IMPLEMENTATION PAPER ON MACHINE LEARNING BASED SECURITY SYSTEM FOR OFFICE PREMISES
0 likes12 views
1. The document discusses the implementation of a machine learning-based security system for office premises using user authentication. 2. The proposed system uses four-step security including login credentials, one-time passwords, and face recognition to authenticate users and restrict unauthorized access, while also featuring auto-saving of data to servers and automatic logouts. 3. The system aims to provide strong security, integrity, and confidentiality of data by making unauthorized access more difficult through multi-factor authentication barriers.
![International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 09 Issue: 04 | Apr 2022 www.irjet.net p-ISSN: 2395-0072
© 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 3943
IMPLEMENTATION PAPER ON MACHINE LEARNING BASED SECURITY
SYSTEM FOR OFFICE PREMISES
Rutuja Patil, Adarsh Charak, Utkarsha Bhad, Sachin Deshmukh
Rutuja Patil Undergraduate Student JSPM’s RSCOE, Maharashtra, INDIA Pune – 411033
Sachin Deshmukh Undergraduate Student JSPM’s RSCOE, Maharashtra, INDIA Pune – 411033
Utkarsha Bhad Undergraduate Student JSPM’s RSCOE, Maharashtra, INDIA Pune – 411033
Adarsh Charak Undergraduate Student JSPM’s RSCOE, Maharashtra, INDIA Pune – 411033
3Example: Prof. Vinodkumar Bhutanal, Dept. of Computer Engineering, JSPM’S RSCOE, Maharashtra, India
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - Safety plays a major role in today’s world. User-authentication can be defined as a procedure used to verify the
identity of the user , authentication is an essential term in the digital world so as to protect the client’s/user’s critical data
stored online or offline. Various mechanisms are gradually evolving to authenticate users in various scenarios. At offices and
work places as well, this user authentication is of utmost importance in order to avoid any mal-practices if any as well as to
maintain individual work privacy in the organization. This project aims at developing a user authentication system for offices
based on their company generated login ID and passwords, One-Time-Password (OTP) generation and face recognition. The
system also has features such as auto-saving data to server and auto-logout
Key Words: AutoSaved, Auto-logout, Face recognition, Login ID and passwords credentials, One-Time-Password
(OTP).
1. INTRODUCTION
As of today, computers are protected with various security software, proxy servers,firewalls,still thedata insidethecomputer
has vulnerabilities. We aim in providing 4 step security so that the data is not accessed illegally. This approach gives more
security by restricting unauthorized access and auto logout feature. Wecanguarantee integrityandconfidentiality. Onestrong
pillar that is one-time password makes it more difficult to get illegal access to data.
The authentication, confidentiality and privacy of data is needed in today’s world. IOT that is the internet of things is evolving
widely. The IoT architecture is just modified version of some existing data communication tools. Due to the vulnerabilities of
the Internet, security and privacy issues should be considered and resolved before the IOT is globally deployed. Direct
interaction of smart devices within the immediate living space of humans intimidates new security vulnerabilities. Research
has been led in developing customized tools for computer securityto establishconfidentiality,integrityandavailability.Incase
of any kind of security failure, our system provides the users data fully data security and assurance of privacy. The aim is to
prevention system that is adaptive and receptive to new threats and provides more control of owner on the data stored on
system by restricting the access to particular user for specific file with limited rights. The idea behind ittocreatesoftwarethat
will protect the data from all kinds of attacks and maintain entire confidentiality of the data.
2. LITERATURE REVIEW
1 There has been a use of three tire architecture in paper“EffectiveAuthenticationForRestricingUnauthorizedUser”,[1]for
providing security to the system. Those tiers are Facial recognition, Fingerprint Scanning and OTP. All these tiers give a
great strength to security of system. Incase if there is an attack on system, the measures are provided so as to not leak the
important data from the system. By using AES 512 all the data is encrypted which can only be decrypted by a specific key.
