WO2001075774A1

WO2001075774A1 - Method for person identification

Info

Publication number: WO2001075774A1
Application number: PCT/RU2001/000010
Authority: WO
Inventors: Andrei Vladimirovich Sviridenko; Jury Vitalievich Morzeev; Sergei Olegovich Novikov
Original assignee: Sviridenko Andrei Vladimirovic; Jury Vitalievich Morzeev; Sergei Olegovich Novikov
Priority date: 2000-04-04
Filing date: 2001-01-12
Publication date: 2001-10-11
Also published as: RU2175148C1

Abstract

The invention relates to the authorized access systems, more specifically to methods for person identification with the aid of a face pattern analysis. The inventive method is characterized by developing a standards set using an optical sensor during determined movements of a head of a person to be identified. Said standards are embodied in the form of discrete two-dimensional information fields containing an information about characteristic phases of the face turn of said person. A horizontal band comprising eyes being emphasized on a white and black face image with the aid of anthropometric data, the face phase of movement is estimated by analyzing the image asymmetry within a visible band. The estimated results are used as a key information for searching in a computer memory the face standards possessing the phases close to the estimated phase and said parts being compared with standard parts identified in the computer memory, a quantitative measure of a similarity thereof is defined.

Description

The method of identifying a person.

The invention is available to the auto-access systems, as well as

Particular ways of identifying people on the basis of the analysis of the structure

His face.

The methods of identifying a person on the basis of analysis are known.

structures of his person on the video image, based on the allocation and

analysis of the characteristic features of the face - the eye, the eyes, the nose, the nose and the face. (see, for example,

Identification methods known for valuation are also known.

an interactive character of a person, in particular through the use and pricing

a set of προφile of rash in the field of vision of the face. (see, for example, ρ. δϊϊsνϊsη eϊ aϊ.,

1987 Ορϊϊsaϊ δasϊegu οg ^"Αtegϊsa," ονν-άϊteηzyuηaϊ ρgοseοΗge τοg ϊe

sηagasϊeπζaϊyuη οι ^* 1ιshηaη Gasez ", ρρ.519-524).

Disposal of the first aid is a temporary and

computing costs of identification and high cost

The facilities are not necessary for the implementation of the process due to the high

Distribution of goods, supplies to an external sensor, equality and

the scope of the coverage of the identity object.

Wastes of the second method are low reliability.

Recognition, unstability of recognition results κ r

Variable variability of the image. 2

For the closest technical essence of the means of identification,

selected as a part of the method, is a method that implies

a violation of the presence of a person in an image that has been obtained with the help of

video cameras, detection of the location of the eyes on the image, the situation

OTHER FACILITIES OF THE FACE OF THE NEGATIVE EYE, DIFFERENCE OF DIFFERENT

SPECIAL FEATURES FOR SOME PARTICIPANTS OF THE PERSON

identities with standards and comparison with others. (see, for example,

ΡаΙеη 5, 164,992, ηον.17,1992).

The disadvantages of production are low reliability and speed.

Recognition and high cost of equipment for sale

recognition.

Β The invention is tasked with enhancing the reliability of recognition,

increase in recognition rate and decrease in the cost of equipment for

identification of a person on the basis of analysis of the structure of his person. These tasks

Decided by the identification of the person in the systems of the author

access to the basis of the analysis of the structure of his person, including the operation

Training for a system with a primary occupation of a group of persons

people with proper identifications, different types of persons,

the identification of char- acteristic characteristics from the materials and the recording of them in memory

computing devices, recognition services received with

With the help of an optical sensor on a real scale of time in a format

in ^χ οdnοgο usτροysτva κοmπyuτeρa videοizοbρazheniya, sοdeρzhaschegο, πο _

3 at the very least, one person, proper identification, analysis

video images with computing equipment for highlighting

persons of a person and the identification of disease signs of an identified person,

comparing these familiarities with the familiarity of components stored in memory

computing devices for making solutions for identification and

system training, and the operation of the learning system of the automated system

access, in a companion set of electronic devices using an optical

a child in the process of accomplishing separate movements of the head man,

Proper identification, such as a collection of standards

Provides a special two-dimensional information field,

containing information about the face phase of the person is mentioned

people, from this information one of the known methods allocate part

in the form of the area of the information field, the corresponding to the face is remembered

a man, inside a separate area, produces a normalization operation

The area, which is located in the premises, is in addition to a second or additional accessory.

