Artificial verification system on the use of safety elements

Abstract

This article shows the development process of an artificial vision program through the Python 3.10.8 software, which verifies and detects if the person uses a surgical mask or cap (hair net or cap) or none of the surgical implements. safety, with the aim of identifying and supervising its correct use in work in the food industry. To start with the development of this program, libraries such as opencv, matplotlib, mediapipe, numpy, imutils and os were installed and used. Thus, these helped to detect faces and 3 programs were created: a program to obtain images (data) by means of video capture of images through the laptop camera, a neural network training program with the collected images and a program that detects and identifies the use or not of these implements. As results, after several tests, a 93.33% validation was obtained.