Palm Biometric Authentication Using Sparse Matched Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing biometric authentication systems are inefficient in reducing the number of processing units, size, production cost, and power consumption, while also requiring additional resources for image processing.
Innovation Solution
A biometric authentication apparatus that intentionally blurs images using a digital blur filter or blurring lens, employing two-dimensional sparse matched filtering and a Halo artifact removal unit to reduce computation costs, with a processing device that captures and processes palm images using a sparse matched filter with a sparse matrix, and a data storage device for vector matching.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional image processing methods are used for biometric authentication, then image quality is maintained, but processing time and computational resources increase significantly
Solution Approach 1:
The patent extracts only the essential biometric features from the palm image using sparse matched filtering, rather than processing the entire image. By extracting only the necessary vector data representing palm characteristics, the system reduces processing time while maintaining authentication accuracy.
Solution Approach 2:
The patent segments the palm image into a grid of cells and processes only the most significant vectors from each cell. This segmentation approach allows the system to focus computational resources on the most discriminating features for authentication, reducing overall processing time.
2Reliability
If multiple processing units are used to improve authentication accuracy, then verification reliability increases, but device complexity and production cost increase
Solution Approach 1:
The patent implements a universal processing device that performs multiple functions: image capture, sparse matched filtering, vector extraction, and authentication. By making the processing device multi-functional, the system achieves reliable authentication without requiring separate dedicated units for each function, thereby reducing overall device complexity.
Solution Approach 2:
The patent combines multiple processing operations into a single integrated processing device. The sparse matched filtering, vector extraction, and authentication functions are merged into one unit, reducing the number of separate components needed while maintaining authentication accuracy through coordinated operation of the combined functions.
3Productivity
If conventional image processing algorithms are applied, then processing completeness is maintained, but power consumption increases
Solution Approach 1:
The patent applies partial action by using sparse matched filtering that processes only the most significant portions of the image data. Instead of processing every pixel and feature, the system identifies and processes only the essential palm characteristics, reducing computational operations and consequently lowering power consumption while maintaining adequate authentication performance.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Invention relates to a biometric authentication apparatus and biometric authentication method for improving the speed of verification by using the sparse matched filtering approach to process the biometric data of a palm. A biometric authentication apparatus comprises an image capturing device (3), that captures images of a palm, a processing device (4) for processing each captured image and data storage device (5), that stores plurality of sets of vectors characterizing biometric data of a hand. Invention is characterized in that it comprises either digital blur filter (11), either blurring lens (10). The use of such a blurring means to blur the image intentionally gives such advantages as each image can be processed performing fewer operations, and apparatus requires fewer processing resources than other state of the art technologies..