Please use this identifier to cite or link to this item: https://dl.ucsc.cmb.ac.lk/jspui/handle/123456789/4421
Title: Face Recognition System for Examinations
Authors: Wickrama, D.R.N.
Issue Date: 4-Aug-2021
Abstract: Examination or the assessment is the ultimate outcome of the education. Each and every curriculum must be concluded with any kind of assessment. Personal identification in the examination systems of Sri Lanka has been performed by getting the candidate’s identity card and check the relevant details of the candidate. The main issue arises in this scenario is the unique identification of each candidate. This is the major issue arises during the examination period faced by the examiners and their officers in charge of the examinations department. It is a hectic method to use the above-mentioned manual method in recognizing the candidates which is not applicable nowadays. In this dissertation it is suggested the unique identification of the candidate by a system and giving out the details of the candidate without performing manually. The preview image of the candidate has been output through the system with the candidate index number, name and any other relevant details. And also, the exam start time and end time will be noted at the time of recognition. The future system would recognize each student and produce the details such as name in full, Index number required for the exam. Even though they have produced their Identity cards or any other materials to prove their identity there is a possibility to pose a fake candidate on behalf of the real candidate. It is essential to recognize their identity without any failure using a bio metric pattern. This project was implemented as the pilot project for the examination systems in candidate face recognizing process, chosen St. Peter’s College Colombo 4 as the client. Face recognition system initially detects the candidate’s face. The detected face later trains and stores inside the file folder. Training process is the most crucial event in this system. If the training process is not successful face detection will fail to perform. During detection time the system allows the candidates to insert the student name and his/her index number of the examination. Finally, the candidate will be able to scan the face prior to the exam. If the recognition is successful, student name and index number will display during the recognition process and a confirmation quote ‘Your Face Recognized’ also displays. Otherwise it displays ‘Your Face not Recognized’. OpenCV/Python suite is being used for developing this face recognition system. The object oriented and structural patterns have been used.CV2, PIL, NumPy,Pickle,OS libraries of image processing used in the face detection, face training and face recognition processes. SQLite is the database management system, stored all data of the candidates. The face recognition system developed from scratch as the stand-alone system in the windows environment. Around fifty to sixty users participated in the evaluation process. The goals of the project achieved successfully. According to the User evaluation feedback, implemented face recognition system for examinations fulfills the requirements of the examination management system in Sri Lanka.
URI: http://dl.ucsc.cmb.ac.lk/jspui/handle/123456789/4421
Appears in Collections:2019

Files in This Item:
File Description SizeFormat 
2016MIT087.pdf750.87 kBAdobe PDFView/Open


Items in UCSC Digital Library are protected by copyright, with all rights reserved, unless otherwise indicated.