Samsung Files Patent to Detect Skin Tone Using Pulse Sensor Technology
Most wearable health sensors were designed and tested primarily on lighter skin tones, which makes their readings less reliable for many people. Samsung is now patenting a way to measure skin tone directly from the same sensor already used to track your pulse.
How Samsung's wrist sensor figures out your skin tone
Imagine two flashlights shining on your wrist at the same time, but with different types of light-filtering lenses. One filter only picks up light that bounced off the very surface of your skin. The other captures light that went deeper, through the layers where your blood actually flows. Compare the two readings, and you can figure out how much melanin (the pigment that determines skin tone) is present.
That's essentially what this Samsung patent describes. The device sends polarized light (think of how polarized sunglasses cut glare) onto your skin and uses two separate receivers to capture the reflected signals. One receiver has a polarizing filter angled to catch only the surface reflection; the other grabs everything else. The difference between those two readings tells the device your skin tone.
Why does that matter on a smartwatch? Because pulse oximeters and heart-rate sensors work differently depending on how much melanin is in your skin, and a device that knows your skin tone can automatically adjust its readings to be more accurate for you specifically.
How polarized light separates surface from deep-tissue signals
The patent describes a wearable sensor with three key components working together.
- A light emitter fitted with a polarizing filter that sends light at a specific angle onto the user's skin.
- A first light receiver with a polarizing filter angled perpendicular (cross-polarized) to the emitter. Cross-polarization means it mostly blocks light that reflected straight off the skin's surface, letting through only photons that scattered deeper into tissue.
- A second light receiver with no polarizing filter, so it captures all returning light, including the surface reflection.
The device generates a PPG signal (photoplethysmogram, the pulse-wave signal that smartwatches already use to measure heart rate) from each receiver separately. The surface-heavy signal from the unfiltered receiver and the deeper-tissue signal from the cross-polarized receiver will differ in a predictable way based on how much melanin sits in the skin's upper layers. Processing circuitry compares the two signals and infers skin tone from that difference.
Critically, the system extracts skin-tone information passively, as a byproduct of the same light pulse used for health monitoring, without requiring a separate dedicated sensor or a special user-initiated calibration step.
What this means for health accuracy on darker skin tones
Pulse oximeters and optical heart-rate sensors have a well-documented accuracy gap on darker skin tones, a problem that attracted significant attention during the COVID-19 pandemic when consumer devices were being used to track blood oxygen. If a wearable can automatically detect skin tone at the hardware level, it can apply the right correction factors without the user ever having to enter anything manually.
For Samsung, which competes directly with Apple's Watch and Google's Fitbit line in the health-tracking market, building bias correction into the sensor itself is a meaningful product differentiator. It also positions the company ahead of expected regulatory pressure in the United States and Europe requiring wearable medical devices to demonstrate accuracy across a wider range of skin tones.
This is genuinely important work, not a patent filed for defensive shelf-sitting. The accuracy gap in optical health sensors across skin tones is a real, documented problem, and embedding the fix at the hardware level is the right approach. Whether Samsung ships this in a Galaxy Watch or keeps it in the lab is the only open question.
Which company should we read for you?
We track 17 companies here. Pro is the same weekly breakdown for any company you choose, delivered privately. Type a name and we'll scope it and send you a quote.
Get one Big Tech patent every Sunday
Plain English, intelligent commentary, no hype. Free.
Editorial commentary on a publicly published patent application. Not legal advice.