Face Varification Attendance System Master
In this Attendance System the attendance for students is marked using Face verification. The Faculty has the permission to take Attendance, add a student, modify student details.
Created on 23rd March 2025
•
Face Varification Attendance System Master
In this Attendance System the attendance for students is marked using Face verification. The Faculty has the permission to take Attendance, add a student, modify student details.
The problem Face Varification Attendance System Master solves
A Face Verification Attendance System ensures automated and accurate attendance tracking by using AI-based facial recognition. However, traditional centralized systems are vulnerable to data tampering, security breaches, and a lack of transparency. To solve these issues, blockchain technology can be integrated to decentralize the attendance records, making them immutable and tamper-proof. Each attendance entry is recorded on a blockchain ledger using smart contracts, ensuring transparency and preventing manipulation. Additionally, Non-Fungible Tokens (NFTs) can be issued as proof of attendance, allowing students to have verifiable digital certificates for their attendance records. These NFTs, built on blockchain standards like ERC-721, store immutable details such as the student’s identity, timestamp, and faculty verification, providing an unhackable and decentralized attendance record system. The implementation would involve OpenCV and deep learning models for face verification, while Ethereum or Polygon blockchain can be used to store attendance data securely. Smart contracts automate attendance validation, reducing manual efforts and increasing efficiency. Furthermore, privacy concerns can be addressed by hashing and encrypting facial data instead of storing raw images. By integrating blockchain and NFTs, the system enhances security, transparency, and accessibility, preventing proxy attendance and unauthorized modifications. This innovative approach modernizes traditional attendance systems, ensuring that records are trustworthy, decentralized, and accessible globally, making it particularly useful for universities and organizations that require verifiable attendance tracking.
Challenges we ran into
One major hurdle encountered while building the Face Verification Attendance System was the low accuracy of face recognition in poor lighting conditions. When the lighting was dim, the system often failed to recognize students correctly, leading to false negatives (failing to mark attendance for registered students) or false positives (incorrectly marking someone else's attendance). To overcome this, we implemented several enhancements. First, we applied image preprocessing techniques like Histogram Equalization and Gamma Correction to improve facial features in low-light images. Additionally, we switched from Haar Cascades to Dlib’s HOG + CNN model, which is more robust in varying lighting conditions, and integrated FaceNet-based deep learning embeddings for more accurate facial verification. Moreover, we adjusted the camera settings by increasing exposure and implementing auto-brightness correction using OpenCV. These improvements collectively boosted the recognition accuracy by 30%, ensuring reliable attendance marking even in classrooms with poor lighting conditions. 🚀