A dynamic gesture recognition system for the Korean sign language (KSL)

Area of Science:

• Computer Science
• Artificial Intelligence
• Human-Computer Interaction

Background:

• Sign language is a crucial communication method for the deaf-mute population.
• Recognizing and translating sign language into text presents significant challenges.
• Existing methods may lack efficiency in real-time gesture classification.

Purpose of the Study:

• To develop an automated system for recognizing Korean Sign Language (KSL).
• To translate recognized KSL gestures into standard Korean text.
• To enhance communication accessibility for deaf-mute individuals.

Main Methods:

• Utilized a pair of data-gloves as the primary sensing device to capture hand and finger movements.
• Proposed an efficient motion classification technique for accurate gesture recognition.
• Implemented a fuzzy min-max neural network for on-line pattern recognition of sign language gestures.

Main Results:

• Successfully demonstrated a system capable of recognizing Korean Sign Language gestures.
• Achieved translation of recognized gestures into Korean text.
• The proposed classification technique and neural network facilitated efficient on-line recognition.

Conclusions:

• The developed system offers a viable solution for translating Korean Sign Language into text.
• Data-gloves combined with fuzzy min-max neural networks provide an effective approach for sign language recognition.
• This technology has the potential to significantly improve communication for the deaf-mute community.