Diseño de un prototipo de sistema inmótico basado en reconocimiento facial usando visión artificial, para controlar el acceso en áreas restringidas.

Abstract

The objective of the present degree work was the design of a prototype of an inmotic system based on facial recognition using artificial vision, to control access in restricted areas, the achievement of the following objectives was proposed: as the main one was the design of the prototype that meets all the requirements, in addition to investigating artificial vision algorithms through which it is possible to perform facial recognition of a person, for this, cascade classifiers were used for face detection with the help of the algorithm known as binary pattern premises histograms (LBPH), after the tests carried out it was concluded that the previously designed algorithm meets the requirements set out, once the artificial vision algorithm was defined, the communication method of the artificial vision algorithm was defined with the different sensors and actuators, this communication was carried out via cable p ar braided due to the different advantages it has, in addition to this, the system was constituted by a mobile application which will receive a notification, the communication between the controller board and the application (app) was carried out through the internet. The main conclusion of the design and implementation process was the use of a raspberry pi controller card which was coupled to the system's own needs. However, its use made it necessary to treat the input and output signals from and to the controller board. It is recommended to take the proper protections when using voltages higher than 3.3 VDC, which is the operating voltage of the controller board, this because the raspberry does not have protection from the pins to the board, therefore incorrect manipulation can cause complete damage to the raspberry.