# US 2023/0144689 A1 · Stripe pattern image collating device, stripe pattern collating method, and computer-readable medium storing program thereof

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2023/0144689 A1
- Document ID: US20230144689A1
- Status: Published application
- Kind: A1
- Application number: 17/913,026
- Filed: 2020-03-27
- Published: 2023-05-11
- Earliest claimed: 2020-03-27
- Assignees: NEC Corporation
- Applicants: NEC Corporation
- Inventors: Akira HIROKAWA; Tatsuya Shimahara
- Web page: https://patentgrep.com/patent/US20230144689A1

## CPC

- G06V 40/1359: {Extracting features related to ridge properties; Determining the fingerprint type, e.g. whorl or loop}
- G06T 5/80: Geometric correction
- G06V 10/806: {of extracted features}
- G06T 2207/20044: Skeletonization; Medial axis transform
- G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL

## Abstract

A stripe pattern image collating device according to the example embodiment includes a feature extracting unit that extracts a feature point and a skeleton from a first stripe pattern image and a second stripe pattern image in which a stripe pattern is formed of ridges, and generates feature point data and skeleton data. A skeleton collating unit that collates two sets of pieces of the feature point data and pieces of the skeleton data that are extracted from each of the first stripe pattern image and the second stripe pattern image, and calculates a collation score. An image analyzing unit that analyzes the second stripe pattern image with respect to an area in which an opposite feature point pair of the first stripe pattern image exists, calculates an image analysis score, and corrects the collation score.

## Claims

### Claim 1

A stripe pattern image collating device comprising:
a feature extracting unit configured to extract a feature point and a skeleton from a first stripe pattern image and a second stripe pattern image in which a stripe pattern is formed of ridges, and generate feature point data and skeleton data;
a skeleton collating unit configured to collate two sets of pieces of the feature point data and pieces of the skeleton data that are extracted from each of the first stripe pattern image and the second stripe pattern image, and calculate a collation score; and
an image analyzing unit configured to analyze the second stripe pattern image with respect to an area in which an opposite feature point pair of the first stripe pattern image exists, calculate an image analysis score, and correct the collation score.

### Claim 2 (depends on 1)

The stripe pattern image collating device according to claim 1, wherein the image analyzing unit extracts ridge direction data and ridge interval data from the skeleton data of the first stripe pattern image, and emphasizes, by using the ridge direction data and the ridge interval data, a ridge contrast of the second stripe pattern image whose image distortion has been corrected by using associating skeleton data of a result of skeleton collation.

### Claim 3 (depends on 2)

The stripe pattern image collating device according to claim 2, wherein the image analyzing unit limits inside of an associating skeleton of the second stripe pattern image to a processing area when an outer skeleton determined to be outside the opposite feature point pair of the first stripe pattern image is an associating skeleton, and extracts the ridge direction data and the ridge interval data in the processing area.

### Claim 4 (depends on 3)

The stripe pattern image collating device according to claim 3, wherein the image analyzing unit extracts a skeleton and a feature point, as a virtual skeleton and a virtual feature point each, from the processing area limited to inside of an associating skeleton of the second stripe pattern image when an outer skeleton determined to be outside the opposite feature point pair of the first stripe pattern image is an associating skeleton, and calculates the image analysis score by collating the virtual feature point with the opposite feature point pair.

### Claim 5 (depends on 4)

The stripe pattern image collating device according to claim 4, wherein the image analyzing unit calculates the image analysis score having a positive value when determining that the virtual feature point is a paired feature point of the opposite feature point pair.

### Claim 6 (depends on 4)

The stripe pattern image collating device according to claim 4, wherein the image analyzing unit analyzes the virtual skeleton and calculates the image analysis score when determining that the virtual feature point is not a paired feature point of the opposite feature point pair.

### Claim 7 (depends on 6)

The stripe pattern image collating device according to claim 6, wherein analysis of the virtual skeleton is performed by comparison with a skeleton of the second stripe pattern image extracted by the skeleton collating unit.

### Claim 8 (depends on 7)

The stripe pattern image collating device according to claim 7, wherein the image analyzing unit sets, as a collating target, an inner skeleton inside an associating skeleton of the second stripe pattern image when an outer skeleton determined to be outside the opposite feature point pair of the first stripe pattern image is an associating skeleton, and calculates the image analysis score according to a ratio of the number of pixels in which the inner skeleton does not exist within a predetermined range from each pixel on the virtual skeleton, to the number of pixels of the entire virtual skeleton.

### Claim 9

A stripe pattern image collating method comprising:
extracting a feature point and a skeleton from a first stripe pattern image and a second stripe pattern image in which a stripe pattern is formed of ridges, and generating feature point data and skeleton data;
collating two sets of pieces of the feature point data and pieces of the skeleton data that are extracted from each of the first stripe pattern image and the second stripe pattern image, and calculating a collation score; and
analyzing the second stripe pattern image with respect to an area in which an opposite feature point pair of the first stripe pattern image exists, calculating an image analysis score, and correcting the collation score.

### Claim 10

A computer-readable medium storing a program causing a computer to execute processing of:
extracting a feature point and a skeleton from a first stripe pattern image and a second stripe pattern image in which a stripe pattern is formed of ridges, and generating feature point data and skeleton data;
collating two sets of pieces of the feature point data and pieces of the skeleton data that are extracted from each of the first stripe pattern image and the second stripe pattern image, and calculating a collation score; and
analyzing the second stripe pattern image with respect to an area in which an opposite feature point pair of the first stripe pattern image exists, calculating an image analysis score, and correcting the collation score.
