CN109934114B - Finger vein template generation and updating algorithm and system - Google Patents
Finger vein template generation and updating algorithm and system Download PDFInfo
- Publication number
- CN109934114B CN109934114B CN201910118830.7A CN201910118830A CN109934114B CN 109934114 B CN109934114 B CN 109934114B CN 201910118830 A CN201910118830 A CN 201910118830A CN 109934114 B CN109934114 B CN 109934114B
- Authority
- CN
- China
- Prior art keywords
- template
- image
- similarity
- finger vein
- class
- 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.)
- Active
Links
Images
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Abstract
The invention relates to the technical field of biological feature recognition, and discloses a finger vein template selection and updating algorithm and a system; firstly, acquiring finger vein images, secondly, carrying out image segmentation and enhancement on the finger vein images based on vein line segmentation based on direction valley detection, binarizing and refining the finger vein images, and then respectively comparing three template selection methods based on a least square method, an intra-class weighted least square method and an inter-class weighted least square method to select an optimal template generation method; and finally, establishing a template database based on the optimal template generation method and updating the template database.
Description
Technical Field
The invention relates to the technical field of biological feature recognition, in particular to a finger vein template generation and update algorithm and a system.
Background
With the rapid development of internet technology, information security issues become more and more important, and how to effectively identify identities to protect personal and property security is an urgent issue to be resolved. Compared to traditional authentication means such as keys and passwords, biometric features based on physiology and behavior are difficult to steal, duplicate and lose. Therefore, the biometric authentication technique has been widely studied and successfully applied to personal authentication. At present, the biological characteristics for identity authentication are mainly divided into two types:
(1) External features: face, fingerprint, iris, etc.
(2) Internal features: finger veins, palm veins, and back veins. The inherent biological features are located under the epidermis, making them difficult to steal and forge compared to external features, and therefore they have a higher security profile.
Although the finger vein biometric technology has achieved great success, there are also drawbacks, such as when the same finger is sampled multiple times at the same time or at different times, there is often a great difference between the acquired images due to the variation of conditions such as finger offset, rotation, bending and illumination, and thus the recognition performance of the system is reduced. To minimize the differences within classes, much work is currently done to capture multiple images as enrollment templates to achieve authentication. However, this solution requires more memory space and increases authentication time. To address this problem, other efforts have proposed generating super enrollment templates, which are then updated during authentication to reduce intra-class variability present in the captured images. The scheme achieves good identification performance. However, the current template generation and updating methods are mainly proposed for biometric identification systems such as fingerprints, faces and irises, and thus they cannot be effectively used in finger vein identification systems. There is currently no relevant template generation and update scheme for finger vein recognition systems. Accordingly, in order to solve the problem, the present invention proposes a finger vein enrollment template generation method to generate a robust enrollment template. Then, in the authentication stage, we propose a template updating method to register the template, realize online updating to the static template of the finger, so as to save multiple changes existing in the acquisition process of the same finger as much as possible, reduce the intra-class difference, and improve the authentication performance of the system.
Disclosure of Invention
Technical problem to be solved
Aiming at the defects of the prior art, the invention provides a finger vein template generation and updating algorithm and a system, and an optimal template is generated by a weighted least square method, so that the storage space is saved and the time is saved. Meanwhile, the template library of the newly acquired finger vein recognition system is combined to realize real-time online updating.
Technical proposal
In order to achieve the above purpose, the invention is realized by the following technical scheme:
a finger vein template generation method, comprising the following operation steps:
and (3) image acquisition: obtaining a corresponding finger vein image through an infrared camera; a registered image acquired a plurality of times for each individual;
image segmentation: because the binary vein template is generated and updated by the method, the acquired image needs to be subjected to feature extraction to obtain a binary image. The finger vein features are extracted herein using current deep learning methods. Firstly, dividing an image into a background area, a fuzzy area and a target area by using a threshold method, and establishing a training set by using background pixel points and vein pixel points. Then, a convolutional neural network model is built and trained using the training data. Finally, the vein features are enhanced by utilizing the trained model, and the enhanced image is binarized;
generating a template: in the process of generating a finger vein template, we propose a template generating method based on a weighted least square method. To reduce intra-class variations, we fuse multiple finger templates of the same class into one optimal template, the main purpose of template generation is to reduce intra-class variations, so we generate one optimal template by minimizing intra-class distances. First, a definition of an optimal template for the finger vein is given. The template generation problem is then converted into an optimized problem based on the definition. Secondly, to improve the performance of the algorithm, matching scores within and between classes are calculated to obtain weights and injected into the objective function. Finally, by solving the optimization problem, a robust registration template is obtained.
