Desenvolvimento de um Dispositivo para Segurança Biométrica Facial Usando Câmera NoIR

Abstract

This work aimed to develop an anti-spoofing detection device for biometric security, using the NoIR camera for Raspberry Pi. The increasing need to enhance security in facial recognition systems, especially against spoofing attempts using photos and videos, motivated this study. The proposed solution combines computer vision techniques, such as the calculation of the Normalized Difference Vegetation Index (NDVI), adapted to distinguish real faces from forgeries. The methodology included the creation of a dataset with real and fake images and the development of algorithms for analyzing these images. The results showed that the system was effective in identifying real faces and detecting spoofing attempts, validating the approach employed and suggesting its potential for practical applications.