CN104331715B - Fingerprint posture antidote based on Template Learning and system - Google Patents
Fingerprint posture antidote based on Template Learning and system Download PDFInfo
- Publication number
- CN104331715B CN104331715B CN201410525149.1A CN201410525149A CN104331715B CN 104331715 B CN104331715 B CN 104331715B CN 201410525149 A CN201410525149 A CN 201410525149A CN 104331715 B CN104331715 B CN 104331715B
- Authority
- CN
- China
- Prior art keywords
- template
- candidate
- fingerprint
- point
- reference point
- 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.)
- Expired - Fee Related
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
- G06V40/12—Fingerprints or palmprints
- G06V40/13—Sensors therefor
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/24—Aligning, centring, orientation detection or correction of the image
Abstract
The present invention proposes a kind of fingerprint posture antidote based on Template Learning, includes the following steps：Multiple candidate templates are generated from several fingerprint images, and multiple candidate templates are filtered to obtain candidate template set；High-quality template is selected from candidate template set；Multiple points with predetermined fingerprint feature are chosen from current finger print image constitute candidate point set according to pre-defined rule；It is selected from candidate point set as a reference point with the highest candidate point of high-quality template similarity；Current finger print image is transformed into the correction realized under preset coordinates system to current finger print image according to reference point.The method of the embodiment of the present invention, correction accuracy height, dependable performance.The present invention also proposes a kind of fingerprint posture correction system based on Template Learning.
Description
Technical field
The present invention relates to fingerprinting technique field more particularly to a kind of fingerprint posture antidotes based on Template Learning
And system.
Background technology
Recently as fingerprint identification technology fast development, the application range of fingerprint recognition system further increases.However
Coordinate system of the same finger between the fingerprint image that different time acquires not fully is unified, and fingerprint characteristic (minutiae point is caused
Deng) there is very big difference in position in the picture.Exactly this difference affects Rapid matching and the encryption of finger print matching system.
The posture correction of fingerprint is considered as solving the problems, such as this important foundation step, and the fingerprint that different times acquire is transformed by it
Under unified coordinate system, the feature description and matching link of fingerprint are greatly simplified.
Existing fingerprint posture antidote is all based on reference point, and reference point includes mainly：Singular point, focus and higher curvature
Point.Fingerprint posture correction algorithm is for fingerprint to be unified under identical coordinate system.Once reference point detection failure, is corrected also just
The performance that can not be carried out, and finger print matching system will be seriously affected.Existing reference point definition is all based on expertise, not
The Statistical Distribution of fingerprint image is utilized well, therefore effect can't be satisfactory.
Invention content
The present invention is directed to solve at least some of the technical problems in related technologies.For this purpose, the present invention
First purpose is to propose that a kind of correction accuracy is high, the fingerprint posture antidote based on Template Learning of dependable performance.
Second object of the present invention is to propose a kind of fingerprint posture correction system based on Template Learning.
To achieve the goals above, the embodiment of first aspect present invention proposes a kind of fingerprint posture based on Template Learning
Antidote includes the following steps：Generate multiple candidate templates from several fingerprint images, and to the multiple candidate template into
Row filtering is to obtain candidate template set；High-quality template is selected from the candidate template set；According to pre-defined rule from current
Multiple points with predetermined fingerprint feature are chosen in fingerprint image constitutes candidate point set；From the candidate point set selection with
The highest candidate point of high-quality template similarity is as a reference point；The current finger print image is converted according to the reference point
The correction to the current finger print image is realized under to preset coordinates system.
Fingerprint posture antidote according to the ... of the embodiment of the present invention based on Template Learning, obtains from several fingerprint images
At least one high-quality template obtains the reference point of current finger print image using high-quality diaphragm plate, will be current according to the reference point of acquisition
Fingerprint image is remedied under scheduled coordinate system, to complete the correction of fingerprint posture.The method of the embodiment of the present invention, correction accuracy
High, dependable performance.
In some instances, described to generate multiple candidate templates from several fingerprint images and include：To fingerprint described in every width
Image is sampled to obtain multiple sampled points；Each sampled point is indicated using the description vectors of predetermined length.
