Mattress Pressure Arrays for Privacy-Safe Sleeping Posture Recognition

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing methods for recognizing sleeping posture at night are hindered by the obstruction of blankets and violate user privacy through camera-based image recognition.

Innovation Solution

A sleeping posture recognition system using a pressure sensor array on a mattress to detect body pressure data, which is converted into a two-dimensional array and processed by a deep neural network for accurate posture recognition, addressing privacy concerns and improving accuracy.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If a camera is used to capture sleeping posture images, then image recognition can be performed, but user privacy is violated and data capture is blocked by blankets

Engineering Contradiction:
Improvesleeping posture recognition accuracyVSAvoidprivacy violation and data capture obstruction
Core Design Contradiction:
Measurement precisionVSObject-affected harmful factors

Solution Approach 1:

The patent replaces the optical measurement system (camera) with a mechanical pressure sensing system. Pressure sensors embedded in the mattress detect body pressure distribution to determine sleeping posture, eliminating the need for visual capture and thus resolving privacy concerns and blanket obstruction issues while maintaining recognition capability

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent introduces pressure distribution data as an intermediary between the user's body and the recognition system. Instead of directly capturing images of the user, the system measures pressure patterns indirectly, which then serve as the basis for posture recognition through deep learning models

Inventive Principle:
Principle #24Intermediary (Mediator)

2Ease of operation

If blankets are used during nighttime sleep, then user comfort and privacy are maintained, but camera-based data capture is blocked

Engineering Contradiction:
Improveuser comfort and privacy protectionVSAvoidsleeping posture detection capability
Core Design Contradiction:
Ease of operationVSMeasurement precision

Solution Approach 1:

The patent replaces optical detection with mechanical pressure sensing, allowing users to sleep with blankets without interfering with detection. The pressure sensors beneath the mattress surface can detect body position through the blanket, maintaining both user comfort and detection capability

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent extracts the detection function from the visual domain to the mechanical domain. By removing the camera and introducing pressure sensors, the system separates detection from visual capture, allowing blankets to remain in place while still enabling posture recognition through pressure distribution patterns

Inventive Principle:
Principle #2Taking out (Extraction)

Applied Scientific Principles

This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.

Function Achieved in This Case

Enhances sleeping posture recognition accuracy and resolves privacy issues by utilizing a mattress-integrated pressure sensor array and deep neural network, ensuring precise posture identification without camera intrusion.

Implementation Method 1

a pressure sensor array is deployed in a sleeping position on a mattress to obtain body pressure data

Methodology Applied
Scientific EffectPiezoresistive effect: Piezoresistive Effect

Data Source

PatentEP4636716A1Sleeping posture recognition method and system based on deep neural network
Publication Date: 2025.10.22 DONGGUAN ONETHIRD SLEEP TECHNOLOGY CO LTD
  • EP4636716A1 patent drawingFigure 1
  • EP4636716A1 patent drawingFigure 2
  • EP4636716A1 patent drawingFigure 3~4

AI summary

Disclosed in the present invention are a sleeping posture recognition method and system based on a deep neural network. The method comprises the steps of: inputting body pressure sample data into a deep neural network for training learning, to obtain a sleeping posture recognition model; obtaining a two-dimensional body pressure array in real time by using a detection device, the detection device obtaining two-dimensional body pressure analog signal data by means of a pressure sensor array, and converting the two-dimensional body pressure analog data into the two-dimensional body pressure array by means of an A/D conversion module; and transmitting the two-dimensional body pressure array to a server for preprocessing, and inputting the preprocessed two-dimensional body pressure array into the sleeping posture recognition model for recognition. According to the present invention, the pressure sensor array is provided at a sleeping position on a mattress to obtain body pressure data in different sleeping postures, and the sleeping posture recognition model is used for performing sleeping posture recognition on the basis of the two-dimensional body pressure array obtained in real time, thereby improving the accuracy of sleeping posture recognition, and solving the problem in the prior art that using a camera to recognize the sleeping posture of a user at night is unlikely to be implemented and invades the privacy of the user.