Face Collation Apparatus with Unregistered Image Registration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing face collation systems struggle with low precision due to variations in facial direction, light, and illumination, causing mismatched face images to be unusable, even if the person is the same as a registered individual.
Innovation Solution
A face collation apparatus that acquires, collates, and stores face images, allowing unregistered images to be displayed and selected for registration as collation face images, improving precision by increasing the database of images and enabling relearning of collation data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If face images are collated using strict matching criteria, then false positives are reduced, but collation precision decreases due to variations in facial direction, light, and illumination
Solution Approach 1:
The system performs preliminary action by acquiring multiple face images of the same person under different conditions (facial directions, lighting, illumination) before collation, and registers them in advance. This allows the collation process to match against a comprehensive set of reference images, improving precision without increasing false positives.
Solution Approach 2:
The system changes parameters by varying facial direction, light direction, and illumination conditions when acquiring multiple face images of the same person. These parameter variations are registered in advance, enabling the collation system to recognize faces under diverse conditions and improve measurement precision while maintaining reliability.
2Productivity
If only matched face images are used for collation, then processing speed is maintained, but collation precision deteriorates due to limited database size
Solution Approach 1:
The system implements self-service by automatically acquiring multiple face images under different conditions and registering them without requiring manual intervention. The system serves itself by expanding its own database with diverse images, improving collation precision while maintaining processing speed through automated operations.
Solution Approach 2:
The system performs preliminary action by pre-acquiring and pre-registering multiple face images of the same person under various conditions before collation is needed. This preliminary database expansion enables faster and more accurate collation without requiring real-time image capture under multiple conditions.
3Measurement precision
If multiple face images under different conditions are registered, then collation precision is improved, but device complexity increases
Solution Approach 1:
The system performs self-service by automatically managing the acquisition, organization, and registration of multiple face images under different conditions. This automated self-service approach improves collation precision while minimizing the increase in device complexity, as the system handles the complexity internally without requiring additional external equipment or manual processes.
4Quantity of substance
If multiple face images are acquired and registered, then a comprehensive database is created, but time consumption increases
Solution Approach 1:
The system performs preliminary action by acquiring and registering multiple face images in advance before collation is needed. This preliminary registration creates a comprehensive database that enables fast and accurate collation operations, trading initial registration time for subsequent processing efficiency.
Solution Approach 2:
The system implements self-service by automatically acquiring and registering multiple face images without requiring manual intervention for each image. This automated process reduces the time consumption associated with manual registration while still building a comprehensive database for improved collation precision.
Data Source
AI summary
To improve precision in collation processes or create a database of collation images by registering at least a part of captured images not matching a previously registered face image, as a collation face image, a face collation apparatus includes an image acquirer that acquires captured face images captured by a capturing device, a registered-person image storage that stores registered-person images that are previously registered face images, an image collator that collates the captured images with the registered-person images, an unregistered-person image storage that stores the captured images that do not match the registered-person images, an image output that outputs at least one of the unregistered-person images stored in the unregistered-person image storage to a display device, and an image register, selected to be registered by a user, that registers the unregistered-person image, among the unregistered-person images output to the display device, as a registered-person image in the registered-person image storage.