In some instances, the candidate point is indicated using the description vectors of predetermined length.
In some instances, using matching error rate or retrieval success rate high-quality mould is selected from the candidate template set
Plate.
A kind of fingerprint posture correction system based on Template Learning is proposed in the embodiment of second aspect of the present invention, including：
Template generation module is filtered the multiple candidate template for generating multiple candidate templates from several fingerprint images
To obtain candidate template set, and high-quality template is selected from the candidate template set；Candidate point generation module, for according to
Pre-defined rule chooses multiple points with predetermined fingerprint feature from current finger print image and constitutes candidate point set；Computing module,
It is as a reference point with the high-quality highest candidate point of template similarity for being selected from the candidate point set；Correct mould
Block, for the current finger print image to be transformed under preset coordinates system to realize to the current finger print according to the reference point
The correction of image.
Fingerprint posture correction system according to the ... of the embodiment of the present invention based on Template Learning, template generation module refer to from several
At least one high-quality template is obtained in print image, candidate point generation module obtains the reference of current finger print image using high-quality diaphragm plate
Point, rectification module according to the reference point of acquisition by under current finger print image flame detection to scheduled coordinate system, to complete fingerprint posture
Correction.The system of the embodiment of the present invention, correction accuracy height, dependable performance.
In some instances, the template generation module generates multiple candidate templates from several fingerprint images and includes：It is right
Fingerprint image described in every width is sampled to obtain multiple sampled points；To each sampled point using predetermined length description to
Amount indicates.
In some instances, using matching error rate or retrieval success rate high-quality mould is selected from the candidate template set
Plate.
In some instances, the candidate point is indicated using the description vectors of predetermined length.
Description of the drawings
Fig. 1 is the flow chart of the fingerprint posture antidote according to an embodiment of the invention based on Template Learning；With
Fig. 2 is the structure diagram of the fingerprint posture correction system according to an embodiment of the invention based on Template Learning.
Specific implementation mode
The embodiment of the present invention is described below in detail, examples of the embodiments are shown in the accompanying drawings, wherein from beginning to end
Same or similar label indicates same or similar element or element with the same or similar functions.Below with reference to attached
The embodiment of figure description is exemplary, it is intended to for explaining the present invention, and is not considered as limiting the invention.
In the description of the present invention, in the mark of step or action front, such as " step S101 "~" step S103 "
Or (1)-(4) are only used for the purpose of the fingerprint posture antidote based on Template Learning of the description embodiment of the present invention, and cannot
It is interpreted as indicating or implies relative ranks relationship, therefore be not considered as limiting the invention.
Referring to Fig.1, a kind of fingerprint posture correction side based on Template Learning is proposed in the embodiment of first aspect present invention
Method includes the following steps：Multiple candidate templates are generated from several fingerprint images, and multiple candidate templates are filtered to obtain
Take candidate template set；High-quality template is selected from candidate template set；It is chosen from current finger print image according to pre-defined rule
Multiple points with predetermined fingerprint feature constitute candidate point set；Selection and high-quality template similarity highest from candidate point set
Candidate point it is as a reference point；Current finger print image is transformed under preset coordinates system to realize to current finger print according to reference point
The correction of image.Concrete implementation process is as follows：
Step S101, generates multiple candidate templates from several fingerprint images, and to multiple candidate templates be filtered with
Obtain candidate template set.
Specifically, in one embodiment of the invention, fingerprint image data library is constituted by several fingerprint images, to fingerprint
Every width fingerprint image in image data base is sampled to obtain multiple sampled points；Predetermined length is utilized to each sampled point
Description vectors E is indicated.
These are candidate template using the description vectors E sampled points indicated, are existed in candidate template apparent non-prime
Template, such as the field of direction consistency of the certain area of the fingerprint image are too strong.Template is repeated there is also many, i.e., similar retouches
Vector is stated in multiple fingerprint images to occur.In one embodiment of the invention, these apparent non-prime templates and again are filtered out
Multiple template, remaining be described just constitute candidate template set.
