Trusted Execution Facial Verification Using Infrared and Depth

The patent segments the verification process into multiple independent stages: first verifying infrared images, then verifying depth images, and finally performing facial feature matching. Each stage operates independently with its own template comparison, creating a multi-layered security architecture that prevents single-point failures and makes tampering significantly more difficult while maintaining user convenience through automated multi-modal verification.

The patent employs a composite verification approach by combining infrared image templates and depth image templates into a unified verification system. This composite structure uses multiple types of biometric data (infrared facial characteristics and depth information) that complement each other, creating a security system where the whole is more secure than the sum of its parts, as tampering with one modality does not compromise the other.