Updating the template: in the authentication process, the collected finger vein features are changed due to the influence of various factors, so that the images acquired by the same finger are difficult to effectively match and authenticate with the registration template. In order to improve the matching precision, we propose a template updating method to realize online updating of the template. First, in the authentication phase, an image to be authenticated is matched with an original registration image, and an authentication image with a matching score greater than a given threshold is stored to update the original registration template. Then, the inter-class distance and the intra-class distance are calculated through the authentication image and the original registration template, and the two distances are fused to obtain the weight of each image. Similar to the template generation, the template updating is converted into an optimization problem through the definition of the template quality, a new registration template is obtained through the least square method, and the original registration template is updated through the template.
A. Validating template quality definitions
The purpose of the existing biometric template generation is to improve the performance of the authentication system, mainly to reduce the verification error rate. Thus, biometric template generation should focus on the reduction of authentication error rate, rather than on the subjective perception of enrollment templates by people, for example, who typically think that images with high contrast can be used as enrollment templates. In the existing recognition system, when a user performs recognition, the system acquires biometric data of a user again, performs image preprocessing on the biometric data, and matches the extracted features with data stored in a dataset. The accuracy of the verification is closely related to the stability of the biometric data of people at different times. In other words, the verification error rate is mainly caused by intra-class variations. Therefore, we consider a high quality finger vein enrollment template to be an image that has the smallest intra-class distance to its cognate images.
B. Weight calculation
The template is formed by fusing a plurality of registered samples. In practice there is a certain difference between the registered samples, so we assume that different samples take different weights. Intuitively, for samples with smaller similarities and larger similarities between classes, a larger specific gravity should be occupied in template generation. Thus, we decide the weight it occupies in the template generation by computing the similarity between this sample and the other registered samples.
1) Similarity calculation: to obtain a robust weight, first, the similarity between two enrollment templates is defined, as in equation (2). Then, each sample is automatically assigned a weight based on the similarity between the classes and the similarity between the classes. Let E and F be binarized feature images extracted from two registered samples, whose sizes are x and y, respectively. The width and height of E are extended to 2w+x and 2h+y, respectively, the extended images of which are shown below.
The similarity of E and F can be calculated by equation (2):
wherein the method comprises the steps of
2) Similarity calculation: for each finger vein image sample, we calculate the distance of this finger to the other finger vein samples in the same category according to equation (2) and calculate its average as the weight of the sample. Assume that there are N categories, each with M registration samples. X is X m,n Refers to the mth sample in the nth category, weight X m,n The calculations are as follows:
d(X i,n ,X m,n ) Finger sample X i,n And sample X m,n Similarity between them. w (w) m,n In fact X m,n And the average similarity to the other M-1 registered samples in the same class.
3) And (5) calculating the similarity between classes: similarly, we calculate inter-class distance acquisition weights for each registered sample. Let w' m,n The weight of the mth sample, which is the nth class, is calculated by the following formula:
d(X m,j ,X m,n ) Finger sample X m,j And sample X m,n Similarity between them. Equation (6) calculates sample X m,n And the average similarity of all samples of the other N-1 categories.
4) Similarity fusion: different samples have different intra-and inter-class similarities, and we fuse their weights by the following formula:
the specific gravity of each of the similarity between the classes and the similarity between the classes in the template generation is determined.
C. Template generation
For the nth category we are mainly to generate a template T n The distance between this template and its registered samples is minimized, so the template generation problem is translated to solve the optimization problem as follows.
