Document Type
Article
Publication Date
Fall 11-6-2024
Abstract
Ensuring security in crowded areas poses a significant problem due to the dense population and the intricate task of monitoring their behaviors. This research presents a real-time security system that utilizes facial expression analysis to identify potential threats. The technology utilizes sophisticated facial recognition and emotion detection techniques to accurately recognize emotions such as wrath, fear, and anxiety. These feelings can serve as indicators of suspicious activities or potential security risks. The system utilizes a Convolutional Neural Network (CNN) to classify facial expressions and recognize objects. In addition, it integrates an SVM classifier trained on features derived from the combined CNN-10 and ResNet50 architecture to improve classification accuracy Moreover, the proposed facial expression analysis model employs distributed processing to ensure that the model can function and make real-time decisions. This technology, when included into the current surveillance infrastructure, offers a strong and effective option for observing and examining crowd behavior. It allows for quick responses to possible security situations. Digital Rights Management (DRM) solutions are utilized to guarantee the security and integrity of identified threats. DRM implements stringent access restrictions, thereby prohibiting unauthorized viewing, duplication, or dissemination of identified surveillance data. Results validate the efficacy of the suggested method in precisely identifying and categorizing facial emotions in real-time, hence improving the overall safety and security of densely populated areas
Recommended Citation
Zayed, Nourhan; I. Sayed, Hassan; and F. El-Kabbany, Ghada, "A Distributed Secure System for Threads Detection in Crowded Environments based on Biometric Facial Expression" (2024). Mechanical Engineering. 131.
https://buescholar.bue.edu.eg/mech_eng/131
