Fake Finger Detection Using Reflected and Transmitted Light Imaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing fingerprint authentication methods struggle to accurately distinguish fake fingers with thin films, particularly transparent or flesh-colored ones, as they resemble the skin's color, leading to incorrect identification.
Innovation Solution
A device and method that capture both reflected and transmitted light images of a finger, classify pixels based on hue and brightness values, and detect overlapping portions between the two images to determine the presence of foreign substances, improving the accuracy in identifying fake fingers with attached thin films.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a fake finger is detected based on a color of a surface of a finger, then the detection method is simple, but a fake finger with a transparent or flesh-colored thin film attached thereto is hardly discriminated
Solution Approach 1:
The patent transitions from single-color detection to multi-dimensional detection by capturing both reflected light images and transmitted light images. This dimensional expansion allows the system to analyze the finger from multiple optical perspectives simultaneously, revealing characteristics that single-color detection cannot distinguish, particularly for transparent or flesh-colored thin films.
Solution Approach 2:
The patent utilizes changes in optical properties by comparing reflected light and transmitted light images. The classifying unit identifies pixels with different optical characteristics (flesh-colored pixels in reflected light vs. white/bright pixels in transmitted light), and the detecting unit finds overlapping portions between these different optical states, enabling detection of thin films that would be invisible in single-color detection.
2Reliability
If a transparent thin film is attached to a fake finger, then the thin film is difficult to detect visually, but the fake finger can be identified by analyzing overlapping portions between reflected and transmitted light images
Solution Approach 1:
The patent introduces transmitted light imaging as an intermediary method to detect thin films that are invisible in reflected light images. The transmitted light image serves as a mediator that reveals the presence of transparent thin films, which then can be cross-referenced with the reflected light image to confirm fake finger detection.
Solution Approach 2:
The system adds the dimension of transmitted light analysis to the traditional reflected light detection. By capturing images from both optical dimensions and finding overlapping portions, the system can detect transparent thin films that would otherwise be undetectable, thereby improving authentication reliability without significantly increasing operational complexity.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enhances the accuracy of fake finger detection by effectively differentiating between genuine skin and attached thin films, preventing erroneous identifications even with transparent or flesh-colored overlays.
Implementation Method 1
an imaging unit that images a reflected light image and a transmitted light image of an authentication object
Implementation Method 2
an imaging unit that images a reflected light image and a transmitted light image of an authentication object
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The accuracy for determining a fake finger in which a thin film is attached to a surface of a finger is improved. A fake-finger determination device includes an imaging unit 10 that images a reflected light image and a transmitted light image of an authentication object serving as a fingerprint authentication object, a classifying unit 31 that classifies each of the reflected light image and the transmitted light image into a foreground portion having a feature of a skin image and a background portion having no feature of a skin image, based on a hue value of a pixel included in each of the images, a detecting unit 32 that compares the foreground portion of one image of the reflected light image and the transmitted light image with the background portion of the other image to detect an overlapping portion, and a determining unit 33 that determines whether or not a foreign substance is present in the periphery of a finger using the overlapping portion.