W in the formula m,n Refers to the weight of the mth sample of the nth class. Equation (8) can be solved by the least squares method. Thus, template T in the nth category n Can be obtained by the formula (9):
from equations (3) and (9), we can get samples of the smallest intra-class distance and the largest inter-class distance in template generation.
Finger vein template update
During the verification process, the finger vein quality may change due to the changing conditions under which the image is generated, resulting in a large intra-class variation. Because of the limited memory and computing power of the finger vein device, it is difficult to obtain multiple finger vein images from the same finger as a registration template. To solve this problem, we propose an online template updating method.
A weight calculation
In the test stage, feature extraction is performed on each input image, and similarity between the feature map and the template is calculated for verification. If its similarity is greater than a predefined certain threshold, the input image (unlabeled image) is stored in the data to update the template. Assume that N templates corresponding to N classes are stored in the recognition system, and K input images have been stored for each class. Since we update the templates with N templates and K input images, we calculate their weights separately.
X k,n Refers to the kth input image in the nth category, k=1, 2. T for original template o,n N=1, 2. Weight of original template is W 0,n To represent.
Wherein w is 0,n Representing its intra-class uniformity to the k input samples. w (w) 0,n From the formula(11) Calculation of
w' 0,n Calculated from equation (12)
Equation (12) actually calculates the template T 0,n Average inter-class similarity with other N-1 templates.
After the weights of the templates are calculated, we calculate the weights of the input images. The kth input sample x k,n The weights of (2) are calculated by equation (13)
In formula (13), w k,n Is the average similarity of the kth input image with the other K-1 input images and its corresponding templates
w' k,n Refers to the average similarity of the kth input image in the nth sample to the other n-1 templates.
B template update
As in equation (8), the modified template T in the nth category n (k) By solving the following objective function
Finally, an improved template is generated
Note that T in formula (9) n * And in equation (17)Is two probability maps. To complete the verification, we binarize them and then store the resulting binarized images for verification of the enrollment template.
Advantageous effects
The invention provides a finger vein template selection and updating algorithm and a system for the first time, and compared with the prior art, the finger vein template selection and updating algorithm has the following beneficial effects:
1. the invention firstly provides the generation and updating of the registration template in the finger vein recognition system.
2. The invention provides a weighted least square method utilizing the inter-class distance and the intra-class distance, and the template generation of the method not only can obtain the optimal template, but also can save a large amount of memory and storage space and save finger vein matching time.
3. The invention provides an online finger vein authentication system template updating algorithm, which can update a registration template in real time, reduce the inter-class variability of finger vein images and improve the recognition accuracy of the system.
4. The weighted least square method can be applied to finger vein recognition systems, can be used for building other biological characteristic templates, and is good in expansibility.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, it being obvious that the drawings in the following description are only some embodiments of the invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
Fig. 1 is a flow chart of finger vein template generation in accordance with the present invention.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention, and it is apparent that the described embodiments are some embodiments of the present invention, but not all embodiments of the present invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Examples:
the finger vein template generation method of the embodiment comprises the following operation steps:
and (3) image acquisition: obtaining a corresponding finger vein image through an infrared camera; a registered image acquired a plurality of times for each individual;
image segmentation: because the binary vein template is generated and updated by the method, the acquired image needs to be subjected to feature extraction to obtain a binary image. The finger vein features are extracted herein using current deep learning methods. Firstly, dividing an image into a background area, a fuzzy area and a target area by using a threshold method, and establishing a training set by using background pixel points and vein pixel points. Then, a convolutional neural network model is built and trained using the training data. Finally, the vein features are enhanced by utilizing the trained model, and the enhanced image is binarized;
generating a template: in the process of generating a finger vein template, we propose a template generating method based on a weighted least square method. To reduce intra-class variations, we fuse multiple finger templates of the same class into one optimal template, the main purpose of template generation is to reduce intra-class variations, so we generate one optimal template by minimizing intra-class distances. First, a definition of an optimal template for the finger vein is given. The template generation problem is then converted into an optimized problem based on the definition. Secondly, to improve the performance of the algorithm, matching scores within and between classes are calculated to obtain weights and injected into the objective function. Finally, by solving the optimization problem, a robust registration template is obtained.
