Optical Action Recognition Wearable Eliminates EMI Noise
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing action recognition systems using surface electromyography (SEMG) face challenges such as limited processing speed, complex signal amplification circuitry, restricted measuring positions, and vulnerability to power source and electromagnetic interference noises.
Innovation Solution
An action recognition system that optically detects muscular variations using an annular body with light emitting and sensing units, generating a light sensing signal to determine user actions, thereby eliminating EMI interference and power source noises and increasing the number of measurable positions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If surface electromyography (SEMG) is used for action recognition, then muscular activity can be detected, but the system is vulnerable to power source noises and electromagnetic noises
Solution Approach 1:
The patent replaces the electrical SEMG detection system with an optical detection system using light emitting units and light sensing units. This substitution eliminates vulnerability to electromagnetic interference and power source noises by using optical signals instead of electrical signals for detecting muscular activity.
Solution Approach 2:
The patent introduces light as an intermediary medium between the muscle and the detection system. The light emitting unit emits light that interacts with the muscle tissue, and the light sensing unit detects the reflected or transmitted light, thereby indirectly measuring muscular activity without direct electrical contact that would be susceptible to electromagnetic interference.
2Reliability
If SEMG signal amplification circuitry is implemented, then signal strength can be increased, but the circuitry becomes relatively complex
Solution Approach 1:
The patent replaces complex electrical signal amplification circuitry with optical detection components. The light sensing units directly convert optical signals from muscle interaction into detectable signals, eliminating the need for complex electrical amplification stages while maintaining signal detection capability.
3Ease of operation
If traditional SEMG measuring positions are used, then action recognition can be performed, but the measuring positions are limited only to the chest and forearms
Solution Approach 1:
The patent creates a universal action recognition system that can be applied to multiple body parts beyond just chest and forearms. The optical detection components can be configured on various movable parts of the body, enabling the same system architecture to recognize actions across different anatomical locations, thereby increasing versatility and adaptability.
4Productivity
If FPGA is used for electromyographic signal processing, then signal processing can be performed, but the processing speed does not easily comply to the requirements for action recognition
Solution Approach 1:
The patent replaces electrical signal processing through FPGA with optical signal detection and processing. The optical system detects muscular variations directly through light-matter interaction, providing inherently faster response times that better comply with real-time action recognition requirements.
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
The system provides accurate and efficient action recognition by optically detecting muscle variations, enhancing operational convenience and improving recognition efficiency while minimizing interference from noise sources.
Implementation Method 1
The light sensing unit is operable to sense a second light beam reflected by the at least a portion of the movable part for generating a light sensing signal
Implementation Method 2
an action recognition system operable to determine an action of a user by optically detecting the muscular variation of a movable part of the user
Data Source
AI summary
A wearable device is illustrated. The wearable device has a body, at least one light emitting unit, at least one light sensing unit and an action recognition module. The wearable device is suitable for wearing on a wrist of a user. The light emitting unit is disposed on an inner side of the body, and the light emitting unit emits a light beam illuminating at least a portion of the wrist. The light sensing unit operatively senses the light beam reflected by the at least portion of the wrist and generates a light sensing signal. The action recognition module is configured to operatively determine a function that corresponds to an action of the user.