Drowsiness detection using machine learning

Project Title

Drowsiness detection using machine learning

Project Area of Specialization Computer ScienceProject Summary

Drowsiness (feeling of sleepiness or lethargy), is a symptom that occurs due to sleep deprivation. Because of this condition of sleepiness many accidents occur mainly on highways and rural roads.

The proposed system is designed to minimize the accidents and occurrences due to drowsiness. As drowsiness is one big cause for accidents around the world and have turned into a world problem thus vigilant and alert systems are necessary in today’s time, that said this system will be designed as a smart system, using artificial intelligence and deep learning, the main objective of the project is to detect the eye movements, facial expressions and behavioral patterns to study the state of the driver. When the symptoms of drowsiness are detected, then the person is alerted with repeated voice sound and a message is forwarded to the owner or related authority. Here, buzzer is implemented for more vigilance.

Project Objectives

the main objective of this project is to detect the symptoms and alert the user using a buzzer.

Project Implementation Method

The proposed system has the following steps involved in processing:

            • Image detection

            • Face detection and information extracting phase

            • Eye region readings

             Overall system represented in the form of a diagram:

Raw data of images is used as an input to the system. The system has to detect if the image has a face or not. If not, it moves on to the next image. When the face is detected, it uses ROI algorithm to reduce to the eye part of the region of face. After this The next step is that the system then converts the image to a gray scale and histogram and sends to the classifier according to the condition put in.

1. 1. 1. Fuzzy Model

This is an alternative model. In this case, the images are preprocessed by using artificial intelligence (linear SVM combined with HOG and ensemble of regression trees) and deep learning techniques (pre-trained CNN), which extract numerical characteristics that can be introduced on a fuzzy inference system (FIS). After this, the FIS returns a numerical output that represents the estimated drowsiness level of the driver, and this value allows the system to raise an alarm if needed.

Benefits of the Project

Health checkup

Reduce the probability and danger of accidents

Conditional Monitor checkup

Technical Details of Final Deliverable

• The symptoms are to be detected and based on the detected symptoms the user will be notified through buzzer.

Final Deliverable of the Project Software SystemCore Industry ITOther Industries Health Core Technology OthersOther Technologies Artificial Intelligence(AI)Sustainable Development Goals Good Health and Well-Being for PeopleRequired Resources

Item Name	Type	No. of Units	Per Unit Cost (in Rs)	Total (in Rs)
			Total in (Rs)	37050
Raspberry pi 4 gb	Equipment	1	36500	36500
reports printouts	Miscellaneous	8	69	550