Updating the template: in the authentication process, the collected finger vein features are changed due to the influence of various factors, so that the images acquired by the same finger are difficult to effectively match and authenticate with the registration template. In order to improve the matching precision, we propose a template updating method to realize online updating of the template. First, in the authentication phase, an image to be authenticated is matched with an original registration image, and an authentication image with a matching score greater than a given threshold is stored to update the original registration template. Then, the inter-class distance and the intra-class distance are calculated through the authentication image and the original registration template, and the two distances are fused to obtain the weight of each image. Similar to the template generation, the template updating is converted into an optimization problem through the definition of the template quality, a new registration template is obtained through the least square method, and the original registration template is updated through the template.
D. Validating template quality definitions
The main purpose of the existing template generation method is to improve performance by reducing the verification error rate. Thus, biometric template generation is primarily defined by minimizing distance, rather than based on subjective perception of enrollment templates by humans. In the existing recognition system, when a user performs recognition, the system acquires biometric data of a user again, performs image preprocessing on the biometric data, and matches the extracted features with data stored in a dataset. The accuracy of verification is closely related to the biometric data of people at different times. In other words, the verification accuracy is mainly caused by intra-class variation, so we assume that one high quality finger vein has the smallest intra-class distance among all enrollment templates.
E. Weight calculation
The template is formed by fusing a plurality of registered samples. In practice there is a certain difference between the registered samples, so we assume that different samples take different weights. Intuitively, samples with greater similarity and less similarity between classes should have a greater specific gravity in template generation. Thus, we decide the weight it occupies in the template generation by calculating the distance between this sample and the other registered samples.
5) Similarity calculation: to obtain a robust weight, first, the similarity between two enrollment templates is defined, as in equation (2). Then, each sample is automatically assigned a weight based on the similarity between the classes and the similarity between the classes. Let E and F be binarized feature maps extracted from two registered samples, the sizes of which are x, y, respectively. The width and height of E are extended to 2w+x and 2h+y, respectively, the extended images of which are shown below.
The similarity of E and F can be calculated by equation (2):
wherein the method comprises the steps of
6) Similarity calculation: for each finger vein image we calculate the distance of this finger to the other fingers in the same category according to equation (2), calculate its average and take it as a weight. Assume now that there are N categories, each with M registration samples. Xm, n refers to the m-th sample in the n-th class, and the weights Xm, n are calculated as follows:
d (Xi, n, xm, n) refers to the similarity between samples Xi, n and samples Xm, n. Wm, n is effectively the average similarity of Xm, n to other M-1 registration samples in the same class.
7) And (5) calculating the similarity between classes: similarly, we calculate inter-class distance acquisition weights for each registered sample. Weight W' m,n Refers to the mth sample of the nth class, which is calculated by the following formula:
d (Xi, n, xm, n) refers to the similarity between samples Xi, n and samples Xm, n. Equation (6) calculates the average similarity of samples Xm, N and all samples of the other N-1 categories.
8) Similarity fusion: different samples have different intra-and inter-class similarities, and we fuse their weights by the following formula:
the specific gravity of each of the similarity between the classes and the similarity between the classes in the template generation is determined.
F. Template generation
For the nth class we mainly aim to generate a template Tn such that there is minimal intra-class variation from this template to its registered samples, so the template generation problem translates into a solution to the optimization problem.
Wm, n in the formula refers to the weight of the mth sample of the nth class. Equation (8) can be solved by the least squares method. Thus, the templates Tn in the nth category can be derived by equation (9):
from equations (3) and (9), we can get samples of the smallest intra-class distance and the largest inter-class distance in template generation.
Finger vein template update
During the verification process, the finger vein quality may change due to the changing conditions under which the image is generated, resulting in a large intra-class variation. Because of the limited memory and computing power of the finger vein device, it is difficult to obtain multiple finger vein images from the same finger. To solve this problem, we propose a method of continuously updating templates through online operation.