Step S102 selects high-quality template from candidate template set.
In one embodiment of the invention, the good of template is weighed using matching error rate or retrieval success rate
It goes bad and determines at least one high-quality template T.
In one embodiment of the invention, the similarity of two width fingerprint images is exported using matching algorithm.Particularly,
Before carrying out matching algorithm, the fingerprint image in fingerprint image data library need to be carried out using the candidate template that step S101 is obtained
Correction, then carries out the matching of two width fingerprint images again.
Refer to output similarity one rational threshold value of setting first to matching algorithm using the measurement of matching error rate, it will
The output of matching algorithm is converted into matching/two kinds non-matching.Then using the markup information in fingerprint image data library as standard
The matching error rate on entire fingerprint image data library is calculated, matching error rate is low, is represented as high-quality template T.It implemented
Journey is as follows：
Fetch the two width fingerprint image M from same finger1And M2, according to the scheduled method of sampling respectively to M1And M2It carries out
Sampling obtains multiple sampled points with predetermined fingerprint feature.The distribution of sampled point can be circle, square, in rectangle
Uniformly, uneven dot matrix can also be irregular figure.Predetermined fingerprint feature includes the field of direction, cyclic graph or the field of direction
With the combination of cyclic graph.By taking directional diagram as an example, but in actual use it is unlimited this.It is directive by sampling the multiple tools of acquisition in this way
Sampled point (x, y, θ).Particularly, sampled point is utilized indicates with the description vectors of description vectors E equal lengths.
Respectively from M1And M2Multiple sampled points (x, y, θ) in select and the highest sampled point of candidate template similarity as
M1And M2Reference point, i.e. the first reference point and the second reference point.Similarity measurement between vector has very much, in the reality of the present invention
It applies in example using including Euclidean distance and the method for vector angle, but is not limited to both methods.
By two width fingerprint image M of same finger1And M2It is remedied to respectively according to the first reference point and the second reference point predetermined
Coordinate system, respectively obtain correction after fingerprint image M '1With M '2。
M ' is calculated using arbitrary fingerprint matching algorithm1With M '2Between matching rate.Matching rate is more than predetermined threshold value, then sentences
It is high-quality template to determine candidate template.
For coming from two width fingerprint images of same root finger, a high-quality template should can lead to higher matching
Rate；For the two width fingerprint images from different fingers, high-quality template should make the matching rate between them small as possible.
The retrieval success rate of candidate template is evaluated, that is, detects the similarity of a fingerprint image and other fingerprint images, sentences
The disconnected fingerprint image is with the similarity of the corresponding fingerprint image in the markup information of fingerprint image data library in all similarities
Order (similarity is from high to low) is judged to retrieving successfully if order is less than predetermined threshold value.The high-quality degree of one template can
Ratio to be accounted for all retrievals using the number for retrieving successful fingerprint image is weighed, and the more high then template of this ratio is more high-quality.
It should be pointed out that the method for the embodiment of the present invention, can train and learn the high-quality template T of more than one.In order to
Ensure the complementarity between multiple high-quality template T, after the selection for completing first high-quality template T, can partly or entirely go
Except then the high-quality template T fingerprint images solved find second high-quality template T again, three or more high-quality template T
The case where can with and so on.
Step S103 chooses multiple points with predetermined fingerprint feature from current finger print image according to pre-defined rule and constitutes
Candidate point set.
Specifically, in one embodiment of the invention, current finger print image is adopted according to the scheduled method of sampling
Sample obtains multiple sampled points with predetermined fingerprint feature.The distribution of sampled point can be round, square, equal in rectangle
Even, uneven dot matrix can also be irregular figure.Predetermined fingerprint feature includes the field of direction, cyclic graph or the field of direction with
The combination of cyclic graph.By taking directional diagram as an example, but in actual use it is unlimited this.In this way by sampling the directive point of tool obtained
(x, y, θ) is used as candidate point P.In a width fingerprint image, multiple candidates of the combination of discrete several positions of selection and angle
Point P is as candidate point set.
