CN112633226A - Face verification method and system based on quaternion fractional order pseudo-Zernike moment - Google Patents

Face verification method and system based on quaternion fractional order pseudo-Zernike moment Download PDF

Info

Publication number
CN112633226A
CN112633226A CN202011622736.4A CN202011622736A CN112633226A CN 112633226 A CN112633226 A CN 112633226A CN 202011622736 A CN202011622736 A CN 202011622736A CN 112633226 A CN112633226 A CN 112633226A
Authority
CN
China
Prior art keywords
fractional order
order pseudo
zernike
face image
quaternion
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202011622736.4A
Other languages
Chinese (zh)
Other versions
CN112633226B (en
Inventor
俞铭
褚红健
王声柱
刘琴
葛淼
周金国
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Guodian Nanjing Automation Co Ltd
Original Assignee
Guodian Nanjing Automation Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Guodian Nanjing Automation Co Ltd filed Critical Guodian Nanjing Automation Co Ltd
Priority to CN202011622736.4A priority Critical patent/CN112633226B/en
Publication of CN112633226A publication Critical patent/CN112633226A/en
Application granted granted Critical
Publication of CN112633226B publication Critical patent/CN112633226B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/16—Human faces, e.g. facial parts, sketches or expressions
    • G06V40/168—Feature extraction; Face representation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10—Complex mathematical operations
    • G06F17/15—Correlation function computation including computation of convolution operations
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10—Complex mathematical operations
    • G06F17/16—Matrix or vector computation, e.g. matrix-matrix or matrix-vector multiplication, matrix factorization
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/22—Matching criteria, e.g. proximity measures

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Mathematical Physics (AREA)
  • Pure & Applied Mathematics (AREA)
  • Mathematical Optimization (AREA)
  • Mathematical Analysis (AREA)
  • Computational Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • Oral & Maxillofacial Surgery (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Health & Medical Sciences (AREA)
  • Software Systems (AREA)
  • Databases & Information Systems (AREA)
  • Algebra (AREA)
  • Computing Systems (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Multimedia (AREA)
  • Evolutionary Computation (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Artificial Intelligence (AREA)
  • Human Computer Interaction (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • General Health & Medical Sciences (AREA)
  • Evolutionary Biology (AREA)
  • Image Analysis (AREA)
  • Collating Specific Patterns (AREA)

Abstract

The invention discloses a face verification method based on quaternion fractional order pseudo-Zernike moments, which comprises the following steps: collecting a color face image of an identity person to be verified; dividing the color face image into three channels of RGB (red, green and blue)R、IG、IBThree images; respectively carrying out fractional order pseudo-Zernike moment transformation on the three images to obtain three fractional order pseudo-Zernike moment feature matrixes; adopting a quaternion fractional order pseudo-Zernike moment transformation rapid calculation method to transform the three fractional order pseudo-Zernike moment feature matrices into quaternion fractional order pseudo-Zernike moment feature matrices so as to obtain an identification feature matrix of the color face image; using K-NN algorithm to compare the recognition feature matrix of the color face image with the recognition feature matrix of the color face imageAnd matching the pre-established recognition feature matrixes in the face image feature library to finish face verification. The color face features extracted by using the quaternion fractional order pseudo-Zernike moment have higher robustness, and the provided rapid calculation algorithm effectively improves the verification rate of the face verification model.