A weight calculation
In the test stage, feature processing is performed on each input image, and similarity between the feature map and the template is calculated for verification. If its similarity is greater than a predefined certain threshold, this input image (unlabeled image) will be stored to update the template during the verification process. Assume that there are N templates in N categories, and each category stores K input images. Since the updating of the templates is based on N categories and K input images, we calculate their weights separately.
Xk, N refers to the kth input image in the nth category, k=1, 2. T for original template o,n N=1, 2. Weight of original template is W 0,n To represent.
Wherein w is 0,n Refers to intra-class homogeneity of the K input samples. w (w) 0,n Calculated from formula (11)
w' 0,n Calculated from equation (12)
Equation (12) calculates the template T 0,n Average inter-class similarity with other N-1 templates.
The Kth input sample x k,n The weights of (2) are calculated by equation (13)
In formula (13), w k,n Is the average similarity of the kth input to the other K-1 inputs
w' k,n Refers to the average similarity of the kth input in the nth sample to the other n-1 templates.
B template improvement
As in equation (8), the modified template T in the nth class n (k) By calculating the following objective function
Finally, an improved template is generated
Note that T in formula (9) n * And in equation (17)Is two probability maps. To complete the verification, they should be binarized and then the stored binarized images used for verification of the enrollment template.
It is noted that relational terms such as first and second, and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.
The above embodiments are only for illustrating the technical solution of the present invention, and are not limiting; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims (2)
1. The finger vein template generation method is characterized by comprising the following operation steps:
and (3) image acquisition: obtaining a corresponding finger vein image through an infrared camera; a registered image acquired a plurality of times for each individual;
image segmentation: because the binary vein template is generated and updated, the acquired image needs to be subjected to feature extraction to obtain a binary image; extracting finger vein features by using a current deep learning method; firstly, dividing an image into a background area, a fuzzy area and a target area by using a threshold method, and establishing a training set by using background pixel points and vein pixel points; then, a convolutional neural network model is established, and training data is utilized to train the convolutional neural network model; finally, the vein features are enhanced by utilizing the trained model, and the enhanced image is binarized;
generating a template: in the process of generating a finger vein template, a template generating method based on a weighted least square method is provided; in order to reduce the intra-class variation, a plurality of finger templates of the same class are fused into an optimal template, and the main purpose of template generation is to reduce the intra-class variation, so that an optimal template is generated by minimizing the intra-class distance; firstly, defining an optimal template of a finger vein; then, converting the template generation problem into an optimized problem based on the definition; secondly, in order to improve the performance of the algorithm, calculating matching scores in and among classes to obtain weights, and injecting the weights into an objective function; finally, obtaining a robust registration template by solving the optimization problem;
updating the template: in the authentication process, the collected finger vein features are changed due to the influence of various factors, so that the images acquired by the same finger are difficult to realize effective matching and authentication with the registration template; in order to improve the matching precision, a template updating method is provided, and online updating of the template is realized; firstly, in an authentication stage, matching an image to be authenticated with an original registration image, and storing the authentication image with the matching score larger than a given threshold value to update the original registration template; then, calculating the inter-class distance and the intra-class distance through the authentication image and the original registration template, and fusing the two distances to obtain the weight of each image; similar to the template, the template is updated and converted into an optimization problem through the definition of the template quality, a new registration template is obtained through the least square method, and the original registration template is updated by using the template;
