EOG-Based Eye-Writing Systems: 2013-2025 Review

This paper comprehensively reviews eye-writing systems based on electrooculographic (EOG) signals. These systems, which enable communication by utilising the potential changes caused by eye movements, are generally classified into two main groups in the literature: virtual keyboard-based and pattern recognition-based. Virtual keyboard systems rely on thresholding methods and do not require artificial intelligence algorithms. On the other hand, in pattern recognition systems, character input is provided by users following specific eye movement patterns, and these patterns are analysed using artificial intelligence-based classification algorithms. In this context, the hardware, signal processing techniques, classification algorithms and interface designs have been systematically analysed by evaluating the studies published in the last twelve years.