Hands-Free Virtual Accessory Try-On via Motion Detection

The patent replaces mechanical touchscreen interactions with computer vision-based motion detection. The system captures video of the user's face and detects predefined target motions (such as raising hands, pointing, or head movements) to control accessory selection, eliminating the need for physical touchscreen contact while maintaining selection accuracy through recognized motion patterns

The system enables users to control the virtual try-on process through their own natural body motions rather than requiring external input devices. By detecting motions performed by the user themselves (such as hand raises or head tilts), the system allows intuitive hands-free operation where the user's natural movements directly control the accessory application process

The patent defines and prepares a set of predefined target motions in advance (such as specific hand gestures, head movements, or body poses) before the user interacts with the system. During operation, the system simply detects which predefined motion the user performs, rather than requiring complex real-time interpretation of arbitrary movements, thereby reducing processing complexity while maintaining hands-free operation

The system transforms continuous video data into discrete motion categories by detecting changes in key parameters (such as position, orientation, or pose) of body parts. By monitoring specific parameters (hand position, head angle, etc.) and comparing them against thresholds that define predefined motions, the system simplifies the complex video stream into actionable control signals without requiring excessive computational resources