Face Recognition Based Automated Student Attendance System

Abstract

The Oxford Dictionary defines a face because the a part of a man or woman\'s head from the brow to the chin. In human interactions, the face is the maximum giant issue as it incorporates vital facts approximately an person. All people will well known human beings from their faces. The proposed answer is to broaden an running prototype of a gadget that could facilitate elegance attendance control for the academics withinside the school rooms through detecting the faces of students from an photograph taken in a school room. The database can shop the faces of students, as soon as the face of the person fits with one in all of the faces held withinside the database then the attendance is recorded. In latest years, evaluation has been allotted and face reputation and detection structures are advanced. A wide variety of those are used on social media structures like Facebook, Banking apps, authorities offices, etc. In today’s aggressive world, with very much less school room time and growing operating hours, teachers can also additionally want equipment that may assist them to manipulate treasured elegance hours efficiently. Instead of that specialize in teaching, teachers are caught with finishing a few formal duties, like taking attendance, keeping the attendance document of every scholar, etc. Manual attendance marking unnecessarily consumes school room time, while clever attendance via face reputation strategies enables in saving the school room time of the lecturer. Attendance marking via face reputation may be implied with inside the school room through shooting the photograph of the scholars with inside the school room thru the digital digicam installed.

Introduction

I. INTRODUCTION

Maintenance of attendance is noticeably essential altogether on the institutes for checking the overall performance of college students. Some institutes take attending manually the use of paper or file-primarily based totally procedures and some have followed methods of automated attendance the use of a few biometric strategies. The latest method for taking attendance is through calling out the call or rolls wide variety of the pupil to document attendance. It's an extended and much less green approach of marking attending because of as all of us recognize the data written withinside the paper usually can be misplaced or is much less correct because of college students of resultant mark each other's attending proxy. Therefore, to get to the bottom of those problems and keep away from errors, we suggest computerizing this approach through offering a gadget that document and manipulate college students attending robotically without a want for lecturer’s interference.

Every biometric gadget includes the enrollment method in the course of which unique functions of a person are saved withinside the database after which there are methods of identity and verification. These 2 methods evaluate the biometric characteristic of a person with formerly saved biometric info captured on the time of enrollment.  Biometric templates are of many types like Fingerprints, Eye Iris, Face, Signature, and Voice. Our gadget makes use of the face reputation method for the Automated Attendance Management System.  By thinking about this fact our gadget goes to be faster and accurate in marking the attendance of person college students.  Face reputation includes steps, withinside the first step faces are detected withinside the photograph after which those detected faces are as compared with the database for verification.  Face detection is hired to locate the location of face area and face reputation is hired for marking the attendance.  The data can shop the faces of students.  Once the face of the pupil fits with one in every of the faces saved withinside the database then the attendance is recorded.  Various robust algorithms are advanced that provide correct overall performance to address face detection and reputation troubles. These algorithms or strategies are the maximum effectively and broadly used for face detection and reputation programs.

PURPOSE OF THE SYSTEM

Instead of the use of the traditional strategies, this proposed gadget targets to broaden an automatic gadget that information the scholar's attendance through the use of facial reputation technology. The foremost goal of this paintings is to make the attendance marking and control gadget green, time saving, easy and easy.

II. SYSTEM ANALYSIS

A. EXISTING SYSTEM

In current gadget, real-time disadvantage of manually taking attendance through line the call or roll wide variety of the scholar to document attendance. It changed into an extended and lesser green approach of marking attendance due to the fact as we recognize the facts written withinside the paper may be misplaced or may be much less correct due to the fact college students frequently mark every other's attendance proxy.

B. Proposed System

The foremost goal of this venture is to get to the bottom of the troubles encountered withinside the preceding attending gadget while reproducing a sparkling revolutionary realistic gadget that could provide comfort to the institution. In this venture, a realistic tool could be advanced this is able to spotting the identification of each person and in the end document down the data right into a database gadget. Apart from that, an interface could be advanced to deliver visible get entry to to the facts methods.

III. DEVELOPMENT ENVIRONMENT

A. Hardware Requirements

1. RAM : eight GB Ram
2. Processor : Intel i5 Processor or More
3. Hard Disk : 1TB

B. Software Requirements

1. Front End : Jupyter Notebook with Python Idle
2. Operating gadget : Windows10
3. Platform : Anaconda Navigator
4. Backend : Machine Learning
5. Framework : Tensorflow , skikit learn

IV. MODULE DESCRIPTION

A. Parameters

1. Radius: Generally 1 is about as a radius for the round neighborhood binary sample which denotes the radius across the relevant pixel.
2. Neighbours: The wide variety of pattern factors surrounding the relevant pixel that's typically eight. The computational price will boom with boom in wide variety of pattern factors.
3. Grid X: The wide variety of cells alongside the horizontal route is represented as Grid X. With the boom in wide variety of cells the grid will become finer which ends up in boom of dimensional characteristic vector.
4. Grid Y: The wide variety of cells alongside the vertical route is represented as Grid Y. With the boom in wide variety of cells the grid will become finer which ends up in boom of dimensional characteristic vector.

B. Algorithm Training

For the education motive of the dataset of the facial photos of the human beings to be identified together with the particular ID is needed in order that the offered method will make use of the furnished facts for perceiving an enter photograph and offering the output. Same photos require equal ID.

C. Computation Of The Algorithm

The intermediate photograph with advanced facial traits which corresponds to the unique photograph is created withinside the first step. Based at the parameters furnished, sliding window concept is used to be able to achieve. So facial photograph is transformed into grey scale. A 3x3 pixels window is taken which also can be expressed as a 3x3 matrix which incorporates the depth of every pixel (zero-255). After this we do not forget the relevant cost of the matrix which we take as the brink. This cost defines the brand new values acquired from the eight neighbours. A new binary cost is about for every neighbour of the relevant cost. For the values identical to or extra than the brink cost I could be the output in any other case zero could be the output. Only binary values could be gift withinside the matrix and the concatenation is accomplished at every role to get new values at every role.

