Default Image

Months format

Show More Text

Load More

Related Posts Widget

Article Navigation

Contact Us Form

404

Sorry, the page you were looking for in this blog does not exist. Back Home

Reduce Accidents with Data Science by Analysing Drowsiness, Yawns, And Blinks

 Driver's exhaustion is something that mainly causes deadly road accidents all around the world. If you are a frequent traveller and have travelled long distances by car driving yourself, then you must have faced drowsiness, yawns, and frequent blinks at some point. Many of us don't want to admit these things, but these small mistakes can lead to severe accidents. Therefore, if you have too faced this, it needs to be addressed, and you need to take every measure required to prevent it. 

Data Science


Moreover, according to the survey, 1 out of 5  road accidents are due to drowsiness, yawns, or blinks. Thus, you should treat any brief state of drowsiness, yawns, or blinks with a proper solution of detention and prevention. So, if you do not know how to prevent this situation, do not worry. We are here with appropriate guidance to drive safely, avoiding such conditions. 


Understanding Data Science

Data science is the process of the collection of data and information to draw meaningful insights. A data science certification course can help you upskill and draw insights from data. This helps the organisations to make valuable decisions and enrich their company's growth. Data science in itself is a vast concept that is used for the identification of various solutions. It has advanced tools like financial logs, sensors, marketing forms, etc., making working with extensive data easier and faster. 



Factors For Drowsiness Detention 

Drowsiness is not being in the state of sleep, but it is in which a person becomes unconscious from time to time while driving. And this negatively impacts human behaviour and activity. The National Highway Traffic Safety Administration estimates that over 72,000 vehicles crash every year due to drowsy driving. Let's see the factors that the data scientists study for the drowsiness detention. 


Physical Indicators  

Physical indicators like frequency of shutting your eyes, yawning, and other irregular physical activities are recorded, and then the analysts analyse those activities. Some more characteristics, like outer and inner mouth corners, nose tip, centre and the corner of the eyes, etc., can also be registered. And if talking of physiological signals, then heart rate, EEG, and pulse rate are considered good indicators of drowsiness. 

However, after studying the various data on this issue, the data scientists concluded that when a person is tired, there is an increase in the power of theta and alpha bands of EEG. And this indicates that increasing EEG signals are signs of drowsiness. 


Behaviour Of The Vehicle

Speed, driving angle, and vehicle position are certain factors that have limitations. While driving, the driver must see the type of road, weather condition, and the type of car. If the driver knows about these conditions, and still the vehicle is at the wrong angle or even at a fast speed, then this behaviour of the vehicle is said to be in the hand of an unconscious or tired driver. And thus, these are the things that an analyst observes, studies, and then concludes with accurate solutions.

 

Computerised Vision Analysis 

A proper computerised analysis detects the facial expression and the vehicle's behaviour to obtain the data. Data scientists usually take informative pictures from those collections and discard the irrelevant pictures. 


Signs Of Drowsiness

There are several signs that the data scientists find after studying and observing the collected informative data. The most common signs are as follows: 


Yawning 

Yawning is a typical indicator of a driver's fatigue and drowsiness. By measuring the driver's hypervigilance, especially in the mouth, through video recordings, scientists conclude that yawning is a facial feature that indicates the driver's drowsiness. The yawing analysis method includes face extraction, mouth localization, and circular hough transform.

 

Blinking 

Blinking is another common sign of a driver's drowsiness, but this sign is recognized only when the driver is not wearing any glasses. The eye blink detection module goes through facial and eye region detection, which sometimes may be faulty due to the limitation of the mobile camera.

Both yawning and blinking are only considered when there is enough light to display their facial changes. Thus, the modules performed for their detection are not accurate at night. 


Machine Learning Methods Used in Data Science To Reduce Accidents

Data science has stretched its hand to identify and provide better solutions to road accidents due to drowsiness, yawning, and blinks. And this has come a long way, concluding with the amazing development of the device. When the driver is drowsy, the device attached to the vehicle senses the activity and then sounds the alarm to alert the driver. 


Some machine learning methods that are used to detect the drowsiness of drivers are: 

Facial Action Coding System 

It is a common and popular expression coding system. Generally, one facial expression is divided into almost 46 part movements to analyse the facial action of the driver. You can detect the motions of the head through the accelerometer. Furthermore, FACS is still under development, and scientists are figuring it to capture the emotional states. 


Support Vector Machines

This ML method is also for face detection, and it captures the eye pictures and then sends it to the ML algorithms to further the process. It determines if the eyes are open or closed. If closed, then it sends an alarm alert to the driver. 


Hidden Markov Model 

This statistical model is mainly used to predict the hidden physical state using the obtained states. It can track the eyes based on their colours and geometrical characteristics. 

The above methods are performed with Python and the inbuilt OpenCV library, which is designed for all detection features. Apart from this, various steps are involved in the detection of drowsiness. They are video recording, face detection, facial feature extractions, their analysis, and then classification is done for decision making. Moreover, you can extensively use machine learning to avoid aeroplane or pilot accidents also, and you can learn it all by getting an MTech in data science India. Check out this course from Great Learning, which you can pursue at your ease and convenience. 

No comments:

Post a Comment