HEART RATE MONITORING: THE ANALYSIS OF FINGERTIP VIDEO CAPTURED VIA SMARTPHONE

Current new generations of smart phones have high technical potential where they can benefit from them in several areas, including telecare applications. This research aims to develop a simple system to analyze and study of fingerprint images from the video captured for monitoring the heart rate (HR) without the need for sensors. The mean frame difference (MFD) vector is the averaged difference between each two consecutive frames, candidate peaks chosen are positive signals and higher than their two neighborhoods, and outliers are removed using the Quartile method. Virtual peaks are estimated from their calculated signals and the mean-error (E) is computed to choose the proper peaks. Experimental results are compared against Nonin Go2 Fingertip Pulse Oximeter readings and show that the overall performance of our system is 87.91%.