Supramolecular Wood-Based Human–Computer Interaction Interface for Text Recognition

Text recognition is a cornerstone of human-computer interaction (HCI), yet conventional isotropic sensing interfaces are intrinsically constrained in resolving the complex directional features of handwriting, particularly the intricate stroke orientations of Chinese characters. Here, leveraging the intrinsic anisotropic architecture of natural wood, we developed a mechanically robust, highly stretchable, and sensitive hydrogel-based interface capable of accurately distinguishing diverse handwriting trajectories. The resulting wood-based soft hydrogel used for human-computer interaction (WSH-HCI) achieves 25 MPa tensile strength along the fiber axis─6.5× higher than in the transverse direction─while maintaining stable multidirectional signal discrimination and up to 96% accuracy in Chinese basic-stroke recognition. These performances arise from preserved aligned cellulose nanofiber channels and a strong PVA-cellulose hydrogen-bonding network. Demonstrations include real-time recognition of English letters and Chinese characters. This work provides a viable pathway toward high-performance anisotropic hydrogel sensors for advanced handwriting recognition and intelligent human-computer interactions.