Dynamic Facial Gesture Authentication for Biometric Security

The patent combines facial feature recognition with dynamic gesture recognition into a unified authentication system. The terminal captures both static facial features and dynamic facial gestures simultaneously, merging two authentication factors into one integrated process that maintains convenience while significantly improving security against password theft and spoofing attacks.

The patent introduces dynamic elements into authentication by requiring users to perform specific facial gestures (such as opening mouth, closing eyes, or facial expressions) in response to prompts. This dynamic component transforms static password authentication into a time-varying biometric authentication process, making it resistant to replay attacks and stolen credential usage.

The patent utilizes the existing camera module of the terminal to perform multiple functions: capturing static facial images for feature extraction and capturing video sequences for dynamic gesture recognition. By making the camera serve dual purposes, the system avoids adding dedicated hardware sensors, thereby limiting the increase in device complexity while achieving enhanced authentication reliability.

The terminal itself performs the authentication processing locally using its built-in processor and algorithms. The system extracts facial features, recognizes gestures, and makes authentication decisions without requiring external authentication servers or additional processing equipment. This self-contained approach minimizes system complexity by eliminating dependencies on external infrastructure.