Visión artificial para el control domótico de una persona con movilidad reducida en el geriátrico mi dulce hogar

Abstract

This study presents the development of a home automation system based on artificial vision in an environment established in reference to a patient with reduced mobility, which require the attention of a worker of the residence to perform common activities of the common areas. We investigated the use of artificial vision that could satisfy the acquisition of information for processing post operation of the system, through a webcam could obtain the information which needed the algorithm to perform the actions established in each movement through Visual Studio Code. With the use of the Haar model for face detection, the main characteristics of the model studied were obtained in real time, such as the four-frame model, in addition to the visualization of specific points of the face for the processing of information prior to the activation of the control elements of the system. Through the Raspberry embedded system and its operating system Raspberry Pi OS with which the artificial vision algorithm for face recognition and nose movement detector was compiled, for its subsequent use in the home automation system which works with Node-network connecting to a MQTT server the ESP development cards controlling its operation in the established environment. Finally, the algorithm data are sent through the serial port for later use in Node-network through the serial node which will receive the message where it is determined to which device the information will be sent, this data will be sent through the MQTT server which will publish a message to the ESP sub client, which in Node-network will send data to the firebase database to present them in a unidirectional platform. The system was able to recognize the user's faces to communicate with the IoT device to control the elements of the environment.