Further, in one embodiment of the invention, candidate point P is described using the description vectors of predetermined length,
As description vectors D.Particularly, the length of description vectors D is identical as the length of description vectors E of high-quality template.
Step S104 is selected as a reference point with the highest candidate point of high-quality template similarity from candidate point set.
Specifically, the high-quality template T-phase obtained from selection in the candidate point set that step S103 is obtained and step S102
Like the highest candidate point R as a reference point of degree.Similarity measurement between vector has very much, in an embodiment of the present invention using packet
Euclidean distance and the method for vector angle are included, but is not limited to both methods.
Current finger print image is transformed under preset coordinates system to realize to current finger print figure by step S105 according to reference point
The correction of picture.
It is corrected based on reference point current finger print image.When comparing two width fingerprint image I1, I2When, it should utilize
Image after being corrected based on same template (reference point) is matched or is retrieved.It should be noted that in addition to individually correcting,
It can be according to reference point by I1Transform to I2Coordinate system；Or by I1Transform to I2Coordinate system.As long as the two images after correction
The identical point of middle coordinate represents fingerprint same position, specific transform method can there are many kinds of.To the skill of this field
All it is known for art personnel, does not repeat here.
Fingerprint posture antidote according to the ... of the embodiment of the present invention based on Template Learning, obtains from several fingerprint images
At least one high-quality template obtains the reference point of current finger print image using high-quality diaphragm plate, will be current according to the reference point of acquisition
Fingerprint image is remedied under scheduled coordinate system, to complete the correction of fingerprint posture.The method of the embodiment of the present invention, correction accuracy
High, dependable performance.
The embodiment of second aspect of the present invention proposes a kind of fingerprint posture correction system 100 based on Template Learning, such as Fig. 2
It is shown, including：Template generation module 10, candidate point generation module 20, computing module 30 and rectification module 40.
Template generation module 10 carries out multiple candidate templates for generating multiple candidate templates from several fingerprint images
Filtering selects high-quality template to obtain candidate template set from candidate template set.Candidate point generation module 20 is for pressing
Multiple points with predetermined fingerprint feature are chosen from current finger print image constitute candidate point set according to pre-defined rule.Computing module
30 is as a reference point with the highest candidate point of high-quality template similarity for being selected from candidate point set.Rectification module 40 is used for
Current finger print image is transformed into the correction realized under preset coordinates system to current finger print image according to reference point.
Template generation module 10 carries out multiple candidate templates for generating multiple candidate templates from several fingerprint images
Filtering selects high-quality template to obtain candidate template set from candidate template set.
Specifically, in one embodiment of the invention, fingerprint image data library is constituted by several fingerprint images, to fingerprint
Every width fingerprint image in image data base is sampled to obtain multiple sampled points；Predetermined length is utilized to each sampled point
Description vectors E is indicated.
These are candidate template using the description vectors E sampled points indicated, are existed in candidate template apparent non-prime
Template, such as the field of direction consistency of the certain area of the fingerprint image are too strong.Template is repeated there is also many, i.e., similar retouches
Vector is stated in multiple fingerprint images to occur.In one embodiment of the invention, these apparent non-prime templates and again are filtered out
Multiple template, remaining be described just constitute candidate template set.
In one embodiment of the invention, the good of template is weighed using matching error rate or retrieval success rate
It goes bad and determines at least one high-quality template T.
In one embodiment of the invention, the similarity of two width fingerprint images is exported using matching algorithm.Particularly,
Before carrying out matching algorithm, the candidate template that template generation module 10 obtains need to be utilized to the fingerprint image in fingerprint image data library
As being corrected, the matching of two width fingerprint images is then carried out again.
