Motion Amplification and Pulse Measurements

Overview

Collaboration with Andrew Steele to recreate smartwatch functionality using a camera and software.
Smartwatch sensors use flashing lights, typically green, to measure pulse and oxygen levels.
Andrew's software aims to replicate these measurements using visual data from a camera pointed at the speaker's face.

Uses green light as it's the most effective for detecting pulse due to its depth penetration in the skin.
Green light offers a balance between depth penetration and clarity of the signal.
Blue light doesn't penetrate deep enough; red light goes too deep, picking up other signals.
Amplification of red, green, and blue channels from the video of a face to detect heartbeat visually.
Results validated with an ECG strapped to the speaker's chest and a stethoscope for audio confirmation.

Heart rate changes with breathing due to the chest cavity volume changes affecting blood flow.
A faster heartbeat when lungs expand and slower when they contract.
Watches can measure this variability to determine breathing rate.

Uses red and invisible infrared light to measure blood oxygen levels.
Red light reflects more when blood is oxygenated, and less with deoxygenated blood.
Infrared shows the opposite: deoxygenated blood reflects more infrared light.
Watches measure the relative changes in these light reflections to determine oxygen levels.
Ambient light and positioning factors less impactful with this dual light measurement approach.

Shows the potential for non-invasive monitoring, such as checking a premature baby's vitals by filming them remotely.
Demonstrates the intricate relationship between technology and biology, specifically in how light and sensors can read vital signs.

KiwiCo as an educational tool for children, encouraging STEM learning through monthly projects.
Mention of long-term benefits in promoting curiosity and problem-solving skills in children.