Head Position Interpolation Using Secondary Landmark Tracking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional vision-based biometric systems, such as those used in vehicle welcome modes, face challenges in continuous tracking of individuals when their head turns beyond a certain angle, leading to increased latency and computational load, and require restarting the welcome mode timer, which decreases the system's effectiveness.
Innovation Solution
The biometric recognition module interpolates between head turns by tracking secondary human characteristics like body parts or clothing, allowing continuous user authentication and reducing latency by temporarily storing facial data and using gait recognition to enhance intent confidence, thereby maintaining position tracking and optimizing computational load.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the system uses vision-based biometric recognition to track users, then user authentication accuracy is improved, but system complexity and computational load increase when users turn their heads beyond 30 degrees
Solution Approach 1:
The system introduces an intermediary tracking mechanism that uses secondary body parts (hands, feet, clothing) as mediators when facial recognition fails due to head turns. This intermediary approach maintains tracking continuity without requiring complex real-time facial analysis at extreme angles, thus reducing computational load while preserving authentication accuracy.
Solution Approach 2:
The system dynamically switches between facial recognition mode and secondary body part tracking mode based on the user's head orientation. When the head is within 30 degrees, facial recognition is used; when it exceeds this angle, the system transitions to tracking secondary body parts. This dynamic adaptation optimizes the balance between accuracy and computational complexity.
2Measurement precision
If the system continuously tracks users using facial recognition, then tracking precision is improved, but latency increases and welcome mode timer must restart frequently
Solution Approach 1:
The system performs preliminary authentication using facial recognition when the user's face is visible, storing the authenticated state and user identity in advance. When the user turns their head and facial recognition becomes unreliable, the system continues tracking using pre-stored identity information combined with secondary body part tracking, avoiding the need to restart authentication and reducing latency.
Solution Approach 2:
The system maintains continuous tracking by seamlessly transitioning between facial recognition and secondary body part tracking. This continuity ensures that the welcome mode timer does not need to restart, as the tracking action remains uninterrupted even when the primary facial recognition method becomes less effective due to head turns.
3Measurement precision
If the system uses facial recognition within a narrow field of vision, then authentication accuracy is improved, but tracking reliability decreases when users are at odd angles
Solution Approach 1:
The system implements multi-functionality by using multiple tracking targets: facial features for primary authentication and secondary body parts (hands, feet, clothing) for supplementary tracking. This universal approach allows the system to maintain tracking reliability across various user positions and angles, while preserving authentication accuracy when the face is visible.
Solution Approach 2:
The system segments the tracking function into multiple independent components: facial feature tracking for authentication and secondary body part tracking for position maintenance. This segmentation allows each component to operate optimally within its own domain, with facial tracking providing high accuracy when available and body part tracking providing reliable position estimation when facial features are obscured.
Data Source
AI summary
A method for maintaining tracking using non-identifiable biometrics includes determining, via a computer vision system, an identity of a user using a set of facial features, associating the set of facial features with a head and a secondary human landmark that does not include the set of facial features, and determining that the set of facial features is unlocatable by a computer vision system. The method includes tracking the user, via the computer vision system, using the head and the secondary human landmark.