Refer to output similarity one rational threshold value of setting first to matching algorithm using the measurement of matching error rate, it will
The output of matching algorithm is converted into matching/two kinds non-matching.Then using the markup information in fingerprint image data library as standard
The matching error rate on entire fingerprint image data library is calculated, matching error rate is low, is represented as high-quality template T.It implemented
Journey is as follows：
Fetch the two width fingerprint image M from same finger1And M2, according to the scheduled method of sampling respectively to M1And M2It carries out
Sampling obtains multiple sampled points with predetermined fingerprint feature.The distribution of sampled point can be circle, square, in rectangle
Uniformly, uneven dot matrix can also be irregular figure.Predetermined fingerprint feature includes the field of direction, cyclic graph or the field of direction
With the combination of cyclic graph.By taking directional diagram as an example, but in actual use it is unlimited this.It is directive by sampling the multiple tools of acquisition in this way
Sampled point (x, y, θ).Particularly, sampled point is utilized indicates with the description vectors of description vectors E equal lengths.
Respectively from M1And M2Multiple sampled points (x, y, θ) in select and the highest sampled point of candidate template similarity as
M1And M2Reference point, i.e. the first reference point and the second reference point.Similarity measurement between vector has very much, in the reality of the present invention
It applies in example using including Euclidean distance and the method for vector angle, but is not limited to both methods.
By two width fingerprint image M of same finger1And M2It is remedied to respectively according to the first reference point and the second reference point predetermined
Coordinate system, respectively obtain correction after fingerprint image M '1With M '2。
M ' is calculated using arbitrary fingerprint matching algorithm1With M '2Between matching rate.Matching rate is more than predetermined threshold value, then sentences
It is high-quality template to determine candidate template.
For coming from two width fingerprint images of same root finger, a high-quality template should can lead to higher matching
Rate；For the two width fingerprint images from different fingers, high-quality template should make the matching rate between them small as possible.
The retrieval success rate of candidate template is evaluated, that is, detects the similarity of a fingerprint image and other fingerprint images, sentences
The disconnected fingerprint image is with the similarity of the corresponding fingerprint image in the markup information of fingerprint image data library in all similarities
Order (similarity is from high to low) is judged to retrieving successfully if order is less than predetermined threshold value.The high-quality degree of one template can
Ratio to be accounted for all retrievals using the number for retrieving successful fingerprint image is weighed, and the more high then template of this ratio is more high-quality.
It should be pointed out that the template generation module 10 of the embodiment of the present invention, can train learn more than one it is high-quality
Template T.It, can portion after the selection for completing first high-quality template T in order to ensure the complementarity between multiple high-quality template T
Point or all remove the fingerprint image that the high-quality template T has been solved, then find second high-quality template T again, three or more
The case where mostly high-quality template T can with and so on.
Candidate point generation module 20 is used to choose from current finger print image according to pre-defined rule multiple with predetermined fingerprint
The point of feature constitutes candidate point set.
Specifically, in one embodiment of the invention, current finger print image is adopted according to the scheduled method of sampling
Sample obtains multiple sampled points with predetermined fingerprint feature.The distribution of sampled point can be round, square, equal in rectangle
Even, uneven dot matrix can also be irregular figure.Predetermined fingerprint feature includes the field of direction, cyclic graph or the field of direction with
The combination of cyclic graph.By taking directional diagram as an example, but in actual use it is unlimited this.In this way by sampling the directive point of tool obtained
(x, y, θ) is used as candidate point P.In a width fingerprint image, multiple candidates of the combination of discrete several positions of selection and angle
Point P is as candidate point set.
Further, in one embodiment of the invention, candidate point P is described using the description vectors of predetermined length,
As description vectors D.Particularly, the length of description vectors D is identical as the length of description vectors E of high-quality template.
Computing module 30 is used to select with the highest candidate point of high-quality template similarity as reference from candidate point set
Point.
Specifically, selection is obtained with template generation module 10 from the candidate point set that candidate point generation module 20 obtains
High-quality template T-phase seemingly spend highest candidate point R as a reference point.Similarity measurement between vector has very much, the present invention's
It includes Euclidean distance and the method for vector angle to be used in embodiment, but is not limited to both methods.
Rectification module 40 is used to that current finger print image to be transformed under preset coordinates system to realize to current according to reference point
The correction of fingerprint image.
