A Secure Cryptographic System Based on Steady-State Visual Evoked Potential Brain-Computer Interface Technology

Addressing the vulnerability of contact-based keyboard password systems to disclosure, this paper proposes and validates the feasibility of a non-contact secure password system based on brain-computer interface (BCI) technology that detects steady-state visual evoked potential (SSVEP) signals. The system first lets a testee look at a digital stimulus source flashing at a specific frequency, and uses a wearable dry electrode sensor to collect the SSVEP signal. Secondly, a canonical correlation analysis method is applied to analyze the frequency of the stimulus source that the testee is looking at, and feeds back a code result through headphones. Finally, after all password codes are input, the system makes a judgment and provides visual feedback to the testee. Experiments were conducted to test the accuracy of the system, where twelve stimulus target frequencies between 10-16Hz were selected within the easily recognizable flicker frequency range of human brain, and each of them was tested for 12 times. The results demonstrate that this SSVEP-BCI-based system is feasible, achieving an average accuracy rate of 97.2%, and exhibits promising applications in various domains such as financial transactions and identity recognition.