CN112149525A - Face recognition method based on Laplace peak clustering - Google Patents
Face recognition method based on Laplace peak clustering Download PDFInfo
- Publication number
- CN112149525A CN112149525A CN202010927909.7A CN202010927909A CN112149525A CN 112149525 A CN112149525 A CN 112149525A CN 202010927909 A CN202010927909 A CN 202010927909A CN 112149525 A CN112149525 A CN 112149525A
- Authority
- CN
- China
- Prior art keywords
- nodes
- laplace
- network
- node
- face
- 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.)
- Pending
Links
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
- G06V40/172—Classification, e.g. identification
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
Landscapes
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Human Computer Interaction (AREA)
- Multimedia (AREA)
- Image Analysis (AREA)
- Collating Specific Patterns (AREA)
Abstract
A face recognition method based on Laplace peak clustering is characterized in that according to an existing face image data set, the similarity of any two face images is calculated by using a CWSSIM algorithm to serve as the distance between the images, the face image data set is converted into a complete weighting network, and the weight of a connected edge is the distance between image nodes; calculating a degree matrix, a Laplace matrix and a Laplace energy value of a network, further calculating the Laplace centrality and the shortest distance of network nodes, obtaining a decision index of the nodes according to the product of the Laplace centrality and the shortest distance of the nodes, selecting the nodes with the decision index larger than a preset threshold value as cluster centers, attributing other nodes to the class represented by the cluster center closest to the cluster centers, and finishing the classification of the nodes and the identification of the face image. The method extracts the central image based on the Laplace centrality, and then identifies the face image through clustering, so that the speed and the accuracy of face identification are improved.
Description
Technical Field
The invention relates to the field of image recognition, in particular to a face recognition method based on Laplace peak clustering.
Background
The rapid development of machine learning promotes social informatization and networked development, wherein algorithms of a large number of machine learning and deep learning are applied to the aspects of the traditional industry, the development process of intellectualization of the traditional industry is promoted, and the field of image recognition is rapidly promoted and developed due to the deep learning, a large number of face recognition applications appear in society, such as face recognition used in community entrance guard, face recognition used in tourist attractions and face recognition used in online payment. The clustering algorithm is also very active in the field of machine learning, the clustering algorithm can extract the internal features of data and cluster data points with similar attributes into a category, the data points in the same category are close in distance and have similar features, but the data points in different categories are low in similarity. The application of clustering algorithm in the field of face recognition requires that various portrait pictures (front face, side face, etc.) of a person are correctly labeled with the user's label.
The application of face recognition can increase social operation efficiency, for example, the efficiency of tourists entering a garden can be improved by adopting face recognition in a tourist attraction; the community entrance guard adopts face recognition to ensure that residents are safer; the face recognition is adopted for online payment, so that the complicated operation of inputting the password is omitted for the user, and the payment is more convenient and safer. The face recognition product is widely applied to the fields of finance, judicial sciences, military, public security, frontier inspection, government, aerospace, electric power, factories, education, medical treatment, numerous enterprises and public institutions and the like. With further maturity of the technology and improvement of social acceptance, the face recognition technology is applied to more fields.
In order to complete the task of face recognition, various face recognition algorithms exist at present, such as a recognition algorithm based on face feature points, a recognition algorithm based on templates, a recognition algorithm based on neural networks, and the like. The results of the algorithms for recognizing the images are accurate, but the time complexity is high, and the time for training the algorithms is too long. In order to divide the face image more quickly and accurately, the invention provides a face identification method based on Laplace peak clustering.
Disclosure of Invention
In order to overcome the defect that the algorithm time complexity of a training model in the field of face recognition is too high at present, the invention provides a rapid and efficient face recognition method based on Laplace peak clustering.
The technical scheme adopted by the invention for solving the technical problems is as follows:
a face recognition method based on peak value clustering comprises the following steps:
the method comprises the following steps: according to the existing face image set with the total number of N, calculating the similarity d between any two face images i and j by using a CWSSIM algorithmij；
Step two: converting the face image set with the total number of N into a weighted undirected fully-connected network G (V, E), wherein V is a node, E is a connecting edge, each node represents a face image, the weight of the connecting edge between the nodes is the distance between corresponding face images, and the weight matrix of the network is
Step three: construction degree matrix A (G)
Wherein
Representing the sum of the connecting edge weights of the ith personal face image node and all other nodes;
step four: calculating a laplace matrix l (G) of network G (a) (G) -d (G);
step five: calculating the Laplace energy value of the network G
Step six: calculating the center of Laplacian c of any node i in the network Gi；
Wherein E isL(Gi) Representing the Laplace energy value of the network without the node i;
step seven: for any node i in the network, computing the node i and all the Laplace centralities are greater than ciThe distance between the nodes of the face image is obtained, and the shortest distance of the node i is obtained
Step eight: calculating decision index of node ii＝ci×i；
Step nine: repeating the sixth step to the eighth step, calculating decision indexes of all nodes in the network G, obtaining nodes with the decision indexes larger than a preset threshold value as a clustering center, and distributing labels;
step ten: and for other face image nodes without labels, assigning the other face image nodes to the class to which the existing labels with the nearest distance belong, finishing the classification of all face images and obtaining the face recognition result.