Description

Face verification method and system based on quaternion fractional order pseudo-Zernike moment
Technical Field
The invention belongs to the field of face recognition, and particularly relates to a face verification method and a face verification system based on quaternion fractional order pseudo-Zernike moments, which can be used for log-in verification and attendance checking of workers and shifts in a rail transit maintenance management system and can perform face recognition quickly and robustly.
Background
Compared with the verification mode of the traditional rail transit maintenance management system, the human face verification has better safety, confidentiality and convenience. The face recognition technology begins in the 60's of the 20 th century, and has very wide application in the fields of personal authentication and appraisal and criminal investigation. In recent 20 years, experts and scholars propose a plurality of face recognition algorithms, but the existing face recognition algorithms have the problems of low recognition speed and low algorithm robustness when the illumination condition is poor and the facial expression and posture change is rich. The face recognition technology is divided into two categories, namely, extraction based on key points and extraction based on features, and the key of the face recognition algorithm based on feature extraction is whether the extracted features have stronger robustness and faster calculation speed. Transform domain features are more robust against interference than spatial domain features and a small number of transform domain coefficients can carry a larger amount of information. The fractional order transformation is a whole process of orthogonal transformation which can show the transformation of a signal from a time domain to a frequency domain, and the fractional order transformation not only contains the frequency domain characteristics of the signal, but also contains the time domain characteristics of the signal and has wide application in digital image processing. In the last two decades, color image quaternion representation has been widely used in color image processing to account for variations in lighting conditions and the effects of color on recognition. The color image quaternion representation mode is provided to solve the influence of illumination condition change and color on an image identification algorithm to a great extent.
Disclosure of Invention
The invention aims to provide a face verification method and system based on a quaternion fractional order pseudo-Zernike moment, and aims to solve the problems that in the prior art, a traditional rail transit maintenance management system user login and attendance face verification module is low in feature extraction speed and robustness.
In order to achieve the purpose, the invention adopts the following technical scheme:
on the other hand, the invention provides a face verification method based on quaternion fractional order pseudo-Zernike moments, which comprises the following steps:
collecting a color face image of an identity person to be verified;
dividing the color face image into three channels of RGB (red, green and blue) IR、IG、IBThree images;
will IR、IG、IBPerforming fractional order pseudo-Zernike moment transformation on the three images respectively to obtain three fractional order pseudo-Zernike moment feature matrices;
transforming the fractional order pseudo-Zernike moment feature matrix into a quaternion fractional order pseudo-Zernike moment feature matrix by adopting a quaternion fractional order pseudo-Zernike moment transformation rapid calculation method to obtain an identification feature matrix of the color face image;
and matching the recognition characteristic matrix of the color face image with a recognition characteristic matrix in a pre-established face image characteristic library by using a K-NN algorithm to finish face verification.
Further, the construction algorithm of the fractional order pseudo-Zernike moments is defined as follows:
for a two-dimensional continuous function f (r, theta) in polar coordinates, the fractional order pseudo-Zernike moments are:
Figure BDA0002872674510000031
wherein α is a fractional order; n and m are nonnegative integers, which are the order and the repetitiveness of the fractional order pseudo Zernike moment respectively;j is a unit of an imaginary number,
Figure BDA0002872674510000032
FrPRαn,m(r) is a radial polynomial defined as:
Figure BDA0002872674510000033
wherein k is a summation index;
for a gray scale image f (x, y) of size N × N, its discrete fractional Zernike moments are defined as:
Figure BDA0002872674510000034
wherein, FrPRαn,m(r) is a radial polynomial, rx,y and θx,yRespectively shown as the following formula:
Figure BDA0002872674510000035
Figure BDA0002872674510000036
wherein ,
Figure BDA0002872674510000037
further, the quick calculation method for the quaternion fractional order pseudo-Zernike moment transformation is obtained by deducing the relation between the fractional order pseudo-Zernike moment and the quaternion fractional order pseudo-Zernike moment, and the specific operation formula is as follows:
Figure BDA0002872674510000041
wherein ,
Figure BDA0002872674510000042
