# US 12,118,824 B2 · Face detection method and server

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

- Number: US 12,118,824 B2
- Document ID: US12118824B2
- Status: Granted
- Kind: B2
- Application number: 17/169,937
- Filed: 2021-02-08
- Published: 2021-09-16
- Granted: 2024-10-15
- Earliest claimed: 2020-03-12
- Assignees: I-Pro Co., Ltd.
- Applicants: Panasonic I-Pro Sensing Solutions Co., Ltd.
- Inventors: Takashi Kamio; Eisaku Miyata
- Web page: https://patentgrep.com/patent/US12118824B2

## CPC

- G06V 40/173: {face re-identification, e.g. recognising unknown faces across different face tracks}
- G06V 20/40: in video content (extracting overlay text G06V20/62; video retrieval G06F16/70; processing of video elementary streams in video servers H04N21/234; processing of video elementary streams in video clients H04N21/44)
- G06V 40/161: {Detection; Localisation; Normalisation}
- G06V 40/166: {using acquisition arrangements}
- G06V 40/172: {Classification, e.g. identification}
- G07C 9/257: {electronically (G07C9/26 takes precedence)}
- G08B 13/19613: {Recognition of a predetermined image pattern or behaviour pattern indicating theft or intrusion}
- G06V 10/22: by selection of a specific region containing or referencing a pattern; Locating or processing of specific regions to guide the detection or recognition
- G06V 10/25: Determination of region of interest [ROI] or a volume of interest [VOI]
- G06V 20/52: Surveillance or monitoring of activities, e.g. for recognising suspicious objects (recognising microscopic objects G06V20/69)
- G06V 40/50: Maintenance of biometric data or enrolment thereof
- G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G07C: TIME OR ATTENDANCE REGISTERS; REGISTERING OR INDICATING THE WORKING OF MACHINES; GENERATING RANDOM NUMBERS; VOTING OR LOTTERY APPARATUS; ARRANGEMENTS, SYSTEMS OR APPARATUS FOR CHECKING NOT PROVIDED FOR ELSEWHERE
- G08B: SIGNALLING SYSTEMS, e.g. PERSONAL CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS

## Abstract

A face detection method to be implemented by a camera and a server that are communicably connected to each other includes the steps of: acquiring a captured video captured by the camera; detecting a face of one or more persons based on the captured video; generating one or more face cut-out images by cutting out a region of the detected face; counting a number of detection times that a person who is the same as or similar to a person in the face cut-out image is detected in a predetermined detection period; determining a category of the person based on the number of detection times; and outputting the face cut-out image corresponding to the person, the number of detection times corresponding to the person, and the category corresponding to the person in association with each other.

## Claims

### Claim 1

A face detection method to be implemented by a camera and a server that are communicably connected to each other, the face detection method comprising:
acquiring a captured video captured by the camera;
detecting a face of one or more persons based on the captured video;
generating one or more face cut-out images by cutting out a region of the detected face;
determining whether the detected face corresponds to a first detection condition or a second detection condition based on at least one detection item, the detection item not including a detection period or a number of times of detection;
determining, in response to the detected face corresponding to the first detection condition, whether a person who is same as or similar to a person in the face cut-out image is stored in a first database, and: counting a first number of detection times that the person who is the same as or similar to the person in the face cut-out image and stored in the first database is detected in a predetermined detection period; and storing the detected face in the first database in association with the first number of detection times, with each of detected faces being stored in the first database in association with a corresponding one of first numbers of detection times in response to the detected faces corresponding to the first detection condition;
determining, in response to the detected face corresponding to the second detection condition, whether the person who is the same as or similar to the person in the face cut-out image is stored in a second database, and: counting a second number of detection times that the person who is the same as or similar to the person in the face cut-out image and stored in the second database is detected in the predetermined detection period; and storing the detected face in the second database in association with the second number of detection times, with each of detected faces being stored in the second database in association with a corresponding one of second numbers of detection times in response to the detected faces corresponding to the second detection condition;
determining a category of the person based on one of the first number of detection times or the second number of detection times; and
outputting the face cut-out image corresponding to the person, the one of the first number of detection times or the second number of detection times corresponding to the person, and the category corresponding to the person in association,
wherein the face detection method further comprises:
registering persons in the first database based on face cut-out images from captured images, the persons registered in the first database being suspicious persons; and
registering persons in the second database based on printed images of the persons, the persons registered in the second database being registered customers.

### Claim 2 (depends on 1)

The face detection method according to claim 1, further comprising:
acquiring the first detection condition; and
counting the first number of detection times of the person in the face cut-out image when the person in the face cut-out image is a detection target satisfying the first detection condition; and
determining the category of the person to be a suspicious person when the first number of detection times is a predetermined number of times or more.

### Claim 3 (depends on 2)

The face detection method according to claim 2, wherein
the first detection condition is for detecting a person having the category of a suspicious person.

### Claim 4 (depends on 2)

The face detection method according to claim 2, further comprising:
registering the face cut-out image of the person having the category of the suspicious person as a blacklist face image on each of a plurality of blacklist face images registered in advance by a user.

### Claim 5 (depends on 4)

The face detection method according to claim 4, wherein
each of the plurality of blacklist face images is matched with the face cut-out image to determine whether the blacklist face image that is same as or similar to the person in the face cut-out image is registered in the first database, and
when the face cut-out image that is the same as or similar to the blacklist face image is registered in the first database, an alert notification screen is generated and output.

### Claim 6 (depends on 1)

The face detection method according to claim 1, further comprising:
acquiring the first detection condition;
counting the first number of detection times of the person in the face cut-out image when the person in the face cut-out image is not a detection target satisfying the first detection condition; and
determining the category of the person to be a loyal customer when the first number of detection times is a predetermined number of times or more.