The technical conception of the invention is as follows: calculating the distance between the images by using a CWSSIM algorithm, converting the images into data points in a network, finding a clustering center in the image network based on peak clustering, and allocating labels to image nodes by using the Laplace centrality to identify the labels of the images.
The invention has the beneficial effects that: the method has the advantages that the high efficiency of image classification by Laplace peak clustering is utilized, the accuracy of face recognition is further improved, and the recognition accuracy and efficiency are greatly improved.
Detailed Description
The present invention is further described below.
A face recognition method based on Laplace peak clustering comprises the following steps:
the method comprises the following steps: calculating the similarity d between any two face images i and j by using a CWSSIM algorithm according to the existing face image set with the total number of N being 30ij；
Step two: converting a face image set with the total number of N-30 into a weighted undirected fully-connected network G (V, E), wherein V is a node, E is a connecting edge, each node represents a face image, the weight of the connecting edge between the nodes is the distance between corresponding face images, and the weight matrix of the network is
Step three: construction degree matrix A (G)
Wherein
Representing the sum of the connecting edge weights of the ith personal face image node and all other nodes;
step four: calculating a laplace matrix l (G) of network G (a) (G) -d (G);
step five: calculating the Laplace energy value of the network G
Step six: calculating the center of Laplacian c of any node i in the network Gi；
Wherein E isL(Gi) Representing the Laplace energy value of the network without the node i;
step seven: for any node i in the network, computing the node i and all the Laplace centralities are greater than ciDistance between the nodes of the face imageDistance, and thus the shortest distance of node i
Step eight: calculating decision index of node ii＝ci×i；
Step nine: repeating the sixth step to the eighth step, calculating decision indexes of all nodes in the network G, obtaining nodes with the decision indexes larger than a preset threshold value as a clustering center, and distributing labels;
step ten: and for other face image nodes without labels, assigning the other face image nodes to the class to which the existing labels with the nearest distance belong, finishing the classification of all face images and obtaining the face recognition result.
As mentioned above, the present invention is made more clear by the specific implementation steps implemented in this patent. Any modification and variation of the present invention within the spirit of the present invention and the scope of the claims will fall within the scope of the present invention.
Claims (1)
1. A face recognition method based on Laplace peak clustering is characterized in that: the method comprises the following steps:
the method comprises the following steps: according to the existing face image set with the total number of N, calculating the similarity d between any two face images i and j by using a CWSSIM algorithmij；
Step two: converting the face image set with the total number of N into a weighted undirected fully-connected network G (V, E), wherein V is a node, E is a connecting edge, each node represents a face image, the weight of the connecting edge between the nodes is the distance between corresponding face images, and the weight matrix of the network is
Step three: construction degree matrix A (G)
Wherein
Representing the sum of the connecting edge weights of the ith personal face image node and all other nodes;
step four: calculating a laplace matrix l (G) of network G (a) (G) -d (G);
step five: calculating the Laplace energy value of the network G
Step six: calculating the center of Laplacian c of any node i in the network Gi；
Wherein E isL(Gi) Representing the Laplace energy value of the network without the node i;
step seven: for any node i in the network, computing the node i and all the Laplace centralities are greater than ciThe distance between the nodes of the face image is obtained, and the shortest distance of the node i is obtained
Step eight: calculating decision index of node ii＝ci×i；
Step nine: repeating the sixth step to the eighth step, calculating decision indexes of all nodes in the network G, obtaining nodes with the decision indexes larger than a preset threshold value as a clustering center, and distributing labels;
step ten: and for other face image nodes without labels, assigning the other face image nodes to the class to which the existing labels with the nearest distance belong, finishing the classification of all face images and obtaining the face recognition result.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010927909.7A CN112149525A (en) | 2020-09-07 | 2020-09-07 | Face recognition method based on Laplace peak clustering |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010927909.7A CN112149525A (en) | 2020-09-07 | 2020-09-07 | Face recognition method based on Laplace peak clustering |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112149525A true CN112149525A (en) | 2020-12-29 |
Family
ID=73889379