wherein FrPZMαn,m(fξ) Xi is an element of { R, G, B }, and is a fractional order pseudo Zernike moment of three channels of red, green and blue respectively; re (z) represents the real part of the conventional complex number z, im (z) represents the imaginary part of the conventional complex number z; mu is the unit pure four-element number, mu is ai + bj + ck,
Figure BDA0002872674510000043
i. j and k are units of imaginary numbers.
Further, discretization of the fractional order pseudo Zernike moments can be rapidly achieved by radial polynomials defined in recursion (2), the recursion formula of the fractional order pseudo Zernike moments being similar to the Zernike moments and being expressed as:
Figure BDA0002872674510000044
wherein the coefficient K1,K2,K3 and K4Given by equation 9:
Figure BDA0002872674510000051
further, the construction method of the facial image feature library comprises the following steps:
collecting a color face image of each person;
dividing each color face image into I channels according to RGB channelsR、IG、IBThree images;
will IR、IG、IBPerforming fractional order pseudo-Zernike moment transformation on the three images respectively to obtain three fractional order pseudo-Zernike moment feature matrices;
and transforming the fractional order pseudo-Zernike moment feature matrix into a quaternion fractional order pseudo-Zernike moment feature matrix by adopting a quaternion fractional order fast computation method to obtain an identification feature matrix of each color face image, wherein the identification feature matrices of all the color face images form a face image feature library.
Further, the matching the recognition feature matrix of the color facial image with the recognition feature matrix in the pre-established facial image feature library by using the K-NN algorithm includes:
and taking K-2 in the K-NN algorithm, finding out 2 nearest neighbor features with Euclidean distance from the recognition feature matrix of the color face image from the face image feature library, and taking other nearest neighbor features except the face image feature library as recognition results to finish face verification.
On the other hand, the invention also provides a face verification system based on quaternion fractional order pseudo-Zernike moments, which comprises the following steps:
the face acquisition unit is configured to acquire a color face image of the identity person to be authenticated;
an image preprocessing unit configured to divide the color face image into I channels according to RGB three channelsR、IG、IBThree images;
a first transformation unit configured to transform IR、IG、IBPerforming fractional order pseudo-Zernike moment transformation on the three images respectively to obtain three fractional order pseudo-Zernike moment feature matrices;
the second transformation unit is configured to transform the fractional order pseudo-Zernike moment feature matrix into a quaternion fractional order pseudo-Zernike moment feature matrix by adopting a quaternion fractional order pseudo-Zernike moment transformation rapid calculation method so as to obtain an identification feature matrix of the color face image;
and the feature matching unit is configured to match the recognition feature matrix of the color face image with a recognition feature matrix in a pre-established face image feature library by using a K-NN algorithm to complete face verification.
On the other hand, the invention provides a face verification system based on quaternion fractional order pseudo-Zernike moment, which comprises a processor and a storage medium;
the storage medium is used for storing instructions;
the processor is used for operating according to the instruction to execute the steps of the face verification method based on quaternion fractional order pseudo-Zernike moments.
In another aspect, the present invention provides a computer-readable storage medium, on which a computer program is stored, which when executed by a processor, implements the steps of the aforementioned face verification method based on quaternion fractional order pseudo-Zernike moments.
Compared with the prior art, the invention has the following advantages and beneficial effects: the method adopts quaternion fractional order pseudo Zernike moment transformation as a feature extraction algorithm, and the extracted face features have rotating scaling robustness; the rapid calculation method can rapidly complete the face verification of the workers and the shifts of the rail transit maintenance management system, and effectively improves the verification rate of the face verification model.
Drawings
FIG. 1 is a flowchart of a face verification method based on quaternion fractional order pseudo-Zernike moments according to an embodiment of the present invention;
FIG. 2 is an algorithmic flow diagram of an embodiment of the present invention;
FIG. 3 is a comparison graph of the fast computation and the direct computation of the quaternion fractional order pseudo-Zernike moment transform of the present invention.
Detailed Description
The invention is further described with reference to specific examples. The following examples are only for illustrating the technical solutions of the present invention more clearly, and the protection scope of the present invention is not limited thereby.
As shown in fig. 1, an embodiment of the present invention provides a face verification method based on quaternion fractional order pseudo-Zernike moments, including the following steps:
step S11, collecting the color face image of the identity person to be verified;
