Get the newsletter
Sony · Biotech · Filed Aug 11, 2025 · USPTO published Nov 19, 2026 · This week's release

Sony Patents a Camera System That Keeps Watching a Patient's Breathing After They Move

Sony has filed a patent for a hospital camera that watches a patient breathe and, when the patient rolls over or a nurse steps into the shot, finds a new spot on the body to watch by itself.

A hospital camera points at a resting patient to track chest movements and monitor breathing on a nearby screen. Drawing from patent filing US 2026/0363553 A1.
A hospital camera points at a resting patient to track chest movements and monitor breathing on a nearby screen.All 23 drawings ↓
Publication number
US 2026/0363553 A1(US20260363553A1)
Applicant
SONY GROUP CORPORATION
Filed · Published
Aug 11, 2025 · Nov 19, 2026
Inventors
TAKANORI FUKAZAWA, NANA KAWAMURA, HIROYASU BABA
Status when we published
Examiner raised objections; the company can respond
Document
16 claims
US classification
600/476
Examiner
ROY, BAISAKHI (Art Unit 3797)
Parent application
National Stage Entry of PCT/JP2024/004782 (filed Feb 13, 2024)(PCTJP2024004782)

What Sony's breathing-monitor camera does

You're a nurse in an ICU, and the camera watching your patient's breathing has just lost track. Someone leaned in to adjust a line, or the patient rolled onto their side, and the little box on the screen that follows the chest is now pointing at nothing.

Sony's patent describes a way for the system to fix that on its own. A camera that records both ordinary color video and a distance map follows a spot on the patient's chest or belly using the color video. When that tracking fails, the system switches to the distance map, cuts the picture into small squares, and looks for the squares where the body is rising and falling in a regular rhythm.

It then picks the square with the strongest breathing signal and carries on, with nobody needing to re-aim anything.

From the filing · CLAIM 1
… tracks, based on the RGB image, a measurement region set in a respiration region of the patient, and, when a tracking error of the measurement region occurs, resets the measurement region, based on the Depth image.

Translation: It follows the patient's breathing area using a standard camera and switches to depth data if it loses track.

How the camera finds a new spot to watch

The system is built around an RGB-D camera: one sensor records normal color video, and a second records a depth map, where every pixel says how far away that point is. The filing names a Time of Flight sensor (which times how long light takes to bounce back) as an example of the depth sensor.

A nurse first marks a measurement region, a box on the chest or abdomen. The system follows that box from frame to frame in the color video, using optical flow (estimating how points move between frames). It decides tracking has failed when the match to the surroundings drops below a threshold, for example when the patient moves a lot, someone walks in front of the camera, or a blanket covers the box.

That is when the depth map takes over. The system:

  • Splits the depth image into a grid of candidate boxes;
  • Throws out boxes with too many pixels that returned no valid distance, boxes outside a set distance range (the filing gives 10 cm to 200 cm as an example), and boxes where distances jump around, which usually means the edge of the body;
  • Runs an FFT (math that finds repeating rhythms in a signal) on each remaining box and keeps those with a clear breathing peak;
  • Ranks them by breathing strength and sets the strongest as the new box.

The description also covers using infrared instead of color video in dark rooms, and showing the breathing waveform and rate on a monitor.

From the filing · THE ABSTRACT
… acquires an RGB image obtained by capturing an image of a patient and a Depth image indicating Depth values relative to certain locations on the patient; tracks, based on the RGB image, a measurement region set in a respiration region of the patient …

Translation: The system combines regular video and depth sensing to monitor how a patient is breathing.

Why hospitals could care about hands-free re-aiming

Hospitals have long relied on sensors that touch the patient, and those can give false readings when contact slips. A camera avoids that, but a camera that loses its target after one rolled-over patient or one passing nurse is only useful if someone notices and re-aims it. This patent targets exactly that weak spot, so the monitor could keep reading for the people caring for the patient.

The filing also describes a battery-powered camera and a cloud version that sends only breathing waveform data, so images of the patient stay inside the hospital. The examiner's first review rejected the claims as filed, which is routine at this stage and usually followed by amendments.

This is the 13th Sony filing we've tracked since July in our wearables that read your body watchlist, following applications like one on needle-free glucose and one on brain-reading headphones.

Source: US 2026/0363553 A1 at the USPTO (PDF). Full patent text and figures are in the official publication.

