An introduction on Remote Photoplythesmograpy and the uses in ADAS systems.

Introduction

The heart rate is defined as the number of times a heart beats per minute, as the work done by the person increases HR increases. HR variation has a huge significance to medicine, psychology and many other fields. At present standard technique to measure HR is electrocardiogram(ECG). This system will comprise of a video-based Photoplethysmogram(PPG) that will provide a contact-free determination of a person’s Heart Rate. It will use face detection for ROI constrained near-real time signal analysis. The rPPG based system will eliminate the limitations of the contact based PPG techniques.

OBJECTIVES

  • The main objective is to remotely measure the live heart rate of a person.
  • To study the concepts of face detection in Region Of Interest extraction.
  • To study the concepts of face detection in Region Of Interest extraction.
  • To work on face colour analysis in order to determine the Heart Rate of the test subject.
Region of Interest for the rPPG measurement.

Proposed method

Input Video

RGB sensor like a Camera will be used to input to video to the system.

Pre-Processing

This stage includes ROI definition and tracking using a suitable face detection algorithm.

Extraction Of PPG Signal

This stage includes decomposing the video frame into its RGB components which will be used for PPG signal analysis.

Peak detection and HR Estimation

From the filtered data the peaks will be plotted and the heart rate parameter will be extracted.

Block diagram of the rPPG methodology.

Summary

  • Small changes due to the heart beat which are invisible to human eye can be detected by the camera and processed on a computer to read out the live pulse of the person.
  • The data can be processed over time to predict diseases like Heart-Attack or Hypertension.
  • By using the contact less PPG technique the need for extra equipment for pulse measurement is eliminated.



Enjoy Reading This Article?

Here are some more articles you might like to read next:

  • Automating TensorFlow Lite Builds with GitHub Actions for Raspberry Pi and x86
  • Unlock the Precision in Power Electronics Control with Space Vector PWM