Video Conferencing Face Recognition Using Historical Authentication Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Video conferencing devices face challenges in maintaining high security and image quality due to factors like noise, poor bandwidth, and environmental changes, which can lead to unauthorized access even by authorized persons.
Innovation Solution
A video conferencing device and image analysis method that samples video frames, detects faces and objects, and selects a security mode based on the number of people and facial features, using previously stored information to authenticate authorized persons even in poor image quality conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If strict face recognition is performed to ensure security, then security performance is improved, but false negatives increase when image quality is poor
Solution Approach 1:
The system performs preliminary actions by storing multiple reference images of authorized persons before the actual authentication. When a face is detected, the system compares it against these pre-stored references and also utilizes historical authentication data to make a more reliable determination, thereby reducing false negatives in poor image quality conditions
Solution Approach 2:
The system dynamically adjusts recognition parameters based on image quality assessment. When image quality is determined to be poor (due to noise, motion, or illumination changes), the system modifies its recognition threshold and utilizes additional contextual information from previous frames to maintain accurate authentication
2Reliability
If continuous monitoring of multiple parameters is performed to maintain security, then security reliability is improved, but device complexity increases
Solution Approach 1:
The system segments the security monitoring task into distinct functional modules: image quality assessment module, face detection module, reference comparison module, and historical data analysis module. Each module handles a specific aspect of the authentication process, making the overall complex system more manageable and maintainable while achieving high security reliability
Solution Approach 2:
The image analyzer performs multiple functions using a unified processing framework. It simultaneously assesses image quality, detects faces, compares against references, and analyzes historical data within a single integrated system, reducing the need for separate dedicated components for each function
3Productivity
If frame sampling is performed to reduce processing load, then processing speed is improved, but recognition accuracy may deteriorate
Solution Approach 1:
The system implements periodic action by sampling video frames at optimized intervals rather than processing every frame. The frame sampling rate is dynamically adjusted based on detected motion and authentication state, processing frames more frequently when motion is detected or authentication is pending, and less frequently during stable states, thereby maintaining accuracy while improving processing speed
Data Source
AI summary
The present disclosure provides methods and apparatuses for analyzing an image detected by a camera that includes sampling, by an image converter, a current frame from a video frame, detecting, by an image analyzer, at least one face and an object in the current frame, determining, by the image analyzer based on the detecting of the at least one face and the object, a number of people, a number of faces, and whether a facial feature in the current frame corresponds to an authorized person, and selecting, by the image analyzer, a security mode based on the number of people, the number of faces, and whether the facial feature corresponds to the authorized person. The selecting of the security mode includes accessing, based on the at least one face not being recognized, position information about the authorized person in at least one previous frame stored in a database.