Currently, in the scenario of the city of Manaus, most of the records of attendance of students in educational institutions are made manually, thus making the process quite laborious and repetitive for teachers, added to this, this system does not present to be fully reliable, because the method used shows itself vulnerable to human error. To save time and ensure that the registration is done without any failure, this work develops and applies a system to register students using facial recognition, and also has a feature that triggers an email message warning the absence of the child to the parents responsible if the student is not present in class. The project uses the Ruby language, for the development of the attendance register management system, and the Python language in conjunction with the OpenCV library for facial recognition.