2 Security and protection of personal data online/offline is very critical as data is becoming critical and smartphones are
vulnerable to illegal unauthorized access. User-authentication allows legitimate person to ask the data. Mobiles are made
secured by using the methodology of biometric authentication using user’s physical traits. The survey provides an
overview of the current state-of-the-art approaches for continuous user authentication using behavioral biometrics
captured by smartphones’ embedded sensors, including insights and open challenges for adoption, usability, and
performance.](https://image.slidesharecdn.com/irjet-v9i4578-220928075702-959ce3a3/85/IMPLEMENTATION-PAPER-ON-MACHINE-LEARNING-BASED-SECURITY-SYSTEM-FOR-OFFICE-PREMISES-1-320.jpg)
![International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 09 Issue: 04 | Apr 2022 www.irjet.net p-ISSN: 2395-0072
© 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 3944
3 Biometric User-authentication of a user by using their own unique traits is the most easiest andcommonwaytoidentifya
person. In [3], a multimodal-biometric-user-verification system with similar twin shows the fingerprint, face and lip
classification model using Support Vector Machine.
4 Paper – ‘Multi-Biometric Authentication Using Deep Learning Classifier for Securing of Healthcaredata;[4]itanalysesthe
results of using both fingerprints as well as online signatures. Here the signatures are verified using the dynamic time
warping methodology. The weighted sum rule is applied for the process of biometric fusion (signatures andfingerprints).
Deep learning classifier is proposed in this work for multi-biometrics authentication. When a biometric authentication
request is submitted to improve the authentication performance, the proposed user-authentication system uses deep-
learning to automatically select an accurate matching image.
5 Now-a-days we can see the concept of cloud evolving having benefits like high performance, accessibility, better storing
capacity etc. Storage has become a huge issue for users having large data, cloud provides a solutionofunlimitedstorageto
a user. Yet cloud has various vulnerabilities and users do not wish to compromise security. So the aim is to figure out and
address the security issues by using cross validation strategy.
6 User-Authentication is always done at the start of the program, but in order to understand the genuinenessof theuserthe
authentication must be continuously going on. In [6], Suhail Javed QuraishiandSarabjeetSinghBediproposedthe usageof
key-stroke used as biometric characteristic for continuously authenticating the user. Authenticationusingbiometricuses
basically three stages such as the enrollment stage, the verification stage, and the identification stage. Identification stage
marks the user as an authenticated user only if the input pattern is matched with the profilepatternor elsethesystemwill
automatically logout. There is no restriction on input data during enrolment, verification, and identification stage. The
technique used to classify the authenticated user’s input from others isUnsupervised One-classSupportVector.ThisUser-
authentication system can be used in areas like in non-proctored online-examination, Intrusion and Fraud Detecting
Systems, etc.
7 Paper “A Robust e-Invigilation - System Employing Multimodal Biometric Authentication”, presents the development of
robust, flexible, transparent and continuous authentication mechanism for e-assessments. There has been a significant
growth of students/users using e-learnings and e-assessments. The major concern is that they might cheat using these
resources in their exams. So in order to reduce the chances of cheating in online examinations, biometrics of the student
can be used to check whether the legitimate student is attempting the exam the system gives a continuous user-
identification by using biometrics;: an eye tracker is used to record the student’s eye movement; and, the speech -
recognition is used to detect communication from student’s end. The focus of [7] particularly is on the development 3D
Facial-Authentication. Experiments have also been conducted in order to test these biometricstosee whetherthestudent
is cheating During the experiment, participants' biometric-data,eye-movement,andhead-movementshavebeencollected
using a software.
8 As of today, computers are protected with various security software, proxy servers, firewalls, still the data inside the
computer has vulnerabilities. We aim in providing 4 step security so that the data is not accessed illegally. This approach
gives more security by restricting unauthorized access and auto logout feature. We can guarantee integrity and
confidentiality. One strong pillar that is one-time password makes it more difficult to get illegal access to data.
9 Paper “Effective Authentication for Avoiding Unauthorized User Access”, presented a evolved Role Based Access Control
model by modifying previous role based access control in Structure Query Language. This model evaluates & executes
security policies that contain access conditions against the dynamic nature of data. The goal is to create or use a
mechanism for forward looking, assertive and flexible security-features to regulateaccesstodata inthedata storage.This
is easily achieved by integrating roles & authenticated access rules and implemented through effective audit trail.