by averaging the points of the highlighted area,

The person who is facing the person using masks with different linear

By the size and the next point subtraction of the first translation

ADDITIONAL PRODUCTS FROM THE SECONDARY EXTERNAL DISPLAY WITH

By acquiring such a normalized user equipment

parts of the two-dimensional informational field that are related to the face are remembered

a man, a black and white image of this area, consisting of two 4 gradations of brightness after this noise filtering is done in black and white

image using the anti-white data on black and white

the face image emits a horizontal area, including

imaging of the eyes, evaluates the phase of facial movement by analyzing asymmetry

Images inside the allocated bandwidth, evaluation results are used in

Key information for searching in memory on a computer

Devices of persons with phases close to the valued phase, break up

black-and-white image with two gradations of brightness in parts, for each

parts determine the measure of its significance in the form of a weighting factor, with

the use of one of the known methods produces a comparison of these

parts with relevant parts of the persons found in memory

computing devices and allocates quantitative measure χ

on the basis of the results of this quantitative

DATA AND TAKING INTO ACCOUNT THE DYNAMIC RESPONSE

with ^χ οdsτve, προizvοdyaτ inτegρalnuyu οtsenκu meρy face sχοdsτva with κazhdym of

etalons, selects etalon, in comparison with the best received the best

an integrated assessment of the benefits, based on one of the famous kriteriev

Takes a decision on the inclusion of additional components in the kit

Ethics, Properly Identified Person for Learning

SYSTEMS, AND THEREFORE, MEMORY RECORDING IN THE MEMORY OF COMPUTER DEVICES

reference points obtained from optional products already 5 identities made in the process of identification and in

The following points in time.

The difference in method is to the effect that the collection of electrical components is interrupted.

the use of an optical sensor in the process of settling

the movements of the head person, the proper identification, such as

that a large number of components are supplied with double discrete

INFORMATION FIELDS CONTAINING INFORMATION FOR PHASES

A person is informed of this person, from this information, one of

known facilities allocate part in the form of the area of the information field,

The outward person is impressed by the inside, and the highlighted area

The company manufactures the normalization of the zone, which is concluded in the process

Front and rear accessory equipment by averaging the speed

Dotted area that is relevant to a person’s face using

mask with different linear sizes and the next point

Subtracting an Advanced Image from the Secondary

ADDITIONAL IMAGES, INCLUDED WITH THIS SOFTWARE NORMALIZED

A better view of a part of a two-dimensional information field,

A flashing person in front of the implanted person is affected by a black and white image.

this area, which contains two grades of locomotion, after this occupation

filtering noise on a black and white normalized image, on

the black-and-white image of the face emits a horizontal band,

incl. image of the eyes, for example, using 6 anthropological data evaluate the phase of facial movement through analysis

asymmetries of the image inside the allocated bandwidth, results of the assessment

use as a key information for searching in memory

computing devices of persons with phases close to valued

Phase breaks up a black and white image with two gradations of brightness on

parts, for each part, allocate to the measure of its significance in the form of weight

The coefficient, using one of the well-known methods of production

comparing these parts with the corresponding parts of persons

found in memory of computing devices and allocates

a quantitative measure of their benefit, on the basis of the results obtained in the result

this quantitative data and taking into account the dynamically tuned

Making a decision to solve a problem, make an integral assessment

measures of disappearance of a person with each of the ele- ments, select the etalon, in comparison with

On the other hand, the best integrated assessment of the price was obtained on the basis of

One of the famous crushers accepts a decision on

additional accessories in a set of electronic accessories

to an identifiable person in order to learn the system, by doing so

Record in memory of computing devices of components received from

ADDITIONAL IMAGES ALREADY IDENTIFIED PERSONS MAKED IN

In the process of identification and at the following times.

The reliability of recognition is achieved through the partition

images with a detached person in parts and the introduction of weights 7 factors for every part, and also due to the use of ^χ

above the floating ships.

Enhanced facial access and identification

the person in the proposed method is reached at the expense of the exact distribution

Real-time face phase, filtering

casinos exploring the information channels of an optical sensor in

Combined with automatic training and adaptive education

Pursuing a solution for identification.

Decrease in the cost and expansion of the area of application of systems,

the identities used are accessible at the expense of the opportunity

use of less specific optical sensors, less rigid

Requirements for technical processors of computing devices.

