Fatigue Detection of Workers using Supervised Learning

Author: Nisha Yadav , Kakoli Banerjee and Vikram Bali

Download PDF

Abstract

For worker fatigue detection, machine vision and image processing could be helpful. There have been many research developed in this field in recent years for driver drowsiness detection but not for workers in software industries. A new method for worker fatigue detection has been proposed in this paper that uses eye condition for fatigue state. There are many techniques used for driver drowsiness detection but the same cannot be used for fatigue detection as one of the important factors is screen illumination for the workers who are working day and night on the systems (laptops or computers). Screen illumination is the light of the computer screen or laptop screen that is casted on the workers face that affects the eyes of the workers. Fatigue not only degrades quality but also acts as a health risk factor resulting in sleep disorder, depression, stress and also decreases the productivity of the company. To avoid the mistakes occurring due to fatigue a mechanism is proposed to measure

Keywords

Face Detection, Extraction, Drowsiness, Estimation

Conclusion

The proposed approach presents worker fatigue detection system which is based on computer vision with machine learning. The proposed system uses eyes state of the worker who is working on the system with the help of real time frame capturing from video. In the first, we locate face and eyes of the worker using Haar feature. Shape predictor allows the iris centre detection and the points of intersection for the two to calculate aspect ratio of the eyes. This analysis is confirmed by the facial landmark detection algorithm to find bests results for the benefit for the workers. The results of this approach show a good precision rate for features of the eyes. It can therefore be concluded that worker fatigue can be detected in advance by using the proposed method before any error or fatal health condition occurs. It will not only measure productivity, but it will also ensure software workers ' health condition.

References

-

How to cite this article

Yadav, Nisha, Banerjee, Kakoli and Bali, Vikram (2019). Fatigue Detection of Workers using Supervised Learning. Biological Forum – An International Journal, 11(1): 236-242.