Facial Authentication Device Using Multi-Method Feature Point Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing facial authentication technologies face accuracy declines due to positional deviations of detected face feature points, caused by noise, small face sizes, or partial coverings, which are not effectively addressed by current methods.
Innovation Solution
A facial authentication device and method that detects multiple face-feature-point candidates using various methods, calculates reliability degrees based on statistical information, and selects the most accurate face image for authentication, thereby mitigating the influence of positional deviations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If facial authentication is performed on all frames including the face in a moving image, then the authentication coverage is improved, but the calculation processing time increases and authentication efficiency deteriorates
Solution Approach 1:
The patent applies preliminary action by performing face detection and quality assessment on frames before conducting full facial authentication. The system pre-identifies candidate frames that contain faces and pre-evaluates their quality metrics (blur level, occlusion, pose) to select only suitable frames for authentication, avoiding unnecessary processing of unsuitable frames.
Solution Approach 2:
The patent implements partial action by performing facial authentication on only a selected subset of frames rather than all frames. The system applies authentication processing selectively to frames that meet predetermined quality criteria, performing partial authentication operations on disqualified frames (such as blur detection) while skipping full authentication on unsuitable frames.
2Reliability
If facial authentication is performed on frames with poor quality (blurring, partial covering), then the authentication coverage is improved, but the authentication accuracy deteriorates
Solution Approach 1:
The patent performs preliminary quality assessment of face images before authentication by evaluating blur levels, occlusion程度, and pose angles. Frames that fail to meet quality thresholds are identified in advance and excluded from authentication processing, ensuring that only high-quality frames undergo authentication.
Solution Approach 2:
The system implements feedback mechanisms by continuously monitoring frame quality metrics and using this information to adjust authentication decisions. The quality assessment results feed back into the frame selection process, creating a closed-loop system that adapts to varying image qualities and maintains authentication accuracy.
3Reliability
If multiple face-feature-point candidates are detected using various methods, then the reliability of feature point detection is improved, but the calculation processing complexity increases
Solution Approach 1:
The patent merges multiple face feature point detection methods into a unified detection system. Different detection algorithms (such as landmark-based methods, contour-based methods, and template-matching methods) are combined to detect feature points, with their results integrated to produce more reliable detection outcomes.
Solution Approach 2:
The system creates multiple copies of feature point detection using different algorithms and then compares the results. By generating duplicate detections through various methods and selecting the most consistent or reliable results, the system enhances detection reliability while managing computational complexity.
Data Source
AI summary
This facial authentication device is provided with: a detecting means for detecting a plurality of facial feature point candidates, using a plurality of different techniques, for at least one facial feature point of a target face, from a plurality of facial images containing the target face; a reliability calculating means for calculating a reliability of each facial image, from statistical information obtained on the basis of the plurality of detected facial feature point candidates; and a selecting means for selecting a facial image to be used for authentication of the target face, from among the plurality of facial images, on the basis of the calculated reliabilities.