The increase in recognition rate is achieved at the expense of

Preliminary assessment of the phase of the sale of the property in conjunction with

Accessible accessibility for people who are eligible

identification, and use of floating solutions

good luck.

The proposed method may be used, for example, in the system

AUTHORIZED ACCESS TO PRIVACY INFORMATION CLEARED IN

MEMORY OF COMPUTER DEVICES.

The proposed method for identifying is illustrated

by the investigation of operations shown in Fig. 1, where: .

8

1. The operation of the property of persons entitled to ^χ

Identification

2. The scope of the invention in the investigative aspects

time;

3. Reproduction of the image in the digital

electrical signal;

4. Memorizing a digital electrical signal in

memorizing device of the computer;

5. The operation of isolation in a digital electrical signal of a part,

a person identifiable;

6. The procedure for detecting facial signs of a person;

7. Comparison of identified identities of a person with identities of reference

persons;

8. Acceptance of a decision on identification;

9. Production management;

Yu.Peraction of the implementation of the oppositely operating,

Automatically accessed by auto

Areas of data, for example, components of the operating system

computing devices;

P. TAKING A DECISION TO PREPARE A SYSTEM FOR IMPLEMENTING A SYSTEM

identification above and above higher

quantitatively integrated appraisal of a small appraisal 9 assignments corresponding to the required quanti-

level of education;

12.0 training;

ΙЗ.Οperformation of the signal distribution.

In the first place, the implementation of the method is limited to

Residence of persons with proper identification with

camcorders and computing devices in the process of tracking

people who are subsequently identified as identifying

By locating the electronic display. When tracking ekrapanom

A person who is a lucid person can take a few steps to

vertical and horizontal areas, which are quickly detected

The camcorder and the video are converted into digital information containing

distinctive facial features. Аче Qualitative specifications for

Highlighting facial areas uses signs of facial movement (displacement,

direction and τ.π.), located in the field of vision of the video camera, linear

Sizes of the face and its individual parts and τ.π. From the general volume of digital

information identifies parts that are relevant to these attributes, and

write to the memory of the computing device, having given them the status

reference with storage addresses. We offer you a wide range of products.

special two-dimensional information fields containing

INFORMATION ABOUT CHARACTERISTIC PHASES OF PERSONALIZATION OF THE FACE; 10

When a person sees his video camera

The image is taken into account in the investigation of the time (operation

2). Further, the received video information is encouraged to convert to

digital electrical signal. The digital information is analyzed with

With the help of computing devices and remember. 5th 5

highlight information related to the area of the face.

Allocation of products in accordance with the algorithm provided by

on fig.2.

The method of highlighting the area of the image pertaining to the face,

Includes the following operations:

14 delay of a digital signal of an electric signal;

15 highlighting the user interface of the analyzed images

One of the well-known methods, for example, by dividing

the primary output frequency of the signal is at a high frequency;

16 Preparation of a variety of images corresponding to

moving counters by means of a point-by-point subtraction of speed

point with identical components;

17 Search sites that share the upper boundary of the region

movements corresponding to the upper parts of the head city;

18 division of the side and lower border of the area of movement.

There is a separate area for normalizing the area,

which is in the process of accessing a second and second optional 11 images by averaging the points of the selected area,

The person who is facing the person using masks with different linear

By the size and the next point subtraction of the first translation

ADDITIONAL PRODUCT FROM THE SECONDARY EXTERNAL DISPLAY.

Improved user-friendly access to the device.

two-sided information field, which is related to the face, is remembered

man. Then, from the normalized political picture, one of the

known methods of binarization are affected by a black and white image of this

The area comprising two grades of land. After this issue

filtering noise on a black-and-white image, and using

the anthropological data on a black-and-white image of a person emit

horizontal area, including imaging of the eyes. Beat the black

a white image of a face with two gradations of detail in parts, for each

parts determine the measure of its significance in the form of a weighting factor. Ηa

Operation 6 identifies distinctive signs of the person calling, speaking

First, find the ones closest to your property from the sets,

Stored in memory of computing devices and, secondly, to produce

comparison of π из с ед ед ед ед ед с с с с с с с identities Κ quality

The results of the estimation of asymmetry are used

images in a closet containing the eye and the womb. Evaluation results in the form of

the numerical factor is used as a reference value

Searching for persons in memory. Ачеqualities of knowledge

P _

The 12th type uses the results of comparison with the 7th standard, which is

It is produced using one of the famous kritera, for example,

Criterion of Fischer, which shares an integrated evaluation with ^χ equipment

An image of a person with his or her base on the main premises

OPERATING FUNCTIONS BY DEFAULT PROCESSING

original physical activity, provided that the maximum is reached

Functions.

