Arm-Worn EMG Sensor for Sign Language Gesture Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional sign language recognition systems are cumbersome, interfere with daily activities, and require frequent removal due to sweat issues, making them inconvenient for long-term use.
Innovation Solution
A sign language recognition system that uses a sensor measurement device worn around the arm to acquire electromyogram and inertial signals, processing these signals to detect gestures through feature vector production and neural network-based recognition, allowing for accurate and rapid identification of sign language gestures without the need for continuous device wear.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If sign language gloves are worn for recognition, then gesture detection capability is improved, but comfort and ease of use deteriorate due to sweat and need for frequent removal
Solution Approach 1:
The patent extracts the sensing function from the hand (glove) and relocates it to the arm (wristband). The sensor measurement device worn on the arm captures electromyogram signals from arm muscles, separating the detection function from direct hand contact, thereby eliminating sweat discomfort while preserving gesture recognition capability
Solution Approach 2:
The patent introduces an intermediary mechanism: instead of directly sensing hand gestures, the system senses muscle activation signals (electromyogram) from the arm that precede and control hand movements. This intermediary approach allows indirect but accurate detection of gesture intent without physical contact issues
2Measurement precision
If camera-based recognition systems are used, then gesture analysis capability is improved, but portability and convenience deteriorate due to device size and processing time
Solution Approach 1:
The patent replaces the mechanical/optical camera-based gesture capture system with an electrical signal-based detection system. Instead of capturing visual images of hand movements, the system directly measures electromyogram signals from arm muscles, eliminating the need for cameras and complex image processing while improving portability
Solution Approach 2:
The system detects muscle activation signals before the actual gesture is completed. By capturing the preparatory muscle signals that precede hand movements, the system can recognize gestures earlier and with simpler processing, reducing both device complexity and recognition time
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
Enables rapid and accurate recognition of sign language gestures, providing a convenient and unobtrusive method for communication that does not interfere with daily activities.
Implementation Method 1
an acquisition unit configured to acquire an electromyogram signal of a user from a sensor measurement device worn around an arm of the user
Implementation Method 2
a search unit configured to search a signal corresponding to an integrated feature vector in a database on the basis of the integrated feature vector obtained by integrating the first feature vector and the second feature vector
Data Source
AI summary
Provided is a sign language recognition system, and the sign language recognition system includes an acquisition unit configured to acquire an electromyogram signal of a user from a sensor measurement device worn around an arm of the user, an extraction unit configured to extract a muscle active section from the electromyogram signal to detect a sign language gesture of the user, a producing unit configured to produce a first feature vector by performing signal processing to the muscle active section, a search unit configured to search a signal corresponding to the first feature vector in a database, and an output unit configured to output a text corresponding to the searched signal.