Exhibeo: espelho inteligente para uso doméstico

Abstract

The internet of things comes in a growing and accelerated development, interconnecting the objects of daily use with the information coming from the internet. In this way, the demand of devices that allow the accomplishment of ordinary tasks in parallel with other digital activities is increasing. The objective of this work is to develop an embedded system that integrates with a mirror providing information such as time and weather information, task lists, event notifications, news and videos, as determined by the user. The system consists of a piece of glass, a wooden frame, a reflective insulfilm sheet, an LCD monitor, a webcam for user recognition and gesture capture, and a Raspberry Pi microcomputer for data processing and interactive interface, developed using the Python language. The library used to recognize gestures and the face of the user is OpenCV. The communication between the embedded device and the Android device was done through WiFi. Regarding the performance of the graphic interface and its interactions, we observed an accuracy of 80% in facial recognition and an accuracy of 85% in the recognition of gestures. The overall data synchronization presented a delay smaller than 0.3s. It can be concluded from the results obtained that the project fullfilled its objective successfully.