10 Paper [10] presented a state of art about biometric hand, different techniques used. Biometric is essentially used to avoid
risks of password easy to find or Stoll; with as slogan save Time and Attendance. The measurement andanalysisofunique
physical or behavioral characteristics (such as fingerprint or voice patterns) especially as a means of verifying personal
identity.. Biometrics is recognizing a person based on his physical unique characteristics like fingerprint, iris recognition,
face recognition etc. This methodology providesusmoreuniquenessandsecurity.Asthesecharacteristicscannotbe faked,
the chances of the computer system having biometricsgettingaccessedillegallyarenegligible.Traditionallythebiometrics
meant signatures, keys and passwords. Biometrics is defined as ‘Life Measure’ in Greek dictionary. Biometrics is a very
useful method which is widely used in criminal identification and prison security. These can be used to protect the ATMs,
phones, systems, smartcards etc. Thus, Biometric is a widely evolving technology which is very efficient for user
authentication and security.](https://image.slidesharecdn.com/irjet-v9i4578-220928075702-959ce3a3/85/IMPLEMENTATION-PAPER-ON-MACHINE-LEARNING-BASED-SECURITY-SYSTEM-FOR-OFFICE-PREMISES-2-320.jpg)
IMPLEMENTATION PAPER ON MACHINE LEARNING BASED SECURITY SYSTEM FOR OFFICE PREMISES
- 1. International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 09 Issue: 04 | Apr 2022 www.irjet.net p-ISSN: 2395-0072 © 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 3943 IMPLEMENTATION PAPER ON MACHINE LEARNING BASED SECURITY SYSTEM FOR OFFICE PREMISES Rutuja Patil, Adarsh Charak, Utkarsha Bhad, Sachin Deshmukh Rutuja Patil Undergraduate Student JSPM’s RSCOE, Maharashtra, INDIA Pune – 411033 Sachin Deshmukh Undergraduate Student JSPM’s RSCOE, Maharashtra, INDIA Pune – 411033 Utkarsha Bhad Undergraduate Student JSPM’s RSCOE, Maharashtra, INDIA Pune – 411033 Adarsh Charak Undergraduate Student JSPM’s RSCOE, Maharashtra, INDIA Pune – 411033 3Example: Prof. Vinodkumar Bhutanal, Dept. of Computer Engineering, JSPM’S RSCOE, Maharashtra, India ---------------------------------------------------------------------***--------------------------------------------------------------------- Abstract - Safety plays a major role in today’s world. User-authentication can be defined as a procedure used to verify the identity of the user , authentication is an essential term in the digital world so as to protect the client’s/user’s critical data stored online or offline. Various mechanisms are gradually evolving to authenticate users in various scenarios. At offices and work places as well, this user authentication is of utmost importance in order to avoid any mal-practices if any as well as to maintain individual work privacy in the organization. This project aims at developing a user authentication system for offices based on their company generated login ID and passwords, One-Time-Password (OTP) generation and face recognition. The system also has features such as auto-saving data to server and auto-logout Key Words: AutoSaved, Auto-logout, Face recognition, Login ID and passwords credentials, One-Time-Password (OTP). 1. INTRODUCTION As of today, computers are protected with various security software, proxy servers,firewalls,still thedata insidethecomputer has vulnerabilities. We aim in providing 4 step security so that the data is not accessed illegally. This approach gives more security by restricting unauthorized access and auto logout feature. Wecanguarantee integrityandconfidentiality. Onestrong pillar that is one-time password makes it more difficult to get illegal access to data. The authentication, confidentiality and privacy of data is needed in today’s world. IOT that is the internet of things is evolving widely. The IoT architecture is just modified version of some existing data communication tools. Due to the vulnerabilities of the Internet, security and privacy issues should be considered and resolved before the IOT is globally deployed. Direct interaction of smart devices within the immediate living space of humans intimidates new security vulnerabilities. Research has been led in developing customized tools for computer securityto establishconfidentiality,integrityandavailability.Incase of any kind of security failure, our system provides the users data fully data security and assurance of privacy. The aim is to prevention system that is adaptive and receptive to new threats and provides more control of owner on the data stored on system by restricting the access to particular user for specific file with limited rights. The idea behind ittocreatesoftwarethat will protect the data from all kinds of attacks and maintain entire confidentiality of the data. 2. LITERATURE REVIEW 1 There has been a use of three tire architecture in paper“EffectiveAuthenticationForRestricingUnauthorizedUser”,[1]for providing security to the system. Those tiers are Facial recognition, Fingerprint Scanning and OTP. All these tiers give a great strength to security of system. Incase if there is an attack on system, the measures are provided so as to not leak the important data from the system. By using AES 512 all the data is encrypted which can only be decrypted by a specific key. 2 Security and protection of personal data online/offline is very critical as data is becoming critical and smartphones are vulnerable to illegal unauthorized access. User-authentication allows legitimate person to ask the data. Mobiles are made secured by using the methodology of biometric authentication using user’s physical traits. The survey provides an overview of the current state-of-the-art approaches for continuous user authentication using behavioral biometrics captured by smartphones’ embedded sensors, including insights and open challenges for adoption, usability, and performance.
