Vehicle Driver Authentication Using Local Face Feature Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current intelligent vehicle systems rely heavily on network connectivity for driver identification, which can lead to safety concerns when connectivity is lost, and they lack effective methods for real-time monitoring of driver states such as fatigue and distraction.
Innovation Solution
A driving management system that uses a camera assembly to collect video streams of vehicle drivers, performs feature matching with pre-stored face images to authenticate drivers, and conducts real-time driver state detection for fatigue, distraction, and predefined actions, allowing for autonomous vehicle control and data storage without relying on continuous network connectivity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the system uses network connectivity for driver identification, then the accuracy of driver authentication is improved, but the reliability of the system deteriorates when network connectivity is lost
Solution Approach 1:
The system performs preliminary actions by storing driver face images and feature data locally in the vehicle before network connectivity is needed. The data set including multiple face images of registered drivers is pre-loaded into local storage, enabling the vehicle to perform driver identification independently when offline. This preliminary storage of authentication data resolves the contradiction by ensuring the system can maintain reliable operation even when network connectivity is lost.
2Reliability
If the system performs real-time driver state monitoring, then the safety of driving is improved, but the complexity of the system increases
Solution Approach 1:
The camera assembly performs multiple functions: it captures driver face images for identification authentication and simultaneously monitors driver states such as fatigue, distraction, and abnormal behaviors. By making the camera multi-functional, the system improves driving safety through real-time monitoring without proportionally increasing system complexity. The same hardware resource serves dual purposes, resolving the contradiction between safety improvement and complexity increase.
Solution Approach 2:
The system merges the driver identification function and driver state monitoring function into a unified processing framework. Both functions process video stream data from the camera and use similar image processing techniques, allowing the system to combine these functions efficiently. This merging reduces overall system complexity while maintaining comprehensive safety monitoring capabilities.
3Reliability
If the system processes and stores video stream data locally, then the dependence on network connectivity is reduced, but the storage requirements increase
Solution Approach 1:
The system extracts only the essential authentication data (face images and feature information) from the complete video stream and stores these extracted features locally. Rather than storing entire video files, the system processes the video stream in real-time and extracts key facial features for authentication purposes. This extraction approach reduces storage requirements while maintaining the ability to operate independently from network connectivity.
Solution Approach 2:
The system transforms the video stream data into a different parameter representation by converting visual information into extracted facial features and authentication data. This parameter transformation reduces the data volume significantly while preserving the essential information needed for driver identification and state monitoring, thereby reducing storage requirements while maintaining network independence.
Data Source
AI summary
Embodiments of the present disclosure disclose driving management methods and systems, vehicle-mounted intelligent systems, electronic devices, and medium. The method includes: controlling a camera assembly provided on a vehicle to collect a video stream of a vehicle driver; obtaining a feature matching result of a face part of at least one image in the video stream and at least one pre-stored face image in a data set, where the data set stores a pre-stored face image of at least one registered driver; and if the feature matching result represents that the feature matching is successful, controlling the vehicle to execute an operation instruction received by the vehicle. The embodiments of the present disclosure reduce the dependence of the driver identification on a network, can realize feature matching without the network, and further improve the safety guarantee of the vehicle.