In fiscal year 2019, the road transport passengers amounted to around 22.6 trillion passengers per kilometer accross India. The fact that this number is so huge where we can’t rely on other driving passenger for our safety leads us to at least contribute to develop a system which can be used in the mobile vehicles to monitor the drivers at each and every moment.

This project is a drowsiness detection system. It uses a camera to monitor the EAR (eye aspect ratio) of the driver. After it reaches a certain value, an alarm is played to wake the driver and make him focus on the road. If you are travelling in a group, this could be a good time switching drivers.

"40% of the on road accidents occur due to Sleep deprivation"

The following project is an implementation of @akshaybahadur21: Drowsiness Detection System, with additional improvements to give users more detailed information. Below I have mentioned the changes,

• Graphical Output
• Drowsy Count displayed on the screen
• Twilio API integration to send 'sms' to emergency contact number
• Displaying real time data over thingspeak

To view, video file

Whenever the screen displays "Drowsiness Detected", SMS is sent to the verified emergency contact number using Twilio API. TWILIO
Below, these were the frames which were captured at the time of drowsiness	Not a really "good acting" from my end

• You need to run the main_drowsy.py for execution

C:\Users\iwill\Desktop\github\11-Are you Drowsy\Are-you-Drowsy> python main_drowsy.py

• The frames captured via execution of the project will update automatically via this line and will be saved in the Drowsy-Clicks folder

cv2.imwrite('Drowsy-Clicks/drowsy%d.jpg' % drowsy_count,frame)		#extracts that particular frame

• For Twilio API to work uncomment the below line which is commented out by default
• For Twilio Documentation, see

client.api.account.messages.create(to="emergency_number",from_="+19403267422",body=body)

This project has a room for lots of improvement which can be done together to add new features or displaying the data in a more beautiful manner. I hope you can connect with me via my Discord