Ambientes virtuais de aprendizagem com biometria facial

Abstract

Online education has achieved great importance in Brazil’s education. Through the ease of access to the new Information and Communication Technologies, it has expanded in geographical and numerical reach. This way, Online Education has been democratizing teaching and knowledge, taking them to remote areas where face-to-face teaching encounters numerous difficulties in serving the population. The development of tools that support the quality of Virtual Learning Environments (VLE) has become frequent in order to improve Online Education. The topic of this research is Information Security in VLEs, and its hypothesis is that facial biometry can contribute to this aspect through the monitoring of students in the virtual platforms of teaching in courses in the modality of Online Education. Our proposal is to improve the authentication procedure on the pallet with the use of facial biometrics, enabling a verification of the vision with facial recognition. The study was composed of three stages. In the first stage, a bibliographical research was carried out. In the second stage, a questionnaire was developed and applied to the Online Education members at Instituto Federal do Piauí and Universidade Federal do Piauí, in which we detected a positive opinion about the creation of new authentication tools. Based on the bibliographical research and the analysis of the data collected through the questionnaire, the third step consisted in the development of a prototype of facial recognition for VLE. A total of 83 individuals, including students, face-to-face tutors, online tutors, coordinators, content teachers and research professors participated in the study. As a technological innovation for Online Education, our facial recognition prototype was implemented with the PHP and PYTHON languages, and the OPENCV computer vision library for haarcascade facial detection and facial recognition with Local Binary Histogram Patterns (LBPH). The tool was developed in order to recognize the student through facial patterns, recording their presence in Virtual Learning Environments. In the prototype tests, a database with 167 images was used, as well as the registration and recognition of 20 users and fraud simulation tests. For performance tests a database with 226 images was set up. The prototype demonstrated efficient performance and effective facial recognition of users, although it requires new implementations.