Biometric Identification Device Intermediate Image Validation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing biometric fingerprint identification methods are time-consuming due to the need for acquiring and analyzing multiple images to ensure quality, which prolongs the capture and identification process.
Innovation Solution
Implementing an intermediate identification phase during the capture process that selects and validates images based on a more restrictive match rate, allowing early initiation of the identification process without waiting for the completion of the capture operation, and optionally proceeding to a final identification phase if the intermediate phase fails.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple successive images are acquired and analyzed to ensure fingerprint quality, then identification accuracy is improved, but identification time increases
Solution Approach 1:
The patent applies preliminary action by performing intermediate identification during the capture phase before the final identification phase. The system extracts features and performs preliminary comparisons with stored features during image acquisition, allowing early identification decisions without waiting for complete capture of multiple images. This overlapping execution of identification tasks during capture reduces total identification time while maintaining accuracy through subsequent verification phases.
2Reliability
If a series of successive images are captured to guarantee image quality, then reliability of identification is improved, but productivity decreases
Solution Approach 1:
The system performs preliminary feature extraction and intermediate identification during the capture phase itself, rather than waiting for capture completion. This allows the system to begin processing with available images while continuing to capture more images, thereby improving throughput without compromising the reliability guarantee provided by the final identification phase that still uses multiple captured images.
Solution Approach 2:
The identification process is segmented into distinct phases: capture phase with intermediate identification, and final identification phase. The intermediate identification phase processes images independently during capture, while the final phase performs comprehensive verification. This segmentation allows parallel processing and improves overall system productivity while maintaining reliability through the two-stage verification approach.
3Loss of time
If intermediate identification phase is implemented during capture, then processing time is reduced, but computational resource usage increases
Solution Approach 1:
The intermediate identification phase performs partial identification actions during capture by comparing extracted features with stored features at a reduced matching rate threshold. This partial action provides early identification indications without requiring the complete computational effort of full identification, thus reducing overall processing time while consuming fewer computational resources than a complete final identification would require.
Solution Approach 2:
The system changes the matching rate parameter between intermediate and final identification phases. The intermediate phase uses a more restrictive (higher) matching rate requirement for quick elimination of non-matches, while the final phase uses a lower matching rate threshold for comprehensive verification. This parameter adjustment optimizes the balance between processing speed and computational resource consumption at different stages.
Data Source
AI summary
Method of rapid identification of an object, in the course of a process comprising a phase of capturing a series of images of an object and then a final phase of identification; the method comprising in an intermediate phase of identification on the basis of at least one image representing the object selected during the capture of the series of images and the object being identified when the characteristics extracted and the characteristics stored in association with an object identifier correspond mutually to a predetermined degree of correspondence which is more restrictive than a degree of correspondence used during the final identification phase, the final identification phase then not being carried out. Identification device for the implementation of this method.