Biometric access control system through facial recognition with liveness techniques

• Autores: Rodriguez Ponce, Kevin Jose, Gutierrez Sanchez, Frank Jhosep, Mendoza De los Santos, Alberto Carlos
• Formato: artículo
• Fecha de Publicación: 2024
• Institución: Universidad La Salle
• Repositorio: Revistas - Universidad La Salle
• Lenguaje: español
• OAI Identifier: oai:ojs.revistas.ulasalle.edu.pe:article/179
• Enlace del recurso: https://revistas.ulasalle.edu.pe/innosoft/article/view/179; https://doi.org/10.48168/innosoft.s16.a179; https://purl.org/42411/s16/a179; https://n2t.net/ark:/42411/s16/a179
• Nivel de acceso: acceso abierto
• Materia: Attendance control; Artificial inteligence; Facial recognition; Liveness techniques; Control de acceso; Inteligencia artificial; Reconocimiento facial; Técnicas de vivacidad

The purpose of this article is a proposal for a facial recognition system with liveness techniques for access control using neural networks. The main focus has been aimed at improving the security of access to a system through the application of artificial intelligence in biometrics, avoiding any type of fraud and impersonation. Python was used along with the following libraries like Tkinter, Cv2, Numpy, PIL, Imutils, Mediapipe, Os, Math, Dlib, Face-recognition, Csv, Tkcalendar and Bcrypt. The results obtained by carrying out 40 tests with a real person were a similarity accuracy of 82,11%, on the other hand, to verify that the person is carrying out a live recognition, 20 tests were carried out with the photograph of that person, in this case, the system did not allow entry since the liveness verification was not met, thus concluding that the recognition system is effective and guarantees greater security in access control.