Litcius/Paper detail

A Bio-Impedance Analysis Method Based on Human Hand Anatomy for Hand Gesture Recognition

Haofeng Chen, Gang Ma, Peng Wang, Xiaojie Wang

2021IEEE Transactions on Instrumentation and Measurement19 citationsDOI

Abstract

In this paper, we presented a bio-impedance analysis method (BIAM) based on human hand anatomy, and propose a feasible and flexible BIAM system to obtain the bio-impedance signals for different gestures to achieve a high classification accuracy in hand gesture recognition with flexibility in electrode arrangement and fewer electrodes. To verify the proposed method, 11 gestures including two sets: hand gestures and pinch gestures were selected for the experiment. Based on the functional structure of the human hand, we identified appropriate electrode positions and placed five electrodes on the hand surface for bio-impedance signal measurement. Compared to the electrical impedance tomography (EIT) method, which uses a band with the same number of electrodes wrapped around the wrist, the proposed method achieved 98.7% recognition accuracies on the hand gesture set and 97.8% on the pinch gesture set, while the EIT can only achieved 97.1%, and 86.3%, respectively. In particular the proposed method demonstrated the advantage of distinguishing gestures with similar muscle contractions.

Topics & Concepts

GestureComputer scienceArtificial intelligenceElectrical impedanceSet (abstract data type)Gesture recognitionFlexibility (engineering)Computer visionAcousticsSpeech recognitionPattern recognition (psychology)EngineeringMathematicsPhysicsElectrical engineeringProgramming languageStatisticsMuscle activation and electromyography studiesEEG and Brain-Computer InterfacesHand Gesture Recognition Systems
A Bio-Impedance Analysis Method Based on Human Hand Anatomy for Hand Gesture Recognition | Litcius