detect your heart rate from a webcam · remote photoplethysmography · green channel variance · no hardware needed · runs locally
how it works
every heartbeat pushes a pulse of blood through your face capillaries, slightly changing skin color. the green channel is most sensitive — it varies by ~0.5–2% with each beat. this tool averages the green channel of a 120×120 px center ROI, buffers 300 frames, and computes a discrete fourier transform. the peak frequency in 0.7–3.3 hz (42–200 bpm) is your heart rate. works best with good, steady lighting and a still face.
✦ steady lighting · face or wrist in frame · sit still · takes ~10 s to calibrate
camera
why this matters
- rppg is used in telehealth today. apple watch, samsung galaxy watch, and hospital-grade remote patient monitoring all use photoplethysmography. some apps claim to detect atrial fibrillation from a phone camera. the physics is the same.
- the page sees your heartbeat. any site with camera permission can run this continuously in the background. with enough frames (~30s) it can also estimate heart rate variability (hrv), a proxy for stress and autonomic nervous system state.
- why green? hemoglobin absorbs green light (~545 nm) more strongly when oxygenated. the red channel works too but is more sensitive to motion; blue is too absorbed. most rppg systems use a combination weighted by snr.
- limitations. dark skin tones, heavy makeup, and poor lighting reduce snr. motion artifacts (talking, head tilts) are the main source of false readings. signal quality is shown as "confidence" — below 30% the bpm reading is unreliable.