
the enrolled students. Appropriate images of the students
have been taken. The camera has to be setup in such a way
that faces can be captured via camera .

3. International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 07 Issue: 01 | Jan 2020 www.irjet.net p-ISSN: 2395-0072
© 2020, IRJET | Impact Factor value: 7.34 | ISO 9001:2008 Certified Journal | Page 739
Figure: Camera identifying the students while entry into
the college.
Figure: Camera identifing students in side the campus.
CONCLUSION & FUTURE WORK:
In current generation security has been a major concern in
most of the sectors especially in Educational institutions
where it constitutes large numberofpeople.Sotheproposed
setup enables the Institution managementaswell asparents
to have surveillance without interrupting one’s privacy. By
utilizing this setup, it emphasizes the identification of
persons very efficiently. The capability of enhancing
situational awareness across multiple scales of time and
space is the significant feature of this system. Addition of an
alarm to the system adds an advanced security when
prohibited persons try to enter the campus. A videotracking
be added to monitor any suspicious activities.
REFERENCES
[1] Ching-Kai Huang and Tsuhan Chen, “Motion Activated
Video Surveillance Using TI DSP”, DSPS FEST '99, Houston,
Texas, August 4-6, 1999.
[2] Wijnhoven, R., de With, P, “3D wire-frame object-
modeling experiments for video surveillance”, Proc. of 27th
Symposium on Information Theory in the Benelux,2006, pp.
101–108.
[3] Stauffer. C, Grimson W.E.L, “Adaptive background
mixture models for real-time tracking”, IEEE Computer
Society Conference on Computer Vision and Pattern
Recognition. Volume 2., Ft. 1999, pp. 252–258.
[4] Rama, K.G.S., Atrey, P.K., Singh, V.K., Ramakrishnan, K.,
Kankanhalli, M.S., “A Design Methodology For Selection and
Placement of Sensors in Multimedia Surveillance Systems”,
The 4th ACM International Workshop on Video Surveillance
and Sensor Networks, Santa Barbara, CA, USA 2006.
[5] P. K. Atrey, Mohan S. Kankanhalli, Abdulmotaleb El
Saddik, “Confidence Building Among Correlated Streams in
Multimedia Surveillance Systems”, MMM 2007, LNCS 4352,
Part II, 2007, pp. 155–164.
[6] Atrey, P.K., Maddage,N.C., Kankanhalli,M.S.,“Audiobased
event detection for multimedia surveillance”, IEEE
International Conference on Acoustics, Speech and Signal
Processing, 2006, pp 813–816.
[7] Ismail Haritaoglu, David Harwood, Larry S. Davis, “W4S:
A Real-Time System for Detecting and Tracking People in 2
½ D”.
[8] Wren, C. R., Azarbayejani, A., Darrell, T., Pentland, A. P.:
Pfinder:real-time tracking of the human body. IEEE
Transactions Pattern Analysis Machine Intelligence, Vol. 19,
No.7,780–785.(1997)
[9] Axel Baumann, MarcoBoltz,Julia Ebling,MatthiasKoenig,
Hartmut S. Loos, Marcel Merkel, Wolfgang Niem, Jan
KarlWarzelhan, Jie Yu “A Review and Comparison of
Measures for Automatic Video Surveillance Systems”,
Hindawi Publishing Corporation EURASIP Journal on Image
and Video Processing Volume 2008, Article ID 824726, 30
pages doi:10.1155/2008/824726.
[10] Ismail Haritaoglu, Harwood, D., Davis,L,“W4:real-time
surveillance of people and their activities”, IEEE
Transactions on Pattern Analysis and Machine Intelligence
(PAMI), vol. 22, 2000, pp. 809–830.