step S12, dividing the color face image into I according to RGB three channelsR、IG、IBThree images;
step S13, adding IR、IG、IBPerforming fractional order pseudo-Zernike moment transformation on the three images respectively to obtain three fractional order pseudo-Zernike moment feature matrices;
the construction algorithm of the fractional order pseudo Zernike moment is defined as follows:
for a two-dimensional continuous function f (r, theta) in polar coordinates, the fractional order pseudo-Zernike moments are:
Figure BDA0002872674510000071
wherein α is a fractional order; n and m are nonnegative integers, which are the order and the repetitiveness of the fractional order pseudo Zernike moment respectively; j is a unit of an imaginary number,
Figure BDA0002872674510000072
FrPRαn,m(r) is a radial polynomial defined as:
Figure BDA0002872674510000081
where k is the sum index.
For a gray scale image f (x, y) of size N × N, its discrete fractional Zernike moments are defined as:
Figure BDA0002872674510000082
wherein, FrPRαn,m(r) is a radial polynomial, rx,y and θx,yRespectively shown as the following formula:
Figure BDA0002872674510000083
Figure BDA0002872674510000084
wherein ,
Figure BDA0002872674510000085
step S14, transforming the fractional order pseudo-Zernike moment feature matrix into a quaternion fractional order pseudo-Zernike moment feature matrix by adopting a quaternion fractional order pseudo-Zernike moment transformation rapid calculation method to obtain an identification feature matrix of the color face image;
the quick calculation method for the quaternion fractional order pseudo-Zernike moment transformation is obtained by deducing the relation between the fractional order pseudo-Zernike moment and the quaternion fractional order pseudo-Zernike moment, and the specific operation formula is as follows:
Figure BDA0002872674510000091
Figure BDA0002872674510000092
wherein FrPZMαn,m(fξ) Xi is an element of { R, G, B }, and is a fractional order pseudo Zernike moment of three channels of red, green and blue respectively; re (z) represents the real part of the conventional complex number z, im (z) represents the imaginary part of the conventional complex number z; μ is the unit pure four element number, and μ ═ ai + bj + ck.
Figure BDA0002872674510000093
i. j and k are units of imaginary numbers.
The relationship given in equation 7 allows us to achieve discretization of the quaternion fractional order pseudo-Zernike moments quickly with fractional order pseudo-Zernike moments. The discretization of the fractional order pseudo-Zernike moments can also be quickly realized by a recursive algorithm, and the key idea of the recursive algorithm is to recursively calculate the radial polynomials defined in equation 2. The recursive formula for the fractional order pseudo-Zernike moments is similar to the Zernike moments and is expressed as:
Figure BDA0002872674510000094
wherein the coefficient K1,K2,K3 and K4Given by equation 8:
Figure BDA0002872674510000101
and step S15, matching the recognition feature matrix of the color face image with a recognition feature matrix in a pre-established face image feature library by using a K-NN algorithm to finish face verification.
In the step, taking K-2 in the K-NN algorithm, finding out 2 nearest neighbor features with Euclidean distance from the recognition feature matrix of the color face image from the face image feature library, taking other nearest neighbor features except the face image feature library as recognition results, namely two color images with the most similar quaternion fractional order pseudo-Zernike moment features, and finally finishing face verification.
Further, the construction method of the facial image feature library comprises the following steps:
collecting a color face image of each person;
dividing each color face image into I channels according to RGB channelsR、IG、IBThree images;
will IR、IG、IBPerforming fractional order pseudo-Zernike moment transformation on the three images respectively to obtain three fractional order pseudo-Zernike moment feature matrices;
and transforming the fractional order pseudo-Zernike moment feature matrix into a quaternion fractional order pseudo-Zernike moment feature matrix by adopting a pre-derived quaternion fractional order pseudo-Zernike moment transformation rapid calculation method to obtain an identification feature matrix of each color face image, wherein the identification feature matrices of all the color face images form a face image feature library.
As shown in fig. 2, which is an algorithm flowchart of the embodiment of the present invention, taking a rail transit maintenance management system as an example, the method includes the following steps:
step 21, collecting a color face image I of each worker and class, then carrying out fast quaternion fractional order pseudo-Zernike moment transformation on the collected color face image I, and establishing a worker and class face recognition data model of a maintenance management system, wherein the specific steps are as follows:
step 211, map the color faceImage I is divided into three channels I according to RGBR、IG、IBThree images;
and step 212, performing fractional order pseudo-Zernike moment transformation on the images on the RGB three channels respectively to obtain three fractional order pseudo-Zernike moment feature matrices.