It is corrected based on reference point current finger print image.When comparing two width fingerprint image I1, I2When, it should utilize
Image after being corrected based on same template (reference point) is matched or is retrieved.It should be noted that in addition to individually correcting,
It can be according to reference point by I1Transform to I2Coordinate system；Or by I1Transform to I2Coordinate system.As long as the two images after correction
The identical point of middle coordinate represents fingerprint same position, specific transform method can there are many kinds of.To the skill of this field
All it is known for art personnel, does not repeat here.
Fingerprint posture correction system according to the ... of the embodiment of the present invention based on Template Learning, template generation module refer to from several
At least one high-quality template is obtained in print image, candidate point generation module obtains the reference of current finger print image using high-quality diaphragm plate
Point, rectification module according to the reference point of acquisition by under current finger print image flame detection to scheduled coordinate system, to complete fingerprint posture
Correction.The system of the embodiment of the present invention, correction accuracy height, dependable performance.
In the description of the present invention, it is to be understood that, term "center", " longitudinal direction ", " transverse direction ", " length ", " width ",
" thickness ", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom" "inner", "outside", " up time
The orientation or positional relationship of the instructions such as needle ", " counterclockwise ", " axial direction ", " radial direction ", " circumferential direction " be orientation based on ... shown in the drawings or
Position relationship is merely for convenience of description of the present invention and simplification of the description, and does not indicate or imply the indicated device or element must
There must be specific orientation, with specific azimuth configuration and operation, therefore be not considered as limiting the invention.
In addition, term " first ", " second " are used for description purposes only, it is not understood to indicate or imply relative importance
Or implicitly indicate the quantity of indicated technical characteristic.Define " first " as a result, the feature of " second " can be expressed or
Implicitly include at least one this feature.In the description of the present invention, the meaning of " plurality " is at least two, such as two, three
It is a etc., unless otherwise specifically defined.
In the present invention unless specifically defined or limited otherwise, term " installation ", " connected ", " connection ", " fixation " etc.
Term shall be understood in a broad sense, for example, it may be being fixedly connected, may be a detachable connection, or integral；Can be that machinery connects
It connects, can also be electrical connection；It can be directly connected, can also can be indirectly connected through an intermediary in two elements
The interaction relationship of the connection in portion or two elements, unless otherwise restricted clearly.For those of ordinary skill in the art
For, the specific meanings of the above terms in the present invention can be understood according to specific conditions.
In the present invention unless specifically defined or limited otherwise, fisrt feature can be with "above" or "below" second feature
It is that the first and second features are in direct contact or the first and second features pass through intermediary mediate contact.Moreover, fisrt feature exists
Second feature " on ", " top " and " above " but fisrt feature be directly above or diagonally above the second feature, or be merely representative of
Fisrt feature level height is higher than second feature.Fisrt feature second feature " under ", " lower section " and " below " can be
One feature is directly under or diagonally below the second feature, or is merely representative of fisrt feature level height and is less than second feature.
In the description of this specification, reference term " one embodiment ", " some embodiments ", " example ", " specifically show
The description of example " or " some examples " etc. means specific features, structure, material or spy described in conjunction with this embodiment or example
Point is included at least one embodiment or example of the invention.In the present specification, schematic expression of the above terms are not
It must be directed to identical embodiment or example.Moreover, particular features, structures, materials, or characteristics described can be in office
It can be combined in any suitable manner in one or more embodiments or example.In addition, without conflicting with each other, the skill of this field
Art personnel can tie the feature of different embodiments or examples described in this specification and different embodiments or examples
It closes and combines.
Although the embodiments of the present invention has been shown and described above, it is to be understood that above-described embodiment is example
Property, it is not considered as limiting the invention, those skilled in the art within the scope of the invention can be to above-mentioned
Embodiment is changed, changes, replacing and modification.
