Bio-sensor Image Quality Scoring for Authentication Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing bio-sensor authentication systems face challenges in determining the quality of input bio-images, leading to inconsistent authentication results due to varying image quality, necessitating an objective standard for evaluating image quality and enhancing authentication accuracy.
Innovation Solution
A method is introduced to calculate a quality score for bio-images by defining and quantifying features such as brightness, roundness, and inhomogeneity, using a sample database to assess deviations and apply weights to feature scores, which is then used to determine authentication success or failure and adjust weights accordingly, with the option for image processing to enhance image quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If bio-sensor authentication is performed without objective quality assessment, then authentication process is simple, but authentication accuracy deteriorates due to varying image quality
Solution Approach 1:
The patent applies preliminary action by performing quality assessment on the input bio-image before the authentication matching process. The quality assessment module evaluates image quality metrics (sharpness, brightness, contrast) and determines whether the image meets minimum quality thresholds before proceeding to authentication, thereby preventing poor-quality images from compromising authentication accuracy
Solution Approach 2:
The patent segments the authentication system into distinct functional modules: a quality assessment module that evaluates image quality, and a separate authentication module that performs matching. This segmentation allows the quality assessment to be performed as a preliminary step, improving authentication accuracy by filtering out poor-quality images while maintaining a clear separation of concerns in the system architecture
2Measurement precision
If multiple image quality features are quantified and weighted, then image quality assessment accuracy is improved, but calculation complexity increases
Solution Approach 1:
The patent applies parameter changes by quantifying multiple image quality features (sharpness, brightness, contrast) and assigning different weight coefficients to each parameter based on their relative importance. The quality assessment calculates a composite quality score using the formula: Quality Score = Σ(wi × Fi), where wi is the weight coefficient and Fi is the normalized feature value. This weighted parameter approach improves assessment accuracy while keeping the calculation systematic and manageable
3Reliability
If weight coefficients are adjusted based on authentication results, then authentication reliability is improved, but system complexity increases
Solution Approach 1:
The patent implements feedback by using authentication results to adjust the weight coefficients of image quality features. When authentication failures occur, the system analyzes whether poor image quality contributed to the failure and adjusts the weights accordingly. This feedback mechanism improves authentication reliability by adapting the quality assessment criteria to actual performance, while the adjustments are made systematically based on accumulated authentication data
Data Source
AI summary
Provided is a method of evaluating performance of a bio-sensor. The method may include obtaining an input bio-image by using the bio-sensor that is to be evaluated; obtaining a sample database including information of M reference bio-images, where M is a natural number; and calculating a quality score of the input bio-image based on the input bio-image and the information in the sample database. The quality score may provide an objective and quantitative score for evaluating a bio-image.