Step 213, transforming the three fractional order pseudo-Zernike moment feature matrices into a quaternion fractional order pseudo-Zernike moment feature matrix by a quaternion fractional order pseudo-Zernike moment transformation rapid calculation method, and finally obtaining the identification feature matrix of the color face image I.
Step 214, storing the identification feature matrix of each color face image I in the face image storage unit to obtain a face image feature library.
Step 22, collecting the face image I of the work class personnel to be authenticated in the authentication mode1；
Step 23, adopting a face recognition algorithm to collect the face image I1The verification is carried out by the following specific steps:
step 231, displaying the face image I1Obtaining a face image I by using fast quaternion fractional order pseudo-Zernike moment transformation1The identification feature matrix E of (a);
step 232, using K-NN algorithm to convert the face image I1The identification feature matrix E is matched with an identification feature matrix in a face image feature library of the workers and the class, K-2 in a K-NN algorithm is taken, 2 nearest neighbor features of the Euclidean distance of the feature matrix are found out, other nearest neighbor features except the characteristic feature matrix are taken as identification results, namely two color images with the most similar quaternion fractional order pseudo-Zernike moment features, and finally face verification of the workers and the class of the dimension system is completed.
The fast calculation algorithm fully considers the correlation between quaternions when calculating the fraction order quaternion pseudo-Zernike moment and uses a recursive method to carry out fast calculation, thereby solving the problem of high complexity of direct calculation of the quaternion pseudo-Zernike moment. In order to better show the proposed fast calculation effect, a comparison experiment is carried out on the proposed fractional order quaternion pseudo-Zernike moment fast calculation algorithm and a direct algorithm, and the experimental result is shown in FIG. 3. Fig. 3 shows the run time of 80 pictures with size 768 × 1024 when using two different algorithms, as is evident from the figure: the algorithm provided by the invention has higher calculation speed, and the experimental result shows that the algorithm provided by the invention has strong superiority in calculation efficiency.
The embodiment shows that the color face features extracted by using the quaternion fractional order pseudo-Zernike moment have higher robustness, the rapid calculation algorithm provided by the invention has higher calculation speed, the face verification of workers and shifts of the rail transit maintenance management system can be rapidly completed, and the verification rate of the face verification model is effectively improved.
In another embodiment, the present invention further provides a face verification system based on quaternion fractional order pseudo-Zernike moments, including:
the face acquisition unit is configured to acquire a color face image of the identity person to be authenticated;
an image preprocessing unit configured to divide the color face image into I channels according to RGB three channelsR、IG、IBThree images;
a first transformation unit configured to transform IR、IG、IBPerforming fractional order pseudo-Zernike moment transformation on the three images respectively to obtain three fractional order pseudo-Zernike moment feature matrices;
the second transformation unit is configured to transform the fractional order pseudo-Zernike moment feature matrix into a quaternion fractional order pseudo-Zernike moment feature matrix by adopting a quaternion fractional order pseudo-Zernike moment transformation rapid calculation method so as to obtain an identification feature matrix of the color face image;
and the feature matching unit is configured to match the recognition feature matrix of the color face image with a recognition feature matrix in a pre-established face image feature library by using a K-NN algorithm to complete face verification.
Further, the face verification system further includes: and the face image storage unit is configured to store the recognition feature matrix of each color face image so as to obtain a face image feature library.
In another embodiment, the invention provides a face verification system based on quaternion fractional order pseudo-Zernike moments, comprising a processor and a storage medium;
the storage medium is used for storing instructions;
the processor is used for operating according to the instruction to execute the steps of the face verification method based on quaternion fractional order pseudo-Zernike moments.
In another embodiment, the invention provides a computer-readable storage medium having stored thereon a computer program which, when executed by a processor, performs the steps of the aforementioned face verification method based on quaternion fractional order pseudo-Zernike moments.
As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
The above description is only a preferred embodiment of the present invention, and it should be noted that, for those skilled in the art, several modifications and variations can be made without departing from the technical principle of the present invention, and these modifications and variations should also be regarded as the protection scope of the present invention.

