Granted Patent B2

US 12,288,414 · App. 17/577,553 · Granted Apr 29, 2025

Inventors: Asem Othman (Worcester, MA); Richard Tyson (Oxfordshire, GB); Aryana Tavanai (Oxford, GB); Yiqun Xue (Quincy, MA); Andrew Simpson (London, GB)

Assignee: VERIDIUM IP LIMITED

G06V40/1371G06F21/32G06V10/32G06V30/194G06V40/11G06V40/13G06V40/1353G06V40/1359G06V40/1388G06V40/70

Abstract

Claims (45)

1. A method for performing fingerprint recognition, the method comprising:

capturing, by a mobile device having a camera, a storage medium, instructions stored on the storage medium, and a processor configured by executing the instructions, images depicting a plurality of fingers of a subject;

detecting, with the processor using a finger detection algorithm, one or more fingers depicted in at least one of the images;

assessing, with the processor using a convolutional neural network (CNN), a quality score of the at least one image;

processing the at least one image, with the processor using a segmentation algorithm, to identify a respective fingertip segment for at least one finger depicted therein based on the assessed quality score;

measuring, with the processor from the identified respective fingertip segment, one or more features of the at least one finger, wherein the measured feature is a frequency of fingerprint ridges, and wherein the at least one image is scaled based on the frequency of fingerprint ridges and one or more of a prescribed reference frequency and a target resolution;

scaling, the at least one image based on the measured feature; and

storing a biometric identifier, wherein the biometric identifier comprises at least a portion of the scaled image depicting the respective fingertip segment.

2. The method of claim 1 , wherein measuring the frequency comprises: calculating the inverse of a median distance between ridges in a local area of the respective fingertip segment, wherein the distance is measured in a direction perpendicular to a ridge orientation of the at least one finger.

3. The method of claim 1 , wherein the target resolution is a resolution of images stored in a legacy database and that are used for matching against the biometric identifier.

4. The method of claim 1 , wherein the images are captured at a distance from the fingers and thereby depict three-dimensional fingers, and wherein the step of generating the biometric identifier further comprises:

converting an image of the respective fingertip segment to an image of a two-dimensional surface of the fingertip segment using a convolutional neural network (CNN), wherein the CNN is trained on pairs of images that each include a first image of a respective finger captured at a distance using a camera and a second image depicting a two-dimensional surface of the respective finger captured using a contact scanner.

5. A method for performing fingerprint recognition, the method comprising:

capturing, by a mobile device having a camera, a storage medium, instructions stored on the storage medium, and a processor configured by executing the instructions, images depicting a plurality of fingers of a subject;

detecting, with the processor using, a finger detection algorithm, the plurality of fingers depicted in one or more of the images;

processing, with the processor, at least one image using a segmentation algorithm, to identify a respective fingertip segment for one or more fingers depicted in the at least one image;

assessing, with the processor using a convolutional neural network (CNN), a quality score of the at least one image;

extracting, with the processor from the identified respective fingertip segment for the one or more fingers, features of the one or more fingers based on the assessed quality score;

detecting, using a CNN trained to detect minutia points, minutia points within at least one image of the captured images;

determining respective quality scores for the detected minutia points;

selecting a subset of the minutia points for inclusion in the biometric identifier according to the respective quality scores;

generating, with the processor, a biometric identifier including the extracted features and the subset of minutia points; and

storing the generated biometric identifier in the storage medium.

6. The method of claim 5 , wherein the minutia points including fingerprint features.

7. The method of claim 6 , further comprising:

enhancing, the respective fingertip segment.

8. The method of claim 6 , further comprising:

scaling, the respective fingertip segment.

9. A method for performing fingerprint recognition, the method comprising:

capturing, by a mobile device having a camera, a storage medium, instructions stored on the storage medium, and a processor configured by executing the instructions, images depicting a plurality of fingers of a subject;

detecting, with the processor using, a finger detection algorithm, the plurality of fingers depicted in one or more of the images;

processing, with the processor, at least one image using a segmentation algorithm, to identify a respective fingertip segment for one or more fingers depicted in the at least one image;

assessing, with the processor using a convolutional neural network (CNN), a quality score of the at least one image;

extracting, with the processor from the identified respective fingertip segment for the one or more fingers, features of the one or more fingers based on the assessed quality score;

detecting a set of minutia points using a minutia extraction algorithm;

calculating respective quality scores for the minutia points in the set; and

selecting at least a subset of the minutia points in the set for inclusion in the biometric identifier according to the respective quality scores;

generating, with the processor, a biometric identifier including the extracted features and the subset of minutia points; and

storing the generated biometric identifier in the storage medium.

10. The method of claim 9 , wherein the minutia extraction algorithm is a CNN.

11. The method of claim 10 , further comprising:

detecting a background of the at least one image and the respective fingertip segment; and

filtering detected minutia points that correspond to the background from the set.

12. The method of claim 9 , further comprising:

using a plurality of the captured images captured at different respective angles to generate a 3D map of one or more fingers among the plurality of fingers.

Assignments (1)

Continuity (9)

References Cited (161)