International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 11 Issue: 04 | Apr 2024
p-ISSN: 2395-0072
www.irjet.net
Face Recognition Attendance System for Employees Shreya Prabhulkar1, Sayali Angre2, Akshata Dhale3 Mayuresh Mahale4 Department of Information Technology, Terna college of Engineering, Nerul, Navi Mumbai, Maharashtra, India ---------------------------------------------------------------------***--------------------------------------------------------------------2. LITERATURE SURVEY Abstract - The integration of facial recognition technology into employee attendance monitoring systems represents a significant advancement leveraging the distinctive features of the human face as a biometric identifier. This project emphasizes two fundamental phases: face detection and recognition. The utilization of the Haar Cascade algorithm excels in rapidly identifying faces within images through its cascade of classifiers, efficiently focusing computational resources on potential face regions. This collaborative approach underscores the ongoing importance of leveraging the unique characteristics of the human face to enhance the efficiency and effectiveness of attendance tracking systems in organizations. By harnessing the strengths of multiple algorithms, this project aims to elevate the accuracy and reliability of facial recognition-based attendance monitoring, ultimately contributing to streamlined operations and enhanced security protocols within workplaces.
This paper [1] explores the application of the Haar Cascade Classifier Algorithm with OpenCV for face detection in automated attendance tracking systems. It emphasizes the algorithm's scalability, enabling real-time processing suitable for organizations of varying sizes. The accuracy of face recognition using this algorithm is highlighted, aiming to minimize errors associated with manual attendance tracking. The paper acknowledges challenges, such as potential false positives/negatives due to changes in appearance, accessories, or hairstyles. It also notes the technical expertise and time-consuming nature required for implementing and fine-tuning the algorithm. This paper introduces [2] an automatic attendance system using the Convolutional Neural Network (CNN) algorithm for face detection and recognition in classrooms, aiming for high accuracy. Leveraging CNN's proficiency in image recognition, the system ensures reliable student identification under varying conditions. It emphasizes the potential for continuous improvement with additional data while acknowledging the complexity of implementing and training CNN models, requiring deep learning expertise and suitable hardware. The paper also notes CNNs' sensitivity to data noise, urging careful consideration for optimal attendance tracking accuracy.
Key Words: Face Recognition, Human face, detection, attendance tracking.
1.INTRODUCTION In the project focusing on developing a facial recognition system for employee attendance monitoring, the unique features of the human face are harnessed as a biometric identifier. The facial recognition process involves two critical phases: face detection, for quickly identifying the presence of a face, and recognition, where the system distinguishes and matches the face as an individual.
PCA, a valuable statistical technique, is applied in face recognition and image compression, particularly through the Eigen faces approach. This method [3] utilizes a small set of characteristic images to describe variations in face images, showcasing PCA's role in identifying patterns in highdimensional data. Noted for its simplicity, PCA is more accessible to understand and implement than complex algorithms. It achieves reduced dimensionality, speeding up computation and potentially lowering computational requirements. However, effective use of PCA mandates careful data preprocessing, and its adaptability to changes in datasets may be limited, necessitating retraining for optimal performance with new variations.
In this specific project, the Haar Cascade algorithm is incorporated. The Haar Cascade method excels in face detection, efficiently pinpointing the location of faces in images. It works by employing a cascade of classifiers, each designed to quickly reject areas of an image that are unlikely to contain a face, thereby focusing computational efforts on potential face regions. The human face's unique characteristics continue to play a pivotal role in this technology. The primary objective remains to boost the efficiency and effectiveness of the current attendance tracking system in organizations, now empowered by the collaborative strengths of Haar Cascade, Eigenface, and Fisher face algorithms.
© 2024, IRJET
|
Impact Factor value: 8.226
In preparation for the Eigen Faces Recognizer, captured images undergo preprocessing to obtain grayscale and uniformly cropped faces. The [4] algorithm's simplicity makes it accessible for beginners in facial recognition, while its dimensionality reduction enhances computational efficiency. However, Eigenfaces exhibit limited discriminative power, potentially impacting accuracy by not
|
ISO 9001:2008 Certified Journal
|
Page 1220