Claims (6)
1. a kind of fingerprint posture antidote based on Template Learning, which is characterized in that include the following steps：
Multiple candidate templates are generated from several fingerprint images, and the multiple candidate template are filtered to obtain candidate mould
Plate set, wherein the candidate template is the fingerprint image of the field of direction with fingerprint certain area；
High-quality template is selected from the candidate template set, wherein using matching error rate or retrieves success rate from the time
The high-quality template is selected in modeling plate set, is selected from the candidate template set using the matching error rate described excellent
The step of matter template includes：Fetch the two width fingerprint image M from same finger1And M2, according to the scheduled method of sampling respectively to M1
And M2It is sampled, obtains multiple sampled points with predetermined fingerprint feature；Respectively from fingerprint image M1With fingerprint image M2It is more
It is selected in a sampled point (x, y, θ) with the highest sampled point of candidate template similarity as M1And M2The first reference point and second
Reference point；By two width fingerprint image M of same finger1And M2It is remedied to respectively according to the first reference point and the second reference point predetermined
Coordinate system, respectively obtain correction after fingerprint image M '1With M '2；M ' is calculated using fingerprint matching algorithm1With M '2Between
With rate, if matching rate is more than predetermined threshold value, judge that candidate template is high-quality template；
Multiple points with predetermined fingerprint feature are chosen from current finger print image constitute candidate point set according to pre-defined rule；
It is selected from the candidate point set as a reference point with the high-quality highest candidate point of template similarity；And
The current finger print image is transformed under preset coordinates system to realize to the current finger print figure according to the reference point
The correction of picture.
2. the method as described in claim 1, which is characterized in that described to generate multiple candidate template packets from several fingerprint images
It includes：
Fingerprint image described in every width is sampled to obtain multiple sampled points；
Each sampled point is indicated using the description vectors of predetermined length.
3. the method as described in claim 1, which is characterized in that the candidate point is indicated using the description vectors of predetermined length.
4. a kind of fingerprint posture correction system based on Template Learning, which is characterized in that including：
Template generation module carries out the multiple candidate template for generating multiple candidate templates from several fingerprint images
Filtering and selects high-quality template to obtain candidate template set from the candidate template set, wherein the candidate template is
The fingerprint image of the field of direction with fingerprint certain area using matching error rate or retrieves success rate from the candidate template collection
The step of selecting high-quality template in conjunction, the high-quality template selected from the candidate template set using the matching error rate
Including：Fetch the two width fingerprint image M from same finger1And M2, according to the scheduled method of sampling respectively to M1And M2It is adopted
Sample obtains multiple sampled points with predetermined fingerprint feature；Respectively from fingerprint image M1With fingerprint image M2Multiple sampled points
It is selected in (x, y, θ) with the highest sampled point of candidate template similarity as M1And M2The first reference point and the second reference point；It will
Two width fingerprint image M of same finger1And M2Scheduled coordinate system is remedied to according to the first reference point and the second reference point respectively,
Respectively obtain the fingerprint image M ' after correction1With M '2；M ' is calculated using fingerprint matching algorithm1With M '2Between matching rate, if
Matching rate is more than predetermined threshold value, then judges that candidate template is high-quality template；
Candidate point generation module, it is multiple with predetermined fingerprint feature for being chosen from current finger print image according to pre-defined rule
Point constitutes candidate point set；
Computing module, for being selected from the candidate point set with the high-quality highest candidate point of template similarity as ginseng
Examination point；And
Rectification module, for the current finger print image to be transformed under preset coordinates system to realize to institute according to the reference point
State the correction of current finger print image.
5. system as claimed in claim 4, which is characterized in that the template generation module generates more from several fingerprint images
A candidate template includes：
Fingerprint image described in every width is sampled to obtain multiple sampled points；
Each sampled point is indicated using the description vectors of predetermined length.
