Fingerprint recognition device and smart card

Patent metadata

Field	Value
Publication number	US-12249175-B2
Application number	US-202217569059-A
Country	US
Kind code	B2
Filing date	Jan 5, 2022
Priority date	Apr 26, 2021
Publication date	Mar 11, 2025
Grant date	Mar 11, 2025

Abstract

A fingerprint recognition device is provided. The fingerprint recognition device includes an image acquisition module acquiring a fingerprint image including an input fingerprint, a preprocessing module generating a preprocessed image by preprocessing the fingerprint image, a minutiae extraction module extracting coordinates of each of minutiae and orientation points of the input fingerprint from the preprocessed image and a fake detection module receiving regions-of-interest (ROIs), including the coordinates of each of the minutiae or orientation points of the input fingerprint, and determining whether the input fingerprint is a fake by performing learning using the received ROIs.

First claim

What is claimed is:

1. A fingerprint recognition device, comprising:
an image acquisition module configured to acquire a fingerprint image including an input fingerprint;
a preprocessing module configured to generate a preprocessed image by preprocessing the fingerprint image;
a minutiae extraction module configured to extract coordinates of minutiae of the input fingerprint using the preprocessed image, and configured to extract coordinates of orientation points of the input fingerprint using the preprocessed image, wherein the orientation points are points on ridges of the input fingerprint where slopes of lines tangent to the ridges change by more than a threshold value; and
a fake detection module configured to receive regions-of-interest (ROIs), including the coordinates of the minutiae of the input fingerprint and the coordinates of the orientation points of the input fingerprint, and determine whether the input fingerprint is a fake by performing learning using the ROIs.

2. The fingerprint recognition device as claimed in claim 1 , further comprising an ROI extraction module configured to extract the ROIs using the preprocessed image, and configured to input the ROIs to the fake detection module.

3. The fingerprint recognition device as claimed in claim 1 , wherein the ROIs include patches with a predetermined size around the coordinates of the minutiae of the input fingerprint and around the coordinates of the orientation points of the input fingerprint.

4. The fingerprint recognition device as claimed in claim 1 , wherein the fake detection module is configured to perform learning using a convolutional neural network and using the ROIs as input for the convolutional neural network.

5. The fingerprint recognition device as claimed in claim 1 , further comprising a matching module configured to determine whether the input fingerprint matches a previously-registered fingerprint by comparing the input fingerprint with the previously-registered fingerprint based on the minutiae and orientation points of the input fingerprint.

6. The fingerprint recognition device as claimed in claim 5 , wherein the minutiae extraction module is embedded in the matching module.

7. The fingerprint recognition device as claimed in claim 5 , wherein:
the matching module and the minutiae extraction module are configured as separate elements, and
the matching module is configured to receive the minutiae and orientation points of the input fingerprint from the minutiae extraction module.

8. The fingerprint recognition device as claimed in claim 1 , wherein the minutiae extraction module is configured to determine the coordinates of each of the minutiae of the input fingerprint to be different from the coordinates of each of the orientation points of the input fingerprint.

9. The fingerprint recognition device as claimed in claim 1 , wherein the coordinates of the orientation points of the input fingerprint include respective vector information for each of the orientation points.

10. A fingerprint recognition device, comprising:
an image acquisition module configured to acquire a fingerprint image including an input fingerprint;
a preprocessing module configured to generate a preprocessed image by preprocessing the fingerprint image;
a minutiae extraction module configured to extract coordinates of orientation points of the input fingerprint, including vector information for each orientation point, using the preprocessed image, wherein the orientation points are points on ridges of the input fingerprint where slopes of lines tangent to the ridges change by more than a threshold value;
a matching module configured to determine whether the input fingerprint matches a previously-registered fingerprint by comparing the input fingerprint with the previously-registered fingerprint based on the orientation points of the input fingerprint; and
a fake detection module configured to receive regions-of-interest (ROIs), including the orientation points of the input fingerprint, and configured to determine whether the input fingerprint is a fake by performing learning using the ROIs.

11. The fingerprint recognition device as claimed in claim 10 , wherein the minutiae extraction module is further configured to extract coordinates of minutiae of the input fingerprint, which are different from the coordinates of the orientation points of the input fingerprint, using the preprocessed image.

12. The fingerprint recognition device as claimed in claim 10 , further comprising an ROI extraction module configured to extract the ROIs using the preprocessed image, and configured to input the ROIs to the fake detection module.

13. The fingerprint recognition device as claimed in claim 10 , wherein the fake detection module is configured to perform learning using a convolutional neural network and using the ROIs as input for the convolutional neural network.

14. The fingerprint recognition device as claimed in claim 11 , wherein the ROIs include patches with a predetermined size around the coordinates of the minutiae of the input fingerprint and around the coordinates of the orientation points of the input fingerprint.

15. A smart card, comprising:
a fingerprint sensor configured to acquire a fingerprint image including an input fingerprint; and
a system chip configured to generate a preprocessed image by preprocessing the fingerprint image, extract coordinates of minutiae of the input fingerprint and coordinates of orientation points of the input fingerprint using the preprocessed image, extract regions-of-interest (ROIs), including the coordinates of the minutiae of the input fingerprint and the coordinates of the orientation points of the input fingerprint, and determine whether the input fingerprint is a fake by performing learning using the ROIs, wherein the orientation points are points on ridges of the input fingerprint where slopes of lines tangent to the ridges change by more than a threshold value.

16. The smart card as claimed in claim 15 , wherein the coordinates of the orientation points of the input fingerprint include respective vector information for each of the orientation points.

17. The smart card as claimed in claim 15 , wherein the coordinates of each of the minutiae of the input fingerprint are different from the coordinates of each of the orientation points of the input fingerprint.

18. The smart card as claimed in claim 15 , wherein the system chip is configured to determine whether the input fingerprint matches a previously-registered fingerprint by comparing the input fingerprint and the previously-registered fingerprint based on the coordinates of the minutiae of the input fingerprint and the coordinates of the orientation points of the input fingerprint.

19. The smart card as claimed in claim 15 , wherein the ROIs include patches with a predetermined size around the coordinates of the minutiae of the input fingerprint and around the coordinates of the orientation points of the input fingerprint.

20. The fingerprint recognition device as claimed in claim 9 , wherein for each of the orientation points, the respective vector information is vector information of a line tangent to a ridge at that orientation point or vector information of a line orthogonal to the line tangent to the ridge at that orientation point.

Assignees

• Samsung Electronics Co Ltd

Inventors

• Lee Han Sol
• Kim Duck-Soo
• Bae Kwang Hyuk
• Seong Yeol Min
• Song Moon Kyu
• Song Seong Wook
• Lee Jun Seo

Classifications

• G06V40/1388

  using image processing · CPC title

• G06V10/46

  Descriptors for shape, contour or point-related descriptors, e.g. scale invariant feature transform [SIFT] or bags of words [BoW]; Salient regional features (colour feature extraction G06V10/56) · CPC title

• G06V10/82

  using neural networks · CPC title

• G06V40/1353

  Extracting features related to minutiae or pores · CPC title

• G06V40/1365 Primary

  Matching; Classification · CPC title