Fingerprint Image Correction Model Training Through Feedback Evaluation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Fingerprint images registered in databases often lack accurate correction, with uncorrected images affecting matching reliability, necessitating a method to automatically correct and detect such images with high accuracy.
Innovation Solution
A model generation apparatus and method that selects training data, trains a model to correct fingerprint images, calculates evaluation values, and updates the model based on these values to improve accuracy, enabling detection of uncorrected images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If fingerprint images are automatically corrected using a basic model, then processing speed increases, but correction accuracy decreases compared to forensic examiner correction
Solution Approach 1:
The patent implements a feedback mechanism where the model generates correction results, evaluation values are calculated to assess correction quality, and the model is updated based on these evaluation values. This closed-loop feedback system enables the model to learn from its errors and progressively improve correction accuracy while maintaining automated processing speed.
Solution Approach 2:
The system enables the model to self-improve through automatic evaluation and updating processes. The model corrects fingerprint images, the correction quality is evaluated, and the model is automatically updated without requiring manual intervention from forensic examiners for each correction, achieving both automation and improving accuracy over time.
2Measurement precision
If all fingerprint images are manually corrected by forensic examiners, then correction accuracy is high, but processing time and cost increase significantly
Solution Approach 1:
The system enables automatic self-correction of fingerprint images through the trained model, eliminating the need for manual correction of all images. The model processes images autonomously, providing high-speed automated correction while maintaining acceptable accuracy through the evaluation and update mechanism.
Solution Approach 2:
The patent replaces the mechanical process of manual correction by forensic examiners with an automated computational model. This substitution transitions from human manual operation to machine-based automatic correction, dramatically reducing processing time while maintaining functional adequacy through iterative improvement.
3Measurement precision
If the model is trained with all available fingerprint images, then model accuracy improves, but training complexity and computational resources increase
Solution Approach 1:
The patent segments the training process into iterative stages where the model is trained on subsets of data, evaluated, and updated progressively. Rather than requiring all data to be processed simultaneously, the system divides training into manageable iterations, reducing computational complexity while achieving high accuracy through cumulative learning.
Solution Approach 2:
The system uses partial training data in each iteration rather than requiring complete datasets. The model is trained on selected portions of fingerprint images, evaluated, and updated in stages. This partial action approach reduces training complexity and resource requirements while still achieving high accuracy through repeated iterative improvement.
Data Source
AI summary
In a model generation apparatus, a training data selection means selects sets of training data from a plurality of fingerprint images. A learning means trains a model that corrects a fingerprint image, by using the sets of training data. An evaluation value calculation means calculates evaluation values of results acquired by inputting the plurality of fingerprint images into the trained model. A model update means updates a model to be trained, based on the evaluation values. Next, the training data selection means determines sets of training data to be selected from the plurality of fingerprint images based on the evaluation values.