Then the conversion of this binary cost right into a decimal cost is accomplished that's made the relevant cost of the matrix. It is a pixel of the real photograph. As the method is completed, we get a brand new photograph which serves because the higher traits of the unique photograph.

D. Extraction Of Histogram

The photograph acquired withinside the preceding step makes use of the Grid X and Grid Y parameters and the photograph is break up into more than one grids. Based at the photograph the histogram may be extracted as below:

1. The photograph is in grey scale and every histogram will encompass simplest 256 positions (zero-255) which symbolises the existences of every pixel depth.
2. After this every histogram is created and a brand new and larger histogram is accomplished. Let us think that there are 8x8 grids, then there could be 16.384 positions in general withinside the very last histogram. Ultimately the histogram indicates the functions of the real photograph.

E. Face Recognition

The education of the set of rules is accomplished. For locating the photograph that's equal because the enter photograph, the 2 histograms are as compared and the photograph similar to the closest histogram is returned. Different procedures are used for the calculation of distance among the 2 histograms. Here we use the Euclidean distance primarily based totally at the formula. Hence the end result of this approach is the ID of the photograph which has the closest histogram. It must go back the gap calculated withinside the shape of 'self assurance. Then the brink and the 'self assurance' may be used to routinely examine if the photograph is successfully identified. If the self assurance is much less than the given threshold cost, it means that the photograph has been properly identified through the set of rules.

VI. FUTURE ENHANCEMENT

In destiny paintings, we want to enhance face detection effectiveness with the assist of the interplay amongst our gadget, the scholars, and the faculty. On the alternative hand, our gadget is advanced through desegregation video-streaming provider and lecture archiving gadget, to present extra profound programs withinside the area of distance education, path control gadget (CMS), and guide for university development (FD). We will enhance this method as a result as we have a tendency to run this gadget with extra than college students on a bench and letting them range their positions.

Conclusion

The complete venture has been advanced from the necessities to a entire gadget along assessment and testing. The gadget advanced has executed its goal and objectives. More cautious evaluation is needed on a venture intrinsically. The methods used can be mixed with others to acquire first-class results. Completely specific methods are enforced withinside the beyond in step with the literature review. The end to set the parameters of this a part of the gadget primarily based totally on a totally small elegance length changed into because of the disasters acquired from the popularity a part of the gadget. The length of the photograph may be very vital in face reputation as each pixel counts. The set of rules we have got used might be analyzed with photos of various sizes and those photos might be displaying college students in a school room putting with herbal sitting positions displaying faces of various sizes. The primary concept is, regardless of how the face is turned, we must be capable of middle the eyes and mouth in kind of the equal role withinside the photograph, as a result spotting the man or woman and staining their attendance.

References

[1] Setia Budi, Oscar Karnalim, Erico D. Handoyo, Sulaeman Santoso, Hapnes Toba, Huyen Nguyen, Vishv Malhotra (2018), “A Low Cost Solution to Record Student Attendance in a Classroom”, IEEE International Symposium on Multimedia (ISM). [2] Hidayat, M. A., &Simalango, H. M. (2018), “Students Attendance System and Notification of College Subject Schedule Based on Classroom Using IBeacon”, 3rd International Conference on Information Technology, Information System and Electrical Engineering (ICITISEE). [3] Omar Abdul Rhman Salim, Rashidah Funke Olanrewaju, Wasiu Adebayo Balogun (2018), “Class Attendance Management System Using Face Recognition”, IEEE (7th International Conference on Computer and Communication Engineering (ICCCE)), pp. 93-98. [4] N.Sudhakar Reddy, M.V.Sumanth, S.Suresh Babu (2018), “A Counterpart Approach to Attendance and Feedback System using Machine Learning Techniques”, Journal of Emerging Technology and Innovative Research (JETIR), Volume 5, Issue 12, pp 190-193. [5] Jalendu Dhamija, Tanupriya Choudhury, Praveen Kumar, Yogesh Singh Rathore (2017), “An advancement towards efficient Face Recognition using Live video feed”, International Conference on Computational Intelligence and Networks (ICCIN). [6] Aayush Mittal, Fatima Sartaj Khan, Praveen Kumar, Tanupriya Choudhury (2017), “Cloud Based Intelligent Attendance System through Video Streaming”, IEEE International Conference On Smart Technology for Smart Nation (Smartechcon). [7] Rekha.E, Dr.Ramprasad.P (2017), “An Efficient Automated Attendance Management System Based on Eigen Face Recognition”, IEEE (International Conference on Cloud Computing, Data Science & Engineering – Confluence). [8] Bhumika Agrawal, Chelsi Gupta, Meghna Mandloi, Divya Dwivedi, Jayesh Surana (2017), “GPU Based Face Recognition System for Authentication”, International Journal of Engineering Development and Research (IJEDR), Volume 5, Issue 2. [9] Ishaan Sathe, Chiman Patel, Prasad Mahajan, Tanmay Telang, Sejal Shah (2017), “Automatic Locking Door Using Face Recognition”, International Journal of Engineering Technology Science and Research (IJETSR), Volume 4, Issue 4. [10] Aesha Shah, Kavin Shah, Vidhi Shah, Chintan Shah (2017), “Built-in Face Recognition for Smart Phone Devices”, International Research Journal of Engineering and Technology (IRJET), Volume 04, Issue 04.

Copyright

Copyright © 2022 K. Geetha, Dr. J. Sreerambabu, N. Santhosh . This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.