Advanced Facial Recognition Systems for Real-Time Embedded Applications
Mohammed Guerbaoui, Nouhaila Misbah, Abdelouahed Selmani, Samira El Faiz, Bachir Benhala, Abdelali Ed‐Dahhak, Abdeslam Lachhab · E3S Web of Conferences · 2025
This article explores several classical techniques of facial recognition, assessing their suitability for real-time embedded systems such as digital cameras. It also includes a thorough discussion on the training requirements for various facial recognition methods. Addressing mounting concerns in security and surveillance, this project aims to develop an advanced facial recognition system using a 2D detection approach. Leveraging the Intel RealSense 415 camera connected to a Raspberry Pi 3. An introduction to Haar Cascading models is provided, highlighting their advantages, particularly their ability to ensure acceptable levels of accuracy for facial recognition in unseen image collections. Lastly, the article offers a detailed description and implementation of a functional platform, accompanied by preliminary results.