<![CDATA[The rapid adoption of face recognition technology for attendance systems has raised concerns about its reliability under varying lighting conditions, which often affect real world deployment. This study aims to analyze the reliability of a face recognition based attendance system under diverse lighting scenarios, addressing challenges in accuracy and robustness. The research employs a deep learning approach, utilizing a Convolutional Neural Network (CNN) trained on a dataset of facial images captured under controlled and uncontrolled lighting conditions, ranging from low to high illumination levels. The methodology includes preprocessing techniques for illumination normalization and feature extraction, followed by performance evaluation using metrics such as accuracy, precision, and false acceptance rate. Experimental results demonstrate that the proposed system achieves an accuracy of 92% in optimal lighting but drops to 78% under low light conditions, highlighting the impact of illumination on recognition performance. The integration of adaptive preprocessing techniques improves reliability by 12% in challenging scenarios. This study concludes that while face recognition based attendance systems are highly effective, their reliability in diverse lighting conditions can be significantly enhanced through advanced preprocessing and robust algorithm design, offering practical implications for real time biometric applications in dynamic educational and workplace settings.]]>
Copyrights © 2025
