Related Experiment Video
Updated: May 17, 2026
08:15
Capturing Dynamic Finger Gesturing with High-resolution Surface Electromyography and Computer Vision
Published on: March 28, 2025
Most probable longest common subsequence for recognition of gesture character input
Darya Frolova1, Helman Stern, Sigal Berman
1Telekom Innovation Laboratories at Ben-Gurion University of Negev, Beer-Sheva 84105, Israel. daryaf@bgu.ac.il
IEEE Transactions on Cybernetics
|October 11, 2012
Summary
This study introduces a new method for recognizing dynamic hand gestures in free air using a probabilistic template and a modified longest common subsequence algorithm. The technique achieves over 98% accuracy for isolated gestures, enhancing gesture recognition interfaces.
Area of Science:
- Computer Vision
- Human-Computer Interaction
- Pattern Recognition
Background:
- Dynamic free-air hand gesture recognition is challenging due to trajectory variations.
- Existing methods may not adequately handle probabilistic distortions in gesture data.
Purpose of the Study:
- To develop an advanced trajectory classification technique for unencumbered, dynamic free-air hand gestures.
- To improve the accuracy and robustness of hand gesture recognition systems.
Main Methods:
- An extension of the longest common subsequence (LCS) algorithm, termed the most probable LCS (MPLCS).
- A learning preprocessing stage to create probabilistic 2-D templates for each gesture.
- Utilizing gesture trajectory similarity based on subsequence length and probability.
Main Results:
- Achieved a recognition rate exceeding 98% for pre-isolated digits from video capture.
- The MPLCS algorithm effectively accounts for various trajectory distortions with different probabilities.
- Demonstrated the potential for high accuracy in dynamic gesture recognition.
Conclusions:
- The MPLCS technique offers a promising approach for accurate dynamic free-air hand gesture recognition.
- Integration into gesture recognition interfaces can significantly enhance usability for gesture character input.
- The method provides a robust solution for handling probabilistic variations in gesture trajectories.
