Development Of Automated Attendance System Based On Facial Recognition

Kanagasabai, Aravindan (2018) Development Of Automated Attendance System Based On Facial Recognition. Project Report. UTeM, Melaka, Malaysia. (Submitted)

Text (24 Pages) Development Of Automated Attendance System Based On Facial Recognition.pdf - Submitted Version Download (608kB)

Abstract

This Project is stress on the development of automated attendance system based on facial recognition using the principle component analysis (PCA) method with the Eigenface approach. In the past, student’s attendance management system is an important task done by every institution in order to maintain the student’s academic results and this project able to reduce the chances of forgery attendance marking. The main objective to be achieved is the development of face recognition system based on Eigenface approach using the PCA algorithm which includes the face detection part using the Viola Jones method of detection. To achieve this objective, the MATLAB software build in algorithm such as image processing toolbox and image acquition toolbox is utilized. Using this algorithm, the expected result from this system is the collection of student database through capturing the image of the students for face recognition process which will detect the face of the student from the whole image eliminating background and other elements from the image and save it in the database and ready for face recognition process which will give a message to indicate that the person is recognized or not. The system performance will be analysing based on the accuracy under different lighting condition, varying in number of person and varying the distance between the student and the webcam. The analysis shows that the performance decreases as the number of person in an image is increased. Although the performance decreases, the accuracy obtain for 5 person in an image records a percentage of greater than 80%. Besides that, the analysis results for varying the distance from webcam shows that the increase in the distance between webcam and the students effect the performance of the system where for distance 3 feet the system records a 94% of accuracy but for distance 9 feet records 0% of accuracy. Lastly, the lighting effect under which the image is capture does not affect the systems performance which records 100% of accuracy for both the lighting condition. Therefore, this system perform very well in face recognition which can be used for attendance recording system.

Actions (login required)

View Item