### Claim 7 (depends on 6)

The face detection method according to claim 6, further comprising:
registering the face cut-out image of the person having the category of the loyal customer as an important person face image on each of a plurality of important person face images registered in advance by a user.

### Claim 8 (depends on 7)

The face detection method according to claim 7, wherein
each of the plurality of important person face images is matched with the face cut-out image to determine whether the important person face image that is same as or similar to the face cut-out image is registered in the second database, and
when the face cut-out image that is the same as or similar to the important person face image is registered in the second database, an alert notification screen is generated and output.

### Claim 9 (depends on 1)

The face detection method according to claim 1, further comprising:
acquiring the second detection condition;
counting the second number of detection times of the person in the face cut-out image when the person in the face cut-out image is a detection target satisfying the second detection condition; and
determining the category of the person to be a loyal customer when the second number of detection times is a predetermined number of times or more.

### Claim 10 (depends on 9)

The face detection method according to claim 9, wherein
the second detection condition is for detecting a person having the category of a loyal customer.

### Claim 11 (depends on 9)

The face detection method according to claim 9, further comprising:
registering the face cut-out image of the person having the category of the loyal customer as an important person face image on each of a plurality of important person face images registered in advance by a user.

### Claim 12 (depends on 11)

The face detection method according to claim 11, wherein
each of the plurality of important person face images is matched with the face cut-out image to determine whether the important person face image that is the same as or similar to the face cut-out image is registered in the second database, and
when the face cut-out image that is the same as or similar to the important person face image is registered in the second database, an alert notification screen is generated and output.

### Claim 13 (depends on 1)

The face detection method according to claim 1, wherein
each of a plurality of non-detected face images registered in advance by a user is matched with the face cut-out image to determine whether a non-detected face image that is same as or similar to the face cut-out image is registered, and
when the face cut-out image that is the same as or similar to the non-detected face image is registered, generation of an alert notification screen is omitted.

### Claim 14 (depends on 1)

The face detection method according to claim 1, wherein
when the person in the face cut-out image is re-detected within the detection period, a detection period of the person in the face cut-out image is extended by the detection period from a timing of the re-detection.

### Claim 15

A computer readable non-transitory storage medium on which a face detection program executed by a server communicably connected to a camera is stored, the face detection program causing the server to execute a process including:
acquiring a captured video captured by the camera;
detecting a face of one or more persons based on the captured video;
generating one or more face cut-out images by cutting out a region of the detected face;
determining whether the detected face corresponds to a first detection condition or a second detection condition based on at least one detection item, the detection item not including a detection period or a number of times of detection;
determining, in response to the detected face corresponding to the first detection condition, whether a person who is same as or similar to a person in the face cut-out image is stored in a first database, and; counting a first number of detection times that the person who is the same as or similar to the person in the face cut-out image and stored in the first database is detected in a predetermined detection period; and storing the detected face in the first database in association with the first number of detection times, with each of detected faces being stored in the first database in association with a corresponding one of first numbers of detection times in response to the detected faces corresponding to the first detection condition;
determining, in response to the detected face corresponding to the second detection condition, whether the person who is the same as or similar to the person in the face cut-out image is stored in a second database, and: counting a second number of detection times that the person who is the same as or similar to the person in the face cut-out image and stored in the second database is detected in the predetermined detection period; and storing the detected face in the second database in association with the second number of detection times, with each of detected faces being stored in the second database in association with a corresponding one of second numbers of detection times in response to the detected faces corresponding to the second detection condition;
determining a category of the person based on one of the first number of detection times or the second number of detection times; and
outputting the face cut-out image corresponding to the person, the one of the number of first detection times or the second number of detection times corresponding to the person, and the category corresponding to the person in association,
wherein the process further includes:
registering persons in the first database based on face cut-out images from captured images, the persons registered in the first database being suspicious persons; and
registering persons in the second database based on printed images of the persons, the persons registered in the second database being registered customers.

### Claim 16

A server communicably connected to a camera, the server comprising:
an interface that acquires a captured video captured by the camera;
a processor that
detects a face of one or more persons based on the captured video acquired by the interface,
generates one or more face cut-out images by cutting out a region of the detected face,
determines whether the detected face corresponds to a first detection condition or a second detection condition based on at least one detection item, the detection item not including a detection period or a number of times of detection,
determines, in response to the detected face corresponding to the first detection condition, whether a person who is same as or similar to a person in the face cut-out image is stored in a first database, and; counts a first number of detection times that the person who is the same as or similar to the person in the face cut-out image and stored in the first database is detected in a predetermined detection period; and stores the detected face in the first database in association with the first number of detection times, with each of detected faces being stored in the first database in association with a corresponding one of first numbers of detection times in response to the detected faces corresponding to the first detection condition,
determines, in response to the detected face corresponding to the second detection condition, whether the person who is the same as or similar to the person in the face cut-out image is stored in a second database, and; counts a second number of detection times that the person who is the same as or similar to the person in the face cut-out image and stored in the second database is detected in the predetermined detection period; and stores the detected face in the second database in association with the second number of detection times, with each of detected faces being stored in the second database in association with a corresponding one of second numbers of detection times in response to the detected faces corresponding to the second detection condition, and
determines a category of the person based on one of the first number of detection times or the second number of detection times; and
an output that outputs the face cut-out image corresponding to the person, the one of the first number of detection times or the second number of detection times corresponding to the person, and the category corresponding to the person in association,
wherein the processor further:
registers persons in the first database based on face cut-out images from captured images, the persons registered in the first database being suspicious persons; and
registers persons in the second database based on printed images of the persons, the persons registered in the second database being registered customers.