Line Sensor Biometric Verification Reducing Memory
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing biometric fingerprint verification devices require substantial memory and a large sensor surface area, making them costly, energy-intensive, and inefficient, especially in portable devices.
Innovation Solution
A biometric device using a line sensor with a limited number of lines captures fractional fingerprint images, which are assembled into partial fingerprint areas for extracting significant data, reducing memory requirements and sensor size by discarding older data as new images are captured.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a complete fingerprint image is stored in memory for verification, then verification accuracy is improved, but memory size and cost increase
Solution Approach 1:
The fingerprint verification process is divided into two distinct phases: enrollment phase where the complete fingerprint image is captured and stored, and verification phase where only partial fingerprint images are captured and compared against stored templates. This segmentation allows the system to maintain high verification accuracy while using minimal memory during the verification operation, as only template data (not full images) needs to be retained.
Solution Approach 2:
The complete fingerprint image is captured and processed in advance during the enrollment phase, creating a compressed template representation that is stored in memory. This preliminary action transforms the full fingerprint data into a condensed form that retains all necessary verification information while occupying minimal storage space, thus resolving the contradiction between accuracy and memory size.
2Area of stationary object
If a line sensor with limited surface area is used, then sensor size and cost are reduced, but the ability to capture complete fingerprint images is worsened
Solution Approach 1:
The system transitions from a static sensor approach (where the sensor must be large enough to capture the entire fingerprint at once) to a dynamic approach where a small line sensor captures sequential partial images as the finger moves across it. This dynamic capture method allows a compact sensor to accumulate complete fingerprint information over time through relative motion between the finger and sensor.
Solution Approach 2:
The system adds the time dimension to the fingerprint capture process. Instead of capturing the complete fingerprint image in a single static snapshot (2D spatial capture), the line sensor captures sequential partial images over time as the finger moves across it. This temporal dimension allows a small sensor surface area to effectively capture complete fingerprint information that would otherwise require a large sensor.
3Reliability
If memory is used to store fingerprint representations, then verification capability is improved, but energy consumption and cost increase
Solution Approach 1:
The system transforms the fingerprint data from full-image format to a compressed template representation, fundamentally changing the data parameter from pixels to extracted features. This parameter change reduces the memory storage requirement from storing complete images (high energy) to storing only essential verification features (low energy), while maintaining verification capability.
Data Source
AI summary
The invention relates to a biometric device (1) for capturing fingerprint information and for extracting significant data from a partial fingerprint area, comprising processing means (8), a line sensor (2) for consecutively capturing fractional fingerprint images from fractional areas of a finger through a relative sliding movement between the finger and the line sensor, means for consecutively storing the fingerprint information in a first memory (6), decision-making means (3) for deciding when the information stored in the first memory constitutes a partial fingerprint area, extraction means (4) for extracting significant data from the partial fingerprint area stored in the first (7), where the captured fractional fingerprint images are stored in the first memory (6) in such a way that several consecutive fractional images are compared with the previously captured images, and are combined together to form a partial fingerprint area which is large enough for the extraction of the significant data, and where the oldest stored fingerprint image data is discarded from the first memory (6) when new fingerprint image data is stored in the first memory (6). In this way, it is possible to use a line sensor with a limited surface and still be able to reduce the memory requirements by extracting significant data representing the fingerprint.