- 2. International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 09 Issue: 04 | Apr 2022 www.irjet.net p-ISSN: 2395-0072 © 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 3944 3 Biometric User-authentication of a user by using their own unique traits is the most easiest andcommonwaytoidentifya person. In [3], a multimodal-biometric-user-verification system with similar twin shows the fingerprint, face and lip classification model using Support Vector Machine. 4 Paper – ‘Multi-Biometric Authentication Using Deep Learning Classifier for Securing of Healthcaredata;[4]itanalysesthe results of using both fingerprints as well as online signatures. Here the signatures are verified using the dynamic time warping methodology. The weighted sum rule is applied for the process of biometric fusion (signatures andfingerprints). Deep learning classifier is proposed in this work for multi-biometrics authentication. When a biometric authentication request is submitted to improve the authentication performance, the proposed user-authentication system uses deep- learning to automatically select an accurate matching image. 5 Now-a-days we can see the concept of cloud evolving having benefits like high performance, accessibility, better storing capacity etc. Storage has become a huge issue for users having large data, cloud provides a solutionofunlimitedstorageto a user. Yet cloud has various vulnerabilities and users do not wish to compromise security. So the aim is to figure out and address the security issues by using cross validation strategy. 6 User-Authentication is always done at the start of the program, but in order to understand the genuinenessof theuserthe authentication must be continuously going on. In [6], Suhail Javed QuraishiandSarabjeetSinghBediproposedthe usageof key-stroke used as biometric characteristic for continuously authenticating the user. Authenticationusingbiometricuses basically three stages such as the enrollment stage, the verification stage, and the identification stage. Identification stage marks the user as an authenticated user only if the input pattern is matched with the profilepatternor elsethesystemwill automatically logout. There is no restriction on input data during enrolment, verification, and identification stage. The technique used to classify the authenticated user’s input from others isUnsupervised One-classSupportVector.ThisUser- authentication system can be used in areas like in non-proctored online-examination, Intrusion and Fraud Detecting Systems, etc. 7 Paper “A Robust e-Invigilation - System Employing Multimodal Biometric Authentication”, presents the development of robust, flexible, transparent and continuous authentication mechanism for e-assessments. There has been a significant growth of students/users using e-learnings and e-assessments. The major concern is that they might cheat using these resources in their exams. So in order to reduce the chances of cheating in online examinations, biometrics of the student can be used to check whether the legitimate student is attempting the exam the system gives a continuous user- identification by using biometrics;: an eye tracker is used to record the student’s eye movement; and, the speech - recognition is used to detect communication from student’s end. The focus of [7] particularly is on the development 3D Facial-Authentication. Experiments have also been conducted in order to test these biometricstosee whetherthestudent is cheating During the experiment, participants' biometric-data,eye-movement,andhead-movementshavebeencollected using a software. 8 As of today, computers are protected with various security software, proxy servers, firewalls, still the data inside the computer has vulnerabilities. We aim in providing 4 step security so that the data is not accessed illegally. This approach gives more security by restricting unauthorized access and auto logout feature. We can guarantee integrity and confidentiality. One strong pillar that is one-time password makes it more difficult to get illegal access to data. 9 Paper “Effective Authentication for Avoiding Unauthorized User Access”, presented a evolved Role Based Access Control model by modifying previous role based access control in Structure Query Language. This model evaluates & executes security policies that contain access conditions against the dynamic nature of data. The goal is to create or use a mechanism for forward looking, assertive and flexible security-features to regulateaccesstodata inthedata storage.This is easily achieved by integrating roles & authenticated access rules and implemented through effective audit trail. 10 Paper [10] presented a state of art about biometric hand, different techniques used. Biometric is essentially used to avoid risks of password easy to find or Stoll; with as slogan save Time and Attendance. The measurement andanalysisofunique physical or behavioral characteristics (such as fingerprint or voice patterns) especially as a means of verifying personal identity.. Biometrics is recognizing a person based on his physical unique characteristics like fingerprint, iris recognition, face recognition etc. This methodology providesusmoreuniquenessandsecurity.Asthesecharacteristicscannotbe faked, the chances of the computer system having biometricsgettingaccessedillegallyarenegligible.Traditionallythebiometrics meant signatures, keys and passwords. Biometrics is defined as ‘Life Measure’ in Greek dictionary. Biometrics is a very useful method which is widely used in criminal identification and prison security. These can be used to protect the ATMs, phones, systems, smartcards etc. Thus, Biometric is a widely evolving technology which is very efficient for user authentication and security.
