Checking driver behavior with Raspberry Pi

Published on 10 Oct 2020. Written by Vinay Raj

Road accidents are one of the most common accidents that occur frequently in the world. Due to these accidents, many people are losing their lives and several people become handicapped. There are many reasons for the occurring of accidents and one of those is the drowsiness of the driver. This will majorly occur with lorry drivers as they will be driving for long distances and the rest is taken them will be less and leading to drowsiness and occurrence of the accidents.

Read more..

SLNOTE


Project Description

The main aim of the project is to detect the drowsiness of the drivers by facial recognition. This can be done with ease due to an increase in computer technology and also the internet. To make this project work we will be using the camera for live video of the driver and by using the sensor. This IRIS detection is processed by the PANDA algorithm. We can also use an accelerometer to track any sudden changes in the speed of the vehicle. You can also add GPS and GSM based tracker so that the vehicle can be traced all the time and if any accident occurred we can trigger an SMS. And if any abnormal behavior is observed in the driver it will automatically turn on the horn and seat vibrator thereby helps in alerting the driver. 

Concepts Used

• Image Processing
• PANDA algorithms
• HAAR detector algorithms

SLLATEST
Software and hardware components

• Raspberry Pi
• Python

Project Implementation

The whole project runs on the processor Raspberry Pi. All the components like accelerometer, alcohol detector, GPS and GSM module are connected to the Raspberry Pi. When the project is on, it first tries to detect the alcohol levels of the driver. Here, the MQ-3 sensor is used as it measures the ethanol content in the air. It alerts the whole system if the drive is over drunk and tries to drive the vehicle. If this is clear the project will work on the rest behavior activities of the driver it will try to look for the drowsiness of the driver and a camera is used for recording the driver continuously. 

 

This algorithm works on the eye blink where if the eyes of the driver are closed for more time, it triggers the whole system. The detection if the drowsiness is achieved by using the HAAR algorithm. This will be working as follows:

1. In the first step, it read the video from the camera.
2. In this step, the recorded video is converted into frames.
3. Here all the separated frames are read by the process
4. While reading the frames it will find the try to find the edges (here it will try to focus on eyes and facial expressions)
5. After edge features, it will try to find the line and rectangle features.
6. After recognizing all the features of the drivers the process will do the IRIS recognition.
7. If everything is normal, then it discards the current frame and goes to the next frame
8. All steps are repeated from step 4.

By detecting and monitoring the behavior of the driver we can ensure safety on roads. This project makes the drive and other users of the road safe. This even controls the rash driving of the driver in the influence of the alcohol. 

 

SLDYK
Kit required to develop Checking driver behavior with Raspberry Pi:

• Raspberry PI
• Camera

Technologies you will learn by working on Checking driver behavior with Raspberry Pi:

• Image Processing
• Computer Vision

Any Questions?