Get our take in your Top Stories

Liked this breakdown? Add Patentlyze as a preferred source on Google, and our plain-English take shows up more often in your Top Stories the next time Sony patent news breaks.

There are more where this came from

We read every patent application Big Tech publishes and send you the ones worth knowing. Plain English, free, every week.

The drawings · 23 sheets

23 drawing sheets from US 2026/0363553 A1 · click any drawing to enlarge

A depth-sensing camera watching a hospital bed to track breathing, drawing sheet 1 from US 2026/0363553 A1
Sheet 1 A depth-sensing camera watching a hospital bed to track breathing
Selecting an area on the chest to monitor breathing motion, drawing sheet 2 from US 2026/0363553 A1
Sheet 2 Selecting an area on the chest to monitor breathing motion
Breathing signals pulled from video and depth data, drawing sheet 3 from US 2026/0363553 A1
Sheet 3 Breathing signals pulled from video and depth data
Ways to pick a breathing target on the patient, drawing sheet 4 from US 2026/0363553 A1
Sheet 4 Ways to pick a breathing target on the patient
Refocusing on the chest after a patient moves, drawing sheet 5 from US 2026/0363553 A1
Sheet 5 Refocusing on the chest after a patient moves
Spotting the patient again after someone blocks the view, drawing sheet 6 from US 2026/0363553 A1
Sheet 6 Spotting the patient again after someone blocks the view
Hardware that tracks patient movement and sends bedside alerts, drawing sheet 7 from US 2026/0363553 A1
Sheet 7 Hardware that tracks patient movement and sends bedside alerts
How the system monitors breathing and recovers lost tracking, drawing sheet 8 from US 2026/0363553 A1
Sheet 8 How the system monitors breathing and recovers lost tracking
Show the other 15 sheets
How depth data converts chest movement into a breathing rate, drawing sheet 9 from US 2026/0363553 A1
Sheet 9 How depth data converts chest movement into a breathing rate
Calculating how strongly a patient is breathing from motion data, drawing sheet 10 from US 2026/0363553 A1
Sheet 10 Calculating how strongly a patient is breathing from motion data
Steps for finding where to watch breathing after losing track, drawing sheet 11 from US 2026/0363553 A1
Sheet 11 Steps for finding where to watch breathing after losing track
Sorting out background clutter using distance measurements, drawing sheet 12 from US 2026/0363553 A1
Sheet 12 Sorting out background clutter using distance measurements
Mapping uneven depth across the grid to discard edges, drawing sheet 13 from US 2026/0363553 A1
Sheet 13 Mapping uneven depth across the grid to discard edges
Detecting a clear breathing rhythm with a sharp motion spike, drawing sheet 14 from US 2026/0363553 A1
Sheet 14 Detecting a clear breathing rhythm with a sharp motion spike
Identifying a body area with no detectable breathing rhythm, drawing sheet 15 from US 2026/0363553 A1
Sheet 15 Identifying a body area with no detectable breathing rhythm
Picking chest spots with larger chest rises over subtle ones, drawing sheet 16 from US 2026/0363553 A1
Sheet 16 Picking chest spots with larger chest rises over subtle ones
Filtering out the background to spot chest motion, drawing sheet 17 from US 2026/0363553 A1
Sheet 17 Filtering out the background to spot chest motion
Combining color and depth video to find a patient, drawing sheet 18 from US 2026/0363553 A1
Sheet 18 Combining color and depth video to find a patient
A bedside camera sending patient video to an offsite server, drawing sheet 19 from US 2026/0363553 A1
Sheet 19 A bedside camera sending patient video to an offsite server
Parts that split work between camera and server, drawing sheet 20 from US 2026/0363553 A1
Sheet 20 Parts that split work between camera and server
Steps for automatically locking onto chest movement from the start, drawing sheet 21 from US 2026/0363553 A1
Sheet 21 Steps for automatically locking onto chest movement from the start
Tracking a patient moving across the bed, drawing sheet 22 from US 2026/0363553 A1
Sheet 22 Tracking a patient moving across the bed
Computer hardware that runs the respiration tracking software, drawing sheet 23 from US 2026/0363553 A1
Sheet 23 Computer hardware that runs the respiration tracking software
Patent filing page

Watchlists this filing belongs to

All watchlists →
Reader comments

No comments yet

Start the discussion

Real name or a handle, either is fine. Comments are read by a person before they appear, so allow a little time. Keep it about the filing.