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010927909.7A Pending CN112149525A (en) | 2020-09-07 | 2020-09-07 | Face recognition method based on Laplace peak clustering |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112149525A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113426129A (en) * | 2021-06-24 | 2021-09-24 | 网易（杭州）网络有限公司 | User-defined role appearance adjusting method, device, terminal and storage medium |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105631416A (en) * | 2015-12-24 | 2016-06-01 | 华侨大学 | Method for carrying out face recognition by using novel density clustering |
| CN105740842A (en) * | 2016-03-01 | 2016-07-06 | 浙江工业大学 | Unsupervised face recognition method based on fast density clustering algorithm |
| CN107194415A (en) * | 2017-04-28 | 2017-09-22 | 浙江工业大学 | Peak clustering method based on Laplace centrality |
| CN109241201A (en) * | 2018-08-30 | 2019-01-18 | 浙江工业大学 | A kind of Laplce's centrality peak-data clustering method based on curvature |
| CN109255378A (en) * | 2018-08-30 | 2019-01-22 | 浙江工业大学 | A kind of Laplce's centrality peak-data clustering method based on potential energy entropy |
| CN109948534A (en) * | 2019-03-19 | 2019-06-28 | 华侨大学 | The method for carrying out recognition of face is clustered using fast density peak value |
| CN111160077A (en) * | 2018-11-08 | 2020-05-15 | 北京航天长峰科技工业集团有限公司 | Large-scale dynamic face clustering method |
-
2020
- 2020-09-07 CN CN202010927909.7A patent/CN112149525A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105631416A (en) * | 2015-12-24 | 2016-06-01 | 华侨大学 | Method for carrying out face recognition by using novel density clustering |
| CN105740842A (en) * | 2016-03-01 | 2016-07-06 | 浙江工业大学 | Unsupervised face recognition method based on fast density clustering algorithm |
| CN107194415A (en) * | 2017-04-28 | 2017-09-22 | 浙江工业大学 | Peak clustering method based on Laplace centrality |
| CN109241201A (en) * | 2018-08-30 | 2019-01-18 | 浙江工业大学 | A kind of Laplce's centrality peak-data clustering method based on curvature |
| CN109255378A (en) * | 2018-08-30 | 2019-01-22 | 浙江工业大学 | A kind of Laplce's centrality peak-data clustering method based on potential energy entropy |
| CN111160077A (en) * | 2018-11-08 | 2020-05-15 | 北京航天长峰科技工业集团有限公司 | Large-scale dynamic face clustering method |
| CN109948534A (en) * | 2019-03-19 | 2019-06-28 | 华侨大学 | The method for carrying out recognition of face is clustered using fast density peak value |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113426129A (en) * | 2021-06-24 | 2021-09-24 | 网易（杭州）网络有限公司 | User-defined role appearance adjusting method, device, terminal and storage medium |
| CN113426129B (en) * | 2021-06-24 | 2024-03-01 | 网易（杭州）网络有限公司 | Method, device, terminal and storage medium for adjusting appearance of custom roles |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112308158B (en) | Multi-source field self-adaptive model and method based on partial feature alignment | |
| Dong et al. | Automatic age estimation based on deep learning algorithm | |
| Santa Cruz et al. | Deeppermnet: Visual permutation learning | |
| CN110413924B (en) | Webpage classification method for semi-supervised multi-view learning | |
| Niu et al. | Federated learning for face recognition with gradient correction | |
| CN108520275A (en) | A kind of regular system of link information based on adjacency matrix, figure Feature Extraction System, figure categorizing system and method | |
| CN110309331A (en) | A kind of cross-module state depth Hash search method based on self-supervisory | |
| CN107766850A (en) | Based on the face identification method for combining face character information | |
| CN109344759A (en) | A kind of relatives' recognition methods based on angle loss neural network | |
| CN100585617C (en) | Based on sorter integrated face identification system and method thereof | |
| CN104036255A (en) | Facial expression recognition method | |
| CN105046272B (en) | A kind of image classification method based on succinct non-supervisory formula convolutional network | |
| CN111414461A (en) | Intelligent question-answering method and system fusing knowledge base and user modeling | |
| CN109993100A (en) | The implementation method of facial expression recognition based on further feature cluster | |
| CN111401156B (en) | Image identification method based on Gabor convolution neural network | |
| CN104966075B (en) | A kind of face identification method and system differentiating feature based on two dimension | |
| CN113177612B (en) | Agricultural pest image identification method based on CNN few samples | |
| CN107220598A (en) | Iris Texture Classification based on deep learning feature and Fisher Vector encoding models | |
| CN109165698A (en) | A kind of image classification recognition methods and its storage medium towards wisdom traffic | |
| CN107832753B (en) | Face feature extraction method based on four-value weight and multiple classification | |
| Singh et al. | Leaf identification using feature extraction and neural network | |
| CN110414431B (en) | Face recognition method and system based on elastic context relation loss function | |
| CN114282059A (en) | Video retrieval method, device, equipment and storage medium | |
| CN112036511A (en) | Image retrieval method based on attention machine mapping convolutional neural network | |
| CN105868743B (en) | It is a kind of based on the face retrieval method for quickly supervising discrete Hash |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20201229 |
|
| RJ01 | Rejection of invention patent application after publication |