Claims (9)

1. A face verification method based on quaternion fractional order pseudo-Zernike moments is characterized by comprising the following steps:
collecting a color face image of an identity person to be verified;
dividing the color face image into three channels of RGB (red, green and blue) IR、IG、IBThree images;
will IR、IG、IBPerforming fractional order pseudo-Zernike moment transformation on the three images respectively to obtain three fractional order pseudo-Zernike moment feature matrices;
transforming the fractional order pseudo-Zernike moment feature matrix into a quaternion fractional order pseudo-Zernike moment feature matrix by adopting a quaternion fractional order pseudo-Zernike moment transformation rapid calculation method to obtain an identification feature matrix of the color face image;
and matching the recognition characteristic matrix of the color face image with a recognition characteristic matrix in a pre-established face image characteristic library by using a K-NN algorithm to finish face verification.
2. The method of claim 1, wherein the construction algorithm for the fractional order pseudo-Zernike moments is defined as follows:
for a two-dimensional continuous function f (r, theta) in polar coordinates, the fractional order pseudo-Zernike moments are:
Figure FDA0002872674500000011
wherein α is a fractional order; n and m are nonnegative integers, which are the order and the repetitiveness of the fractional order pseudo Zernike moment respectively; j is a unit of an imaginary number,
Figure FDA0002872674500000012
FrPRαn,m(r) is a radial polynomial defined as:
Figure FDA0002872674500000021
wherein k is a summation index;
for a gray scale image f (x, y) of size N × N, its discrete fractional Zernike moments are defined as:
Figure FDA0002872674500000022
wherein, FrPRαn,m(r) is a radial polynomial, rx,y and θx,yRespectively shown as the following formula:
Figure FDA0002872674500000023
Figure FDA0002872674500000024
wherein ,
Figure FDA0002872674500000026
3. the method as claimed in claim 2, wherein the fast computation method of quaternion fractional order pseudo-Zernike moment transformation is derived from the relationship between fractional order pseudo-Zernike moments and quaternion fractional order pseudo-Zernike moments, and the specific operation formula is as follows:
Figure FDA0002872674500000025
wherein ,
Figure FDA0002872674500000031
Figure FDA0002872674500000032
Figure FDA0002872674500000033
Figure FDA0002872674500000034
wherein FrPZMαn,m(fξ) Xi is an element of { R, G, B }, and is a fractional order pseudo Zernike moment of three channels of red, green and blue respectively; re (z) represents the real part of the conventional complex number z, im (z) represents the imaginary part of the conventional complex number z; mu is the unit pure four-element number, mu is ai + bj + ck,
Figure FDA0002872674500000035
i. j, k are imaginary units and Ω is an integration region.
4. The method of claim 2, wherein the discretization of the fractional order pseudo-Zernike moments can be rapidly achieved by radial polynomials defined in the recursive formula (2), which is similar to the Zernike moments and is expressed as:
Figure FDA0002872674500000036
wherein the coefficient K1,K2,K3 and K4Given by equation (9):
Figure FDA0002872674500000037
5. the method of claim 1, wherein the method for constructing the face image feature library comprises:
collecting a color face image of each person;
dividing each color face image into I channels according to RGB channelsR、IG、IBThree images;
will IR、IG、IBPerforming fractional order pseudo-Zernike moment transformation on the three images respectively to obtain three fractional order pseudo-Zernike moment feature matrices;
and transforming the fractional order pseudo-Zernike moment feature matrix into a quaternion fractional order pseudo-Zernike moment feature matrix by adopting a pre-derived quaternion fractional order pseudo-Zernike moment transformation rapid calculation method to obtain an identification feature matrix of each color face image, wherein the identification feature matrices of all the color face images form a face image feature library.
6. The method according to claim 1, wherein the matching the recognition feature matrix of the color facial image with the recognition feature matrix in the pre-established facial image feature library by using the K-NN algorithm comprises:
and taking K-2 in the K-NN algorithm, finding out 2 nearest neighbor features with Euclidean distance from the recognition feature matrix of the color face image from the face image feature library, and taking other nearest neighbor features except the face image feature library as recognition results to finish face verification.
7. A face verification system based on quaternion fractional order pseudo-Zernike moments is characterized by comprising the following components:
the face acquisition unit is configured to acquire a color face image of the identity person to be authenticated;
an image preprocessing unit configured to divide the color face image into I channels according to RGB three channelsR、IG、IBThree images;
a first transformation unit configured to transform IR、IG、IBPerforming fractional order pseudo-Zernike moment transformation on the three images respectively to obtain three fractional order pseudo-Zernike moment feature matrices;
the second transformation unit is configured to transform the fractional order pseudo-Zernike moment feature matrix into a quaternion fractional order pseudo-Zernike moment feature matrix by adopting a quaternion fractional order pseudo-Zernike moment transformation rapid calculation method so as to obtain an identification feature matrix of the color face image;
and the feature matching unit is configured to match the recognition feature matrix of the color face image with a recognition feature matrix in a pre-established face image feature library by using a K-NN algorithm to complete face verification.
8. A face verification system based on quaternion fractional order pseudo-Zernike moments is characterized by comprising a processor and a storage medium;
the storage medium is used for storing instructions;
the processor is configured to operate in accordance with the instructions to perform the steps of the method according to any one of claims 1 to 6.
9. Computer-readable storage medium, on which a computer program is stored which, when being executed by a processor, carries out the steps of the method according to any one of claims 1 to 6.
CN202011622736.4A 2020-12-30 2020-12-30 Face verification method and system based on quaternion fractional order pseudo Zernike moment Active CN112633226B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202011622736.4A CN112633226B (en) 2020-12-30 2020-12-30 Face verification method and system based on quaternion fractional order pseudo Zernike moment

