Comparison of facial recognition systems Fisherfaces and LBPH for access control to restricted areas

Valdeos Acevedo D.M.; Victoria Espinoza Luis L.; Oscco D.H.; Paredes M.G.S.P.; Acuna-Condori K. · 2022 IEEE ANDESCON: Technology and Innovation for Andean Industry, ANDESCON 2022 (2022) · DOI: 10.1109/andescon56260.2022.9989768

Resumen

Face detection and recognition is a topic that has gained great importance in the scientific community in recent years. Its widest application focus on video surveillance, identification, and validation, among others. In order to determine which is more efficient in levels of accuracy and precision, we propose a comparison between the two most popular methods until this date for an identification system through facial recognition: Fisherface and Local Binary Patterns Histogram (LBPH).In addition, each technique was performed under different realtime capture scenarios. The result of our LBPH system obtained a better accuracy high up 90% than Fisherfaces.

Instituciones: Pontificia Universidad Católica del Perú; Universidad Tecnológica del Perú; Universidad de Ingeniería y Tecnología