- 3. International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 09 Issue: 04 | Apr 2022 www.irjet.net p-ISSN: 2395-0072 © 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 3945 3. PROPOSED SYSTEM A face recognition system is a computer application for automatically identifyingorverifyinga personfroma digital image ora video frame from a video source. For the enhancement of security of accounts and privacy of users features like face recognition and One-Time Password (OTP) are used. Face recognition technologyhelpsthemachinetoidentifyeachandevery user uniquely. Fig 1 illustrates the system architecture of proposed system. It consists of a three-step secure authentication process for working employees in their respective organizations. Fig 1: System Architecture of Proposed System As stated above, at first the user needs to enter the unique login Id and password combination provided to them by the company. Once that matches with the database, next step is to generate the OTP on the registered mobile number. If not, further access to login into the system will be denied. If correct generated OTP is entered into the system, the user moves second step ahead to gain access to the system. Once this is done, next comes the face recognition. With the help of webcams present with each and every system, the face in front of cam is detected. In the process of continuous authentication, a live picture of the user is captured through webcam and it is then compared with the image in database.Ifthisimagematches,user can get the access to operate that particular system. Haar Cascade Classifier is used to implement face recognition. An additional feature of autosave and auto-logout is also included in this system. Suppose the employee happens to leave their desk due to some reasons, the webcam will be continuously detecting and capturing the images ifany.Incaseanunknownface image other than that system user, is captured by the webcam, the whole work/data will be automatically saved to the company’s server and the system will automatically logout, so that, the unknown user cannot make any manipulations in the users work. The autosave feature helps the user to re-continue the work form the point he/she had left it. Hence a very efficient, safe and reliable system is developed. I. RESULTS Fig. 2 sign_up_page
- 4. International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 09 Issue: 04 | Apr 2022 www.irjet.net p-ISSN: 2395-0072 © 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 3946 Fig. 3. Login Pages Fig. 4 Home Fig. 5. Enter OTP Fig. 6 FINAL_PAGE
- 5. International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 09 Issue: 04 | Apr 2022 www.irjet.net p-ISSN: 2395-0072 © 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 3947 4. CONCLUSIONS We have used 4 steps for giving the accurate security and authentication purpose. These steps include – 1.credentials for logging in 2. Facial recognition using HAAR-CASCADE 3.One Time Password and 4. Auto-Logout. These are the four pillars of our system which give proper authentication and security.. This approach gives more security by restricting unauthorized access and auto logout feature. We can guarantee integrity and confidentiality. One strong pillar that is one-time password makes it more difficult to get illegal access to data REFERENCES 1. Patil, A., Rana, D., Vichare, S., & Raut, C. (2018). Effective Authentication for Restricing Unauthorized User. 2018 International Conference on Smart City and Emerging Technology (ICSCET). doi:10.1109/icscet.2018.8537323 2. IEEE INTERNET OF THINGS JOURNAL (IOTJ) 21 Sensor-based; Continuous Authentication of Smartphones’ Users Using Behavioral Biometrics: A Contemporary Survey Mohammed Abuhamad, AhmedAbusnaina,DaeHunNyang,and DavidMohaisen arXiv:2001.08578v210 on May 2020 3. Multimodal-Biometric-User-Verification-System with Similar Twin using SVM 2 B.Lakshmi priya, M.PushpaRani International Journal of Recent Technology and Engineering (IJRTE) ISSN:2277-3878,Volume-8Issue-6,march2020 4. Multi-Biometric User-Authentication Technique Using Deep Learning Classifier for Securing of Healthcare Data Dr. GandhimathiAmirthalingam1 , Harrin Thangavel1 Volume 8, No.4, July – August-2019 5. Cloud-Security: To Prevent Unauthorized Access Using an Eeficient Key-Management Authentication Algorithm S.NaveenKumar1, K.1Nirmala2 International Journal of Engineering & Technology, (IJET) - ; 7 (1.1) 2018 On the key-strokes as Continuous User Biometric Authentication Mr.SuhailQuraishi, SarabjeetsinghBedi at International Journal of Engineering and Advanced Technology IJEAT ISSN: 2249 – 8958,, August-2019