A Multi-channel Eye-Mask-based Wearable System for Accurate Eye Movement Recognition and Saccadic Angle Estimation

Advanced wearable systems can accurately capture eye movement patterns, providing users with precise, comfortable, and reliable HCI solutions. In this paper, a Multichannel Eye-mask-based Wearable (MEW) system that offers a portable, high-sensitive eye movement recognition (EMR) and saccadic angle estimation (SAE) via the combination of multi-channel eye-mask electrodes, high-precision EOG acquisition system, and automatic EMR and SAE algorithms is proposed. The multi-channel eye-mask electrodes provide high-spatial-resolution EOG sensing, and the high-precision EOG acquisition system leverages an STM32 microcontroller and a BLE transceiver for wireless EOG data transmission. The proposed system was validated with EOG signals collected from 20 subjects over two days, during which they performed 8-directional EMR and SAE tasks. An average accuracy of 87.2% can be achieved on the EMR task. Besides, the mean absolute error (MAE) and root mean square error (RMSE) for the SAE task are 4.19°and 5.27°, indicating its effectiveness in capturing and decoding multiple eye movements. Overall, the proposed MEW system offers a reliable, portable, and high-sensitive solution in eye movement directions and angle tracking, which is also expected to be easily employed in practical scenarios for EOG-based HCI applications.