similarity is calculated as:
d() Finger sample->And sample->Similarity between; />In fact +.>Average similarity to other M-1 registration samples in the same class;
the similarity between classes is calculated as:
d() Finger sample->And sample->Similarity between; equation (2) calculates the sample +.>Average similarity to all samples of the other N-1 categories;
similarity fusion is as follows:
the finger vein template is generated as follows:
2. The finger vein template generation method according to claim 1, wherein the finger vein template is updated as:
refers to the kth input image in the nth category, k=1, 2,; original template->N=1, 2,..n; weight of original template +.>To represent;
wherein the method comprises the steps ofMean intra-class homogeneity of K input samples; />Calculated from equation (6)
In the formula (9) of the present invention,is the average similarity of the kth input to the other K-1 inputs
template update
Finally, an improved template is generated
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910118830.7A CN109934114B (en) | 2019-02-15 | 2019-02-15 | Finger vein template generation and updating algorithm and system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910118830.7A CN109934114B (en) | 2019-02-15 | 2019-02-15 | Finger vein template generation and updating algorithm and system |
Publications (2)
Publication Number | Publication Date |
---|---|
CN109934114A CN109934114A (en) | 2019-06-25 |
CN109934114B true CN109934114B (en) | 2023-05-12 |
Family
ID=66985615
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910118830.7A Active CN109934114B (en) | 2019-02-15 | 2019-02-15 | Finger vein template generation and updating algorithm and system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109934114B (en) |
Families Citing this family (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110909757B (en) * | 2019-08-20 | 2023-07-14 | 北京北科慧识科技股份有限公司 | Method for selecting and updating template of biological recognition system |
CN112863523B (en) * | 2019-11-27 | 2023-05-16 | 华为技术有限公司 | Voice anti-counterfeiting method and device, terminal equipment and storage medium |
CN111144323A (en) * | 2019-12-28 | 2020-05-12 | 广东智冠信息技术股份有限公司 | Palm vein biological feature recognition registry self-adaptive change method |
CN112102210B (en) * | 2020-11-17 | 2021-03-16 | 北京圣点云信息技术有限公司 | Vein image template updating method and device based on self-learning |
CN112990163B (en) * | 2021-05-18 | 2021-08-06 | 深圳阜时科技有限公司 | Fingerprint calibration method, electronic device and storage medium |
CN113609943A (en) * | 2021-07-27 | 2021-11-05 | 东风汽车有限公司东风日产乘用车公司 | Finger vein recognition method, electronic device and storage medium |
CN113792649A (en) * | 2021-09-13 | 2021-12-14 | 广州广电运通金融电子股份有限公司 | Rapid authentication method, device and medium based on finger vein biological identification technology |
Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2007172175A (en) * | 2005-12-20 | 2007-07-05 | Nippon Telegr & Teleph Corp <Ntt> | Method, device and program for image retrieval, and computer-readable storage medium |
CN103778409A (en) * | 2014-01-02 | 2014-05-07 | 深圳市元轩科技发展有限公司 | Human face identification method based on human face characteristic data mining and device |
JP2015088040A (en) * | 2013-10-31 | 2015-05-07 | マネージメントサービス株式会社 | Authentication device, authentication method, and program |
US9418214B1 (en) * | 2011-12-06 | 2016-08-16 | Imageware Systems, Inc. | Anonymous biometric enrollment |
CN105917353A (en) * | 2013-09-16 | 2016-08-31 | 眼验股份有限公司 | Feature extraction and matching and template update for biometric authentication |
CN105956579A (en) * | 2016-05-27 | 2016-09-21 | 国创科视科技股份有限公司 | Rapid finger vein identification method integrating fuzzy template and point characteristics |
US9471765B1 (en) * | 2015-07-01 | 2016-10-18 | Fingerprint Cards Ab | Fingerprint authentication with template updating |
CN106326886A (en) * | 2016-11-07 | 2017-01-11 | 重庆工商大学 | Finger-vein image quality evaluation method and system based on convolutional neural network |
CN108429614A (en) * | 2018-01-05 | 2018-08-21 | 杭州电子科技大学 | A kind of fuzzy vault implementation method merged with face characteristic grade based on fingerprint |
Family Cites Families (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR100825756B1 (en) * | 2006-12-05 | 2008-04-29 | 한국전자통신연구원 | Method for feature extraction and its apparatus |
