Development of algorithm for the recognition and detection of drowsiness in interprovincial transport drivers (#508)

Authors

León León, Ryan Abraham

Medina Ticona, Vanessa Lisseth

Vargas Vargas, Alexia Jackeline

Abstract

The purpose of this research is to explain the detection through the development of an algorithm for the recognition of drowsiness in drivers, making use of image processing, which is achieved through the application of artificial vision with PYTHON software and facial detection techniques. The application of this project is developed in the transportation sector, within the area of interprovincial service. For the development of this project several steps have been considered. The first is to know and recognize the face of the people so that through recognition can be detected each of the important points for analysis. One of them was both the right eye and the left eye which, through a strategically located camera, can detect the person by capturing an image, then the software can recognize and measure parameters of distances between the dimensions that have the eyes and the time they generate a blink or an extension of the blink. This software recognizes and with the mentioned parameters, makes a decision to define whether or not there is a state of drowsiness within the analyzed processing. Finally, when detecting a defined number of prolonged blinks, it alerts by means of a message that the driver is in a drowsy state. It was determined that the design of the algorithm detected and recognized 100% of the people analyzed. Also, 150 tests were performed which showed a system effectiveness of 99%