Smartwatch Gesture Recognition Using IMU and Pulse Wave Sensing

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing gesture recognition technologies on smart wearable devices suffer from poor environment adaptability and recognition accuracy, particularly in scenarios with significant changes such as outdoor walking or strong sunlight, leading to unstable and inaccurate gesture recognition.

Innovation Solution

A smart wearable device combines inertial measurement unit (IMU) and pulse wave sensor data to recognize gestures by analyzing attitude and pulse wave features, using a gesture classifier to improve recognition accuracy and robustness, and selectively activating sensors to reduce power consumption.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Device complexity

If gesture recognition uses only inertial measurement unit data, then device complexity is reduced, but recognition accuracy and robustness deteriorate in varying environments

Engineering Contradiction:
Improvesensor system complexityVSAvoidgesture recognition accuracy
Core Design Contradiction:
Device complexityVSMeasurement precision

Solution Approach 1:

The patent combines data from two different sensor types (inertial measurement unit and pulse wave sensor) to recognize gestures. The IMU captures motion characteristics while the pulse wave sensor captures physiological changes during gesture execution. By merging these complementary data sources, the system achieves higher recognition accuracy and environmental robustness without proportionally increasing system complexity.

Inventive Principle:
Principle #5Merging (Combining)

2Measurement precision

If pulse wave sensor is continuously activated for gesture recognition, then recognition accuracy is improved, but power consumption increases

Engineering Contradiction:
Improvegesture recognition accuracyVSAvoiddevice power consumption
Core Design Contradiction:
Measurement precisionVSUse of energy by moving object

Solution Approach 1:

The pulse wave sensor is not continuously activated but is periodically enabled only when the inertial measurement unit detects preliminary movement patterns that suggest a gesture is being performed. This periodic activation based on trigger events maintains high recognition accuracy while significantly reducing overall power consumption compared to continuous sensor operation.

Inventive Principle:
Principle #19Periodic action

3Speed

If gesture recognition system activates all sensors immediately upon movement detection, then recognition speed is improved, but power consumption increases

Engineering Contradiction:
Improvegesture recognition speedVSAvoiddevice power consumption
Core Design Contradiction:
SpeedVSUse of energy by moving object

Solution Approach 1:

The system performs preliminary detection using only the inertial measurement unit to identify movement patterns that precede actual gestures. Only after this preliminary detection confirms a gesture is in progress does the system activate the pulse wave sensor. This staged approach maintains fast recognition response while avoiding unnecessary sensor activation and associated power consumption.

Inventive Principle:
Principle #10Preliminary action

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 gesture recognition accuracy and robustness across varying environments, reducing misrecognition and improving user interaction experience by enabling more intuitive gesture-based operations.

Implementation Method 1

collecting first attitude data through an inertial measurement unit

Methodology Applied
Scientific EffectInertial measurement: Inertia

Implementation Method 2

collecting pulse wave data through a pulse wave sensor

Methodology Applied
Scientific EffectPhotoplethysmography: Absorption (EM radiation)

Data Source

PatentEP4310636B1Gesture interaction method, system, and apparatus
Publication Date: 2026.03.25 HUAWEI TECH CO LTD
  • EP4310636B1 patent drawingFigure 1
  • EP4310636B1 patent drawingFigure 2A
  • EP4310636B1 patent drawingFigure 2B~3A

AI summary

This application provides a gesture interaction method. By implementing the method, a smart wearable electronic device such as a smart watch can obtain a plurality of types of data when a user completes a specific gesture. The gesture completed by the user is recognized based on the plurality of types of data, so that accuracy and robustness of gesture recognition can be improved, and a probability of misrecognition is reduced. Particularly, when an environment in which the user is located changes significantly, the electronic device such as the smart watch can accurately recognize a user gesture. Therefore, in a process of controlling an operation of the electronic device through gesture recognition, the user can obtain better use experience.