Using one of the known methods, for example, Euclidean

Measures of likelihood that make a comparison of these previously used parts of the face with

by the corresponding parts of the persons found in the memory

computing devices and allocates quantitative measure χ

on the basis of the results of this quantitative

DATA AND TAKING INTO ACCOUNT THE DYNAMIC RESPONSE

Immediately, an integral assessment of the cost of a person’s death with each of

etalons, selects etalon, in comparison with the best received the best

Integrated assessment of the benefits. Аче How to do the solution

use the cassettes of the Neiman-Piroson according to the agreement

the probability of the correct identification of the property is maximized

a person's face and a given probability of an erroneous matter.

Decision identification (Operation 8) is accepted if received

The quantitatively integrated appraisal is higher than the corresponding rate.

a certain quantitative level of probability. thirteen

In order to increase the reliability of the recognition in the proposed method

The training operation is available 12. On the basis of one of the known ^χ

The clergy, in accordance with this decision 11, are included.

ADDITIONAL EQUIPMENT IN A SET OF ELECTRONIC COMPANIES

to an identified person. When this is done, a recording is made in memory

computing devices derived from optional

images already identified in the process

Identification and at the following times.

The proposed method can find a wide range of applications in the system.

auto access.

The method may be referred to as accessed by the private system.

to sources of information, existing systems, in monitoring systems in

Public places, for example, for the search for people in the region.

Claims

14

ΦΟΡΜULΑ

The method of identifying a person in an automated access system

on the basis of analysis of the structure of his person, including the operation of training

systems with a primary population of people

Appropriate identifications reflecting different phases of persons by identifying

The specifications of the components and the recording of them in memory of the computing

Devices, recognition and reception with the help of an optical device

real-time sensor in an external device format

The video computer system, which is available in the first place, is a single person

person with proper identification, video analysis with

By means of a computing device for distinguishing a person’s face and

identifying disease signs of an unidentified person, a comparison of these

identities with the identities of the components that are stored in memory of a computer

DEVICES FOR FORMING SOLUTIONS FOR IDENTIFICATION AND LEARNING

systems, and the operation system for learning an automated access system

characterized by the fact that a set of electronic components is used with the use of an optical

a child in the process of accomplishing separate movements of the head man,

Proper identification, such as a collection of standards

Provides a special two-dimensional information field,

containing information about the face phase of the person is mentioned

people, from this information one of the known methods allocate part

in the form of the area of the information field, the corresponding to the face is remembered 15 people, including a dedicated area for normalization

by averaging the points of the highlighted area,

The person who is facing the person using masks with different linear

By the size and the next point subtraction of the first translation

ADDITIONAL PRODUCTS FROM THE SECONDARY EXTERNAL DISPLAY WITH

By acquiring such a device from a two-part unit

information field related to a person who is remembered for being impaired

a black and white image of this area, comprising two gradations of a locomotion,

After this, filter the noise on a black and white normalized

image using the anti-white data on black and white

the face image emits a horizontal area, including

Images inside the allocated bandwidth, evaluation results are used in

Key information for searching in memory on a computer

Devices of persons with phases close to the valued phase, break up

black-and-white image with two gradations of brightness in parts, for each

the use of one of the known methods produces a comparison of these

parts with relevant parts of the persons found in memory

computing devices and allocates quantitative measure χ

on the basis of the results of this quantitative 16 data and taking into account the dynamic dynamics of the decisions taken

Immediate appraisal of an assessment of the cost of a person’s death with each of

etalons, selects etalon, in comparison with the best received the best

an integrated assessment of the benefits, based on one of the famous kriteriev

Takes a decision on the inclusion of additional components in the kit

Ethics, Properly Identified Person for Learning

SYSTEMS, AND THEREFORE, MEMORY RECORDING IN THE MEMORY OF COMPUTER DEVICES

reference points obtained from optional products already

identities made in the process of identification and in

The following points in time.