International Research Journal of Engineering and Technology (IRJET) Volume: 09 Issue: 10 | Oct 2022
www.irjet.net
e-ISSN: 2395-0056 p-ISSN: 2395-0072
Face detection based attendance system Md. Ruhen Hossain Bhuiyan, Jebin Ferdousi Department of Electrical and Computer Engineering, North South University, Dhaka, Bangladesh ---------------------------------------------------------------------------***--------------------------------------------------------------------------became higher than ever before. Fourthly, it has the Abstract—The ID card attendance system used in advantage of full automation instead of manual recognition. Fifthly, you need to forget the time fraud. No more buddy favors, since everyone now has to pass a face scanning devices to check-in. Sixthly, facial recognition attendance system can accurately track time and attendance without human error. Seventhly, removes the risk of manual errors. Eightly, it saves time by instantly eliminating the hassle of swiping cards or waving badges around. Ninthly, facial recognition attendance system can moreover hold data that allows the school to study and keep an eye on their students’ statistics and any added reports which are required.
schools in Bangladesh today requires deep learning and neural network techniques more than ever before. The aim of the study is to demonstrate that the Haar wavelet neural network (HWNN) that uses the Harr wavelet as an active function with the help of deep learning can be used in school attendance systems as a facial recognition attendance system. The result shows that the student’s face needs to be identified, and then the face needs to be used as an attendance. The purpose of this research is to examine and critically evaluate recent attendance marking techniques using facial recognition methods. The literature review shows that the intelligent implementation of facial recognition techniques can make attendance management systems more effective. In this paper, through facial recognition, we suggest an ideal model for an automated attendance system.
II. LITERATURE REVIEW A number of techniques for face detection were suggested, i.e. AdaBoost algorithm, FloatBoost algorithm, Support Vector Machines (SVM), Viola Jones detection algorithm and Bayes classification. With the quick face detection algorithm, the effectiveness of the face recognition algorithm can be improved. Some of the previous face recognition methods were appearancebased methods that use texture characteristics applied to the whole face. Some of the other techniques for prior face recognition were feature-based that utilizes geometric features such as mouth, nose, eyes, eye brows, cheeks, and their relationship. After the implementation of the historical igenface method, the research of facial recognition became popular in the early 1990s. In 2014, DeepFace accomplished state-of-the-art precision on the renowned LFW benchmark, for the first time approaching human output on unconstrained situation, by training a 9 layer system on 4 million facial pictures. Inspired by this job, the focus of studies has moved to deep-learning methods, and in just three years the precision has dramatically increased. Deep learning has transformed facial recognition’s study landscape into almost all elements such as algorithm models, training/test information sets, implementation scenarios and even assessment protocols. Our research indicates that the attendance can be registered more effectively by applying the real-time face recognition attendance system. The suggested system will automatically update the attendance. It provides us the precise outcome.
Index Terms—component, formatting, style, styling, insert I. INTRODUCTION These days, the ID card punching system is used for class attendance system. One of the major problems of an ID-cardbased attendance system is that a student can provide the attendance of another student with the help of an ID-card. Even though the other student may not be present in a particular class, his attendance will be given and everyone will think that he was present on that particular day in his class. It will be very difficult to solve this issue because the teacher in a class has to memorize the face of everybody and maintain track of who gives the correct attendance which is really time consuming and sometimes impossible because the teacher has to teach in the class. These days, this fraud is going on so much. This is why we offer you facial recognition attendance system that will fix this kind of fraud and make no mistakes because the images of the faces of the students will be saved as well as their attendance so if any student tries to offer the attendance of another student, he/she will be caught. Some advantages of this system are: Firstly, the security level will improve as a face biometric system improves your security measures. Secondly, it is an easy integration process as facial recognition tools work pretty flawlessly. Thirdly, the accuracy rate is really high because the success level of face tracking technology these days
© 2022, IRJET
|
Impact Factor value: 7.529
|
ISO 9001:2008 Certified Journal
|
Page 1142