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202011622736.4A CN112633226B (en) 2020-12-30 2020-12-30 Face verification method and system based on quaternion fractional order pseudo Zernike moment

Publications (2)

Publication Number Publication Date
CN112633226A true CN112633226A (en) 2021-04-09
CN112633226B CN112633226B (en) 2023-09-22

Family

ID=75287363

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202011622736.4A Active CN112633226B (en) 2020-12-30 2020-12-30 Face verification method and system based on quaternion fractional order pseudo Zernike moment

Country Status (1)

Country Link
CN (1) CN112633226B (en)

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20150023558A1 (en) * 2012-03-30 2015-01-22 Muhittin Gokmen System and method for face detection and recognition using locally evaluated zernike and similar moments
CN104484432A (en) * 2014-12-20 2015-04-01 辽宁师范大学 Color image searching method based on quaternion exponential moment
CN110390307A (en) * 2019-07-25 2019-10-29 首都师范大学 Expression recognition method, expression recognition model training method and device
CN110610144A (en) * 2019-08-28 2019-12-24 首都师范大学 Expression recognition method and system for privacy protection

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20150023558A1 (en) * 2012-03-30 2015-01-22 Muhittin Gokmen System and method for face detection and recognition using locally evaluated zernike and similar moments
CN104484432A (en) * 2014-12-20 2015-04-01 辽宁师范大学 Color image searching method based on quaternion exponential moment
CN110390307A (en) * 2019-07-25 2019-10-29 首都师范大学 Expression recognition method, expression recognition model training method and device
CN110610144A (en) * 2019-08-28 2019-12-24 首都师范大学 Expression recognition method and system for privacy protection

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
杨迪等: "基于伪Zernike矩归一化的人脸特征提取方法", 计算机工程与应用, no. 35 *

Also Published As

Publication number Publication date
CN112633226B (en) 2023-09-22

Similar Documents

Publication Publication Date Title
Yang et al. Fvras-net: An embedded finger-vein recognition and antispoofing system using a unified cnn
Yuan et al. Fingerprint liveness detection using an improved CNN with image scale equalization
Bansal et al. Umdfaces: An annotated face dataset for training deep networks
WO2021227349A1 (en) Front-end facial image encryption and recognition method for biometric privacy protection
Han et al. Multispectral palmprint recognition using wavelet-based image fusion
Yin et al. Dynamic difference learning with spatio–temporal correlation for deepfake video detection
Jagadeesan et al. Cryptographic key generation from multiple biometric modalities: Fusing minutiae with iris feature
Soleymani et al. Generalized bilinear deep convolutional neural networks for multimodal biometric identification
Shao et al. PalmGAN for cross-domain palmprint recognition
Gu et al. Latent fingerprint registration via matching densely sampled points
CN100365645C (en) Identification Method Based on Eyebrow Recognition
Sardar et al. Enhanced biometric template protection schemes for securing face recognition in IoT environment
Aslan et al. VisGIN: Visibility graph neural network on one-dimensional data for biometric authentication
El-Abed et al. Quality assessment of image-based biometric information
Balakumar et al. Secure biometric key generation scheme for cryptography using combined biometric features of fingerprint and iris
CN112633226B (en) Face verification method and system based on quaternion fractional order pseudo Zernike moment
Bharadi et al. Multi-instance iris recognition
CN112766169A (en) Account opening identity authentication method, device, equipment and storage medium
CN112163542A (en) A Palmprint Confidentiality Authentication Method Based on ElGamal Encryption
Lalithamani et al. An efficient approach for non-invertible cryptographic key generation from cancelable fingerprint biometrics
Mohite et al. Thermal and Visual Face Recognition using Eigenfaces and Transfer Learning
Khadilkar et al. Face identification based on discrete wavelet transform and neural networks
Ji et al. A cross domain multi-modal dataset for robust face anti-spoofing
CN114612989A (en) Method and device for generating face recognition data set, electronic equipment and storage medium
Sindhu et al. Fingerprint authentication based on adaptive greedy registration of minutiae pairs

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant