VISION ALGORITHM DEVELOPMENT ARTIFICIAL TO DETECT THE DROWSINESS IN DRIVERS HEAVY MACHINERY MINERS

Authors

  • Figueroa Herrera, Brenda Mariana
  • Sánchez Burgos, Flavio Cesar
  • León León, Ryan Abraham

DOI:

https://doi.org/10.18687/LACCEI2023.1.1.1048

Keywords:

Drowsiness, Python, computer vision, landmarks, threshold.

Abstract

The main objective of this research is to develop an artificial vision algorithm that detects the drowsiness of heavy machinery mining drivers, using an artificial vision architecture with Python software, importing face detection libraries such as shape_predictor_68_face_landmarks with which it will be detected and They will identify each important point. In the input stage, the face will be detected with a camera placed in a strategic point of the vehicle and/or machine. Then the software detects and measures the required points. A threshold 0.22 was used at a time of 60 fps (1sec) to determine if the individual is blinking continuously or due to fatigue, if this is the case an alert message will be sent. In the tests carried out we obtained average positive results of 96.48%.

Downloads

Published

2024-04-16

Issue

Section

Articles