DRIVER DROWSINESS AND FATIQUE DETECTION USING PYTHON

ABSTRACT:-

Drowsiness, fatique and tiredness while driving a car is very common nowadays and this leads to increase in the number of accident . So, now the question how we can decrease the number accident on the road. For decreasing the number of accident on the road we can follow two methods: 1.First method is by detecting the  pulse rate and heart beat of a person using digital equipment. 2. Second method is using open cv in these we detect the facial expression, mainly the eyes of the person as soon as the person eyes get closed for a couple of second the alarm would get raised in the car after applying both the methodology, we came to the conclusion that the second method is more better as it is less expensive and more accurate. Keywords: Detection, Driver Drowsiness.

METHODOLOGY:-

• We have two method to detect the drowsiness of the driver.
• Detecting the pulse rate and heart beat

We can detect the fatique of the driver by measuring the pulse rate and heart beat. But for this driver has to wear the equipment all the time .Also In this method the cost of equipment is very expensive.

• Real Time Computer Vision Systems

In this method, we detect the face expression of the driver such as eyes, mouth, eye bros using the opencv and dlib library. By calculating the euclidean distance of the eye ,we drive a ratio containing horizontal and vertical  part of the eye .if the ratio of eyes decrease to the level of 2 our model raises an alarm for alerting the driver.

The image below shows the indexes of the 68 coordinates:

     The eye aspect ratio (EAR) between height and width of the eye is calculated.

EAR = ||p2 − p6|| + ||p3 − p5|| (1)

 2||p1 − p4||

where p1, . . ., p6 are the 2D landmark locations, depicted in Below Diagram. When an eye is open the EAR remains  constant and gets close to zero while closing an eye.

 RESULTS AND DISCUSSION:-

The Drowsiness detection with Python and OpenCV was implemented using the following steps: Successful

runtime capturing of video with camera. Captured video was separated into frames and each frame were

analysed. Than the Successful detection of face followed by detection of eye is done. If eye remains closed for

successive frames and were detected, then it is classified as drowsy condition else it is considered as normal

blink and the loop of capturing and detecting the image and analysing the situation of driver is carried out again

and again. When the person is in drowsy condition the eye is not surrounded by circle or it is not detected, and

alarming sound is made to awake the driver.

MODULES:-

• Camera: To detect person drowsyness.
• L298 Motor Driver: We Use this L298  Motor Driver for roated motor.
• GPS Sensor: We Use Gps Sensor For get latitude and longitude of that location where drowsyness
• Ultra Sonic Sensor : Ultra Sonic Sensor use for major the distance.
• Database: All information like person image ,latitude and longitude ,speed of vehicle add in database
• Android Application : We use android application in this project.
• Firebase: Get notification in android app when drowsyness detected. and we can see the location name, person image speed and location on google map in android application.

CONCLUSION

This paper proposed a method for detection of driver drowsiness from video. Here, a method for automatically

measuring facial expressions was employed to determine spontaneous behavior during real drowsiness event.

A real-time eye blink detection algorithm was proposed. We displayed that Haar feature-based cascade

classifiers and regression-based facial landmark detectors are accurate enough to reliably estimate the positive

images of face and a proportion of eye openness. This project also shows the significance of using instances of

fatigue and drowsiness conditions in which the person actually fall sleep

HARDWARE:-

• Processor: Intel Core i3 or more.
• RAM: 4GB or more.
• Hard disk: 250 GB or more.
• GPS Sensor
• Ultrasonic Sensor
• L298 Motor Driver
• 12 volt Adapter
• Shop View Sensor
• Camera
• Raspberry pi

SOFTWARE:-

• Operating System : Windows 10, 7, 8.1.
• Android
• Python
• Spyder, Jupyter notebook, Flask.