Biometric Authentication System Using Frame Relay for Bandwidth Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing biometric authentication systems for access control in retail and similar settings are expensive, resource-intensive, and not economically viable for locations with limited bandwidth, failing to provide effective value creation for stakeholders and efficient real-time recognition.
Innovation Solution
A system utilizing a camera with in-built face detection and AI-based sensor light for energy efficiency, coupled with a scalable frame relay solution that reduces bandwidth consumption and supports multiple locations, enabling biometric authentication, attendance capture, and visitor analytics.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If real-time recognition via video streams is implemented, then user identification capability is improved, but bandwidth consumption and resource usage increase significantly
Solution Approach 1:
The patent extracts only the essential facial feature information from video streams rather than transmitting complete video data. The camera captures video, extracts facial features locally, and transmits only these extracted features to the server, significantly reducing bandwidth consumption while maintaining identification accuracy.
Solution Approach 2:
The system performs preliminary facial feature extraction at the camera端 before transmission. By pre-processing the video data to extract only relevant facial features and storing these extracted features locally, the system reduces the amount of data that needs to be transmitted and processed in real-time, thereby reducing bandwidth consumption and server load.
2Measurement precision
If continuous video streaming is used for face detection, then recognition accuracy is improved, but energy consumption and bandwidth usage increase
Solution Approach 1:
Instead of continuous video streaming, the system uses periodic or event-triggered capture. The camera captures images or video frames only when motion is detected or at scheduled intervals, rather than continuously streaming video. This periodic action maintains recognition accuracy for active users while significantly reducing energy consumption and bandwidth usage during idle periods.
3Reliability
If existing biometric authentication systems are deployed, then security functionality is provided, but system cost and complexity increase
Solution Approach 1:
The patent implements a multi-functional system where a single camera device performs multiple functions: facial feature extraction, video capture, image processing, and authentication. The server also handles multiple tasks including feature storage, comparison, and user management. This consolidation of functions into unified devices reduces system complexity while maintaining comprehensive security functionality.
Data Source
AI summary
The present invention provides a system and a method for biometric authentication using facial information to recognize users across different locations. Further, the system generates feature vectors based on the facial information and generates recognition metadata for the identification and categorization of users. The system utilizes a frame relay (FR) connect pipeline to provide a more economically efficient solution, enable locations with improper bandwidth, and support a large number of locations on the same hardware. The system uses an artificial intelligence (AI) engine for predicting one or more categorizations of the user based on the generated one or more recognition metadata.