US8239685B2 (en) * | 2007-10-24 | 2012-08-07 | Electronics And Telecommunications Research Institute | Biometric authentication method |
GB201217721D0 (en) * | 2012-10-03 | 2012-11-14 | Holition Ltd | Video image processing |
US10043058B2 (en) * | 2016-03-09 | 2018-08-07 | International Business Machines Corporation | Face detection, representation, and recognition |
-
2019
- 2019-02-15 CN CN201910118830.7A patent/CN109934114B/en active Active
Patent Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2007172175A (en) * | 2005-12-20 | 2007-07-05 | Nippon Telegr & Teleph Corp <Ntt> | Method, device and program for image retrieval, and computer-readable storage medium |
US9418214B1 (en) * | 2011-12-06 | 2016-08-16 | Imageware Systems, Inc. | Anonymous biometric enrollment |
CN105917353A (en) * | 2013-09-16 | 2016-08-31 | 眼验股份有限公司 | Feature extraction and matching and template update for biometric authentication |
JP2015088040A (en) * | 2013-10-31 | 2015-05-07 | マネージメントサービス株式会社 | Authentication device, authentication method, and program |
CN103778409A (en) * | 2014-01-02 | 2014-05-07 | 深圳市元轩科技发展有限公司 | Human face identification method based on human face characteristic data mining and device |
US9471765B1 (en) * | 2015-07-01 | 2016-10-18 | Fingerprint Cards Ab | Fingerprint authentication with template updating |
CN105956579A (en) * | 2016-05-27 | 2016-09-21 | 国创科视科技股份有限公司 | Rapid finger vein identification method integrating fuzzy template and point characteristics |
CN106326886A (en) * | 2016-11-07 | 2017-01-11 | 重庆工商大学 | Finger-vein image quality evaluation method and system based on convolutional neural network |
CN108429614A (en) * | 2018-01-05 | 2018-08-21 | 杭州电子科技大学 | A kind of fuzzy vault implementation method merged with face characteristic grade based on fingerprint |
Non-Patent Citations (2)
Title |
---|
一种基于模板匹配的手指静脉识别算法;邹晖等;《传感技术学报》;20161031;第29卷(第10期);第1530-1534页 * |
基于局部细节点三维映射的指纹模板生成方法;惠妍等;《中国科学: 信息科学》;20190131;第49卷(第1期);第43-56页 * |
Also Published As
Publication number | Publication date |
---|---|
CN109934114A (en) | 2019-06-25 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109934114B (en) | Finger vein template generation and updating algorithm and system | |
Chlaoua et al. | Deep learning for finger-knuckle-print identification system based on PCANet and SVM classifier | |
Ali et al. | Fingerprint recognition for person identification and verification based on minutiae matching | |
You et al. | On hierarchical palmprint coding with multiple features for personal identification in large databases | |
Gu et al. | Efficient rectification of distorted fingerprints | |
Haji et al. | Real time face recognition system (RTFRS) | |
Jalilian et al. | Enhanced segmentation-CNN based finger-vein recognition by joint training with automatically generated and manual labels | |
Gavisiddappa et al. | Multimodal biometric authentication system using modified ReliefF feature selection and multi support vector machine | |
Bakhshi et al. | End to end fingerprint verification based on convolutional neural network | |
Nguyen et al. | Automatic identification fingerprint based on machine learning method | |
Ebrahimian et al. | Automated person identification from hand images using hierarchical vision transformer network | |
Joshi et al. | Cross-domain consistent fingerprint denoising | |
Daramola et al. | Algorithm for fingerprint verification system | |
Aly et al. | An adaptive multimodal biometrics system using PSO | |
Dey et al. | Unimodal and Multimodal Biometric Data Indexing | |
Kaur et al. | Offline signature verification in Punjabi based on SURF features and critical point matching using HMM | |
Turky et al. | The use of SOM for fingerprint classification | |
Virdaus et al. | Fingerprint verification with crossing number extraction and orientation-based matching | |
Sayeed et al. | An efficient multimodal biometric authentication integrating fingerprint and face features | |
Divakar | Multimodal biometric system using index based algorithm for fast search | |
Khodadoust et al. | A novel indexing algorithm for latent palmprints leveraging minutiae and orientation field | |
Lim et al. | A central point-based analysis for fingerprint liveness detection | |
Singh et al. | Fingerprint indexing using minutiae-based invariable set of multidimensional features | |
Parvathy et al. | FRMSDNET Classifier for Multimodal Feature Fusion Biometric Authentication | |
Andrezza et al. | A Novel Fingerprint Quality Assessment Based on Gabor Filters |
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 |