6. system as claimed in claim 4, which is characterized in that the candidate point is indicated using the description vectors of predetermined length.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410525149.1A CN104331715B (en) | 2014-10-08 | 2014-10-08 | Fingerprint posture antidote based on Template Learning and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410525149.1A CN104331715B (en) | 2014-10-08 | 2014-10-08 | Fingerprint posture antidote based on Template Learning and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104331715A CN104331715A (en) | 2015-02-04 |
| CN104331715B true CN104331715B (en) | 2018-08-28 |
Family
ID=52406436
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410525149.1A Expired - Fee Related CN104331715B (en) | 2014-10-08 | 2014-10-08 | Fingerprint posture antidote based on Template Learning and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104331715B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110895665B (en) * | 2018-09-12 | 2023-03-31 | 上海耕岩智能科技有限公司 | Coordinate transformation method for under-screen imaging, storage medium and electronic equipment |
| JP7311310B2 (en) * | 2018-10-18 | 2023-07-19 | パナソニック インテレクチュアル プロパティ コーポレーション オブ アメリカ | Information processing device, information processing method and program |
Family Cites Families (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN100414558C (en) * | 2002-12-06 | 2008-08-27 | 中国科学院自动化研究所 | Automatic fingerprint distinguishing system and method based on template learning |
| CN101079102A (en) * | 2007-06-28 | 2007-11-28 | 中南大学 | Fingerprint identification method based on statistic method |
| CN101122950B (en) * | 2007-09-04 | 2011-01-19 | 成都方程式电子有限公司 | Fingerprint elastic deformation correction method and device |
| CN101609499B (en) * | 2008-06-20 | 2011-06-08 | 南京理工大学 | Rapid fingerprint identification method |
| US8170346B2 (en) * | 2009-03-14 | 2012-05-01 | Ludwig Lester F | High-performance closed-form single-scan calculation of oblong-shape rotation angles from binary images of arbitrary size using running sums |
| CN102262730B (en) * | 2011-08-15 | 2012-10-03 | 山东志华信息科技股份有限公司 | Fingerprint matching method based on multiple reference point pairs |
| US9092652B2 (en) * | 2012-06-29 | 2015-07-28 | Apple Inc. | Zero reference based ridge flow map |
| CN103077377B (en) * | 2012-12-31 | 2015-07-29 | 清华大学 | Based on the fingerprint correction method of field of direction distribution |
-
2014
- 2014-10-08 CN CN201410525149.1A patent/CN104331715B/en not_active Expired - Fee Related
Also Published As
| Publication number | Publication date |
|---|---|
| CN104331715A (en) | 2015-02-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US8412730B2 (en) | Image search apparatus and method thereof | |
| CN106991705B (en) | Position parameter estimation method based on P3P algorithm | |
| CN109241901B (en) | A kind of detection and recognition methods to the three-dimensional point cloud with hole | |
| KR20130025906A (en) | Face feature-point position correction device, face feature-point position correction method, and face feature-point position correction program | |
| Chen et al. | Local coordinates alignment with global preservation for dimensionality reduction | |
| CN105139007B (en) | Man face characteristic point positioning method and device | |
| CN103559420A (en) | Building method and device of anomaly detection training set | |
| CN112801094B (en) | Pointer instrument image inclination correction method | |
| CN111639647B (en) | Indicator light state identification method and device, computer equipment and storage medium | |
| CN115862821B (en) | Digital twinning-based intelligent operating room construction method and related device | |
| CN112887897A (en) | Terminal positioning method, device and computer readable storage medium | |
| CN104331715B (en) | Fingerprint posture antidote based on Template Learning and system | |
| CN114897110B (en) | Group hole measurement swing angle planning method, readable medium and equipment | |
| CN106909539A (en) | Image indexing system, server, database and related methods | |
| CN106778925B (en) | Face recognition pose over-complete face automatic registration method and device | |
| CN106909552A (en) | Image retrieval server, system, coordinate indexing and misarrangement method | |
| CN106934141B (en) | Acceleration method for calculating resistance based on long-edge cutting | |
| Wang et al. | Improving face recognition by online image alignment | |
| CN108319717B (en) | The calculation method of optimum lattice size range when a kind of vector point is converted to grid point | |
| CN113970996B (en) | Character gesture recognition method and system | |
| CN109409415A (en) | A kind of LLE algorithm kept based on global information | |
| CN106206345A (en) | The method determining to etch the engagement edge roughness of the contact hole in wafer | |
| CN110543896B (en) | Heterogeneous crowdsourcing fingerprint labeling method based on semi-supervised naive Bayes | |
| CN108664848A (en) | The recognition methods of image object and device | |
| CN106844443A (en) | The rapid correction method and system of a kind of map vector |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20180828 Termination date: 20191008 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |