Desarrollo de un prototipo de sistema de prevención de intrusos en residencias, basado en una plataforma Raspberry Pi mediante procesamiento de imágenes.

Abstract

The present project has as objective to design and build a prototype that contributes to the prevention and detection of intruders within a residence, based on open source hardware and software platforms. The system it is composed of a node whose hardware is conformed of a Raspberry card and a camera. Although the camera focuses on the entire background of the scene to be monitored, the system has been designed to establish a specific area of interest, in which the intruder detection process is carried out. This process is carried out by applying the Python programming language and using the open source computer vision library OpenCV. The moment the node detects the presence of an intruder, it proceeds to alert the user via email about the event in progress. Which leads to the execution of the local storage process of the captured image, as well as the recording of the event details in the cloud server with the respective visual evidence. Through the use of a web application, the user can have access to the information hosted on the server, and take the pertinent actions that guarantee their personal integrity and subsequently notify the relief units. Finally, the implemented prototype is subjected to the respective performance tests in the scenarios that are considered to be the main access areas of a residence.