Biometric Indexing System Using Prototype Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current biometric identification systems face inefficiencies in searching large databases due to the need to directly compare probes with all templates, leading to increased search time and resource utilization.
Innovation Solution
The system indexes a biometric corpus using a set of prototypes, reducing the number of templates to be searched by constructing a search corpus based on prototype and typicality scores, allowing for efficient matching and candidate selection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If direct comparison of probes with all templates is performed, then identification accuracy is maintained, but search time increases significantly
Solution Approach 1:
The biometric corpus is segmented into multiple prototype corpuses based on prototype templates. Each prototype corpus contains templates similar to its prototype, allowing the system to divide the large search space into smaller, manageable segments. This segmentation enables the system to search only relevant segments rather than the entire corpus, reducing search time while maintaining accuracy.
Solution Approach 2:
The system performs preliminary indexing by creating prototype corpuses before actual search operations. Prototype templates are pre-computed and stored, allowing rapid retrieval during search. This preliminary action prepares the data structure in advance, eliminating the need to process all templates from scratch during each search, thus significantly reducing search time.
2Reliability
If all templates in biometric corpus are searched, then complete search coverage is achieved, but computational resources are over-utilized
Solution Approach 1:
The system extracts and separates templates into different prototype corpuses based on their similarity to prototype templates. By extracting only the relevant templates that belong to specific prototype corpuses, the system avoids processing irrelevant templates, thereby reducing computational resource utilization while maintaining reliable search coverage within the extracted subsets.
Solution Approach 2:
Different regions of the biometric corpus are treated differently by assigning them to specific prototype corpuses based on local characteristics. Templates with similar characteristics are grouped together, allowing the system to apply search strategies tailored to each local region. This local quality approach ensures efficient processing of relevant templates while skipping irrelevant ones, optimizing computational resource usage.
3Ease of manufacture
If prototypes are randomly selected from biometric corpus, then indexing simplicity is maintained, but search efficiency may be reduced
Solution Approach 1:
The system creates prototype templates by copying and selecting representative samples from the biometric corpus. These prototype templates serve as representative copies that capture the essential characteristics of their source templates. By using these copied prototypes for indexing, the system maintains ease of manufacture (simple copying process) while achieving improved search efficiency through better prototype representation.
Data Source
AI summary
A prototype biometric identification system is disclosed that indexes a biometric corpus into indexed-corpuses using a set of P prototypes before searching for a probe in a search corpus constructed based on the indexed-corpus. The system may index the biometric corpus based on the prototypes directly or based on prototype-typicality scores.