Sistema de vigilancia inteligente implementado en local comercial El Chaval

Abstract

The development of the technical project seeks to implement an intelligent security system in a small commercial establishment. The developed system seeks to provide security to employees and customers by using a security camera that detects objects used to hide the face such as hats, masks, and balaclavas, as well as a microphone that detects threatening words and phrases. Upon detection of the established patterns, an alert message will be sent via SMS to the venue's security manager. Field test results showed that object detection efficiency depends on factors such as light, object orientation and camera movement, while voice detection is affected by noise and distance. The open source object detection system YOLO V5 was used for training and image processing, and Google Speech Recognition was used for voice pattern detection. A Jetson Nano was used to process the algorithms, since it has a GPU that facilitates and improves performance for object detection in real time.