CN106127111A - A kind of intelligent fingerprint verification system based on cloud and method - Google Patents

Abstract

The invention provides a kind of intelligent fingerprint verification system based on cloud and method, relate to field of biological recognition.It is characterized in that, described system includes: finger print acquisition module, image pre-processing module, characteristic extracting module, mode decision module, result display module, data transmission module, high in the clouds data transmission module, Registering modules, fingerprint database and retrieval module；Described finger print acquisition module signal is connected to image pre-processing module；Described image pre-processing module signal is connected to characteristic extracting module；Described characteristic extracting module signal is connected to mode decision module；Described mode decision module by signal is connected to data transmission module；Described data transmission module signal respectively is connected to result display module and high in the clouds data transmission module；Described high in the clouds data transmission module signal respectively is connected to Registering modules and retrieval module；Described Registering modules signal is connected to fingerprint database；Described retrieval module by signal is connected to fingerprint database.

Description

A kind of intelligent fingerprint verification system based on cloud and method

Technical field

The present invention relates to field of biological recognition, particularly to a kind of intelligent fingerprint verification system based on cloud and side Method.

Background technology

Fingerprint identification technology in whole living things feature recognition field in occupation of critical role, the most traditional fingerprint recognition system System has been also up to more satisfactory effect in fingerprint matching accuracy.But, along with the expansion of data message, a lot Under application scenario, the scale of fingerprint database is increasing, when system needs to process jumbo fingerprint database, if adopted By traditional man-to-man fingerprint recognition pattern, then will consume considerable time.This is for the stronger application of requirement of real-time For system, it is clear that be unacceptable.

Additionally, along with information technology and the fast development of cloud, and the proposition of the Internet+theory.More and more Traditional industries develop towards internet arena.Same, fingerprint recognition should also be as cloud as next breach.Due to The speed that cloud processes, can increase the efficiency of process based on distribution type processing method.And along with fingerprint management Data volume incrementally increases, and traditional locally stored method being managed also has already fallen behind.Cloud and fingerprint recognition are carried out In conjunction with, it is the next key point of this field development.

Certainly, the fast throughput relying solely on cloud is not all right all the time, in addition it is also necessary to the searching algorithm to itself Improve.The speed that the most not only can allow retrieval is accelerated, and increases effectiveness of retrieval, it is also possible to make retrieval result more accurate Really.

Summary of the invention

In consideration of it, the invention provides a kind of intelligent fingerprint verification system based on cloud and method, this invention has The advantage such as have that accuracy rate is high, speed fast, intelligent and amount of storage is big.

The technical solution used in the present invention is as follows:

A kind of intelligent fingerprint verification system, it is characterised in that described system includes: finger print acquisition module, image are located in advance Reason module, characteristic extracting module, mode decision module, result display module, data transmission module, high in the clouds data transmission module, Registering modules, fingerprint database and retrieval module；Described finger print acquisition module signal is connected to image pre-processing module；Described figure As pretreatment module signal is connected to characteristic extracting module；Described characteristic extracting module signal is connected to mode decision module；Institute State mode decision module by signal and be connected to data transmission module；Described data transmission module signal respectively is connected to result display mould Block and high in the clouds data transmission module；Described high in the clouds data transmission module signal respectively is connected to Registering modules and retrieval module；Institute State Registering modules signal and be connected to fingerprint database；Described retrieval module by signal is connected to fingerprint database.

Described finger print acquisition module, for gathering the fingerprint image data of user, locates to image in advance by view data transmission Reason module；Described image pre-processing module, carries out pretreatment processing for fingerprint image data, the view data after processing is sent out Deliver to characteristic extracting module；Described characteristic extracting module, for carrying out feature extraction to the view data after processing；Described pattern Judge module, is used for judging that this user is to need new user registration to need for carrying out fingerprint retrieval；Described data transmission module, It is connected for the signal between connection model judge module with high in the clouds data transmission module, and connects result display module and cloud Signal between end data transport module connects；Described result display module, is used for showing that high in the clouds data transmission module sends back The object information come；Described Registering modules, for registering the new finger print information of typing to fingerprint database；Described retrieval module, For retrieving the finger print information of coupling in fingerprint database.

The described image pre-processing module fingerprint image data to collecting carries out the method for pretreatment processing and includes following Step:

Step 1: fingerprint image is carried out image segmentation, extracts the foreground zone of fingerprint image；

Step 2: be standardized processing to fingerprint image by the mean pixel gray scale calculating fingerprint image so that fingerprint In the gray scale of image and the unified critical field of setting contrast to.

Step 3: be filtered image strengthening；

Step 4: image is carried out binaryzation and micronization processes.

Described Finger print characteristic abstract module carries out the method bag of feature extraction to the fingerprint image after pretreatment module process Include following steps:

Step 1: use Poincare Index algorithm to calculate the characteristic information of finger print core point, be designated as: P (x, y, z).

Step 2: when taking the fingerprint minutiae point, first sets up 8 neighborhood pictures for each pixel M in refined image Element district；Wherein M₁-M₈For the adjacent loops around pixel M around pixel；M₁-M₈The value of middle black color dots is set to 1, the value of white point It is set to 0.

Step 3: according to the G-value of the equation below each pixel of calculating:

$G=0.5{\mathrm{\Σ}}_{i=1}^8|M_i-M_{i+1}|,M_9=M_1;$

Step 4: when G-value is 1, can determine that detected M point is crestal line end points, when G-value is 3, then can determine that M point is Crestal line bifurcation；Return after minutiae point being detected this point coordinate (x, y), and returns type T of minutiae point according to the value of G, Read the deflection θ of this point the most again；Thus show that the specific features of a fingerprint minutiae is finally with M (x, y, θ, T) operator Characterize.

A kind of method based on the intelligent fingerprint verification system one of Claims 1-4 Suo Shu, it is characterised in that institute The method of stating comprises the following steps:

Step 1: after system initialization, system start-up；

Step 2: finger print acquisition module is started working, is acquired fingerprint, is sent by the fingerprint image data collected To fingerprint image preprocessing module；

Step 3: the fingerprint image preprocessing module fingerprint image to collecting carries out pretreatment；

Step 4: Finger print characteristic abstract module starts the fingerprint image after fingerprint pretreatment is carried out feature extraction；To extract After eigenvalue send to mode decision module；

Step 5: mode decision module judges that should carry out new user carries out registration and need for carrying out fingerprint retrieval；If Needs carry out new user's registration, then perform step 6, if needing to carry out fingerprint retrieval, then perform step 7；

Step 6: Registering modules sets up the index factor according to eigenvalue, and is deposited by the fingerprint characteristic value that this index factor pair is answered Store up into fingerprint database；

Step 7: retrieval module sets up the index factor according to this feature value, and retrieves in fingerprint database, generates List of matches, then mates fingerprint characteristic according to list of matches；Matching result is sent to display module.

The described index factor is:WhereinRepresent and fingerprint image is divided After m × n block, the angle matrix being made up of fingerprint image each sub-block local ridge orientation field；F_m*nRepresent the local of fingerprint Ridge frequency matrix；D represent with finger print core point P (x, y, z) centered by, in radii fixus R all minutiae point to core point away from From meansigma methods；Δ then represents from nearest three minutiae point of core point, be denoted as Δ=ω 1, ω 2, ω 3} (ω i=θ i-θ c, And-π ＜ ω i ＜ π).

The method that described retrieval module carries out fingerprint retrieval comprises the following steps:

Step 1: assume that the fingerprint index factor to be checked isThen Calculate every similarity score:

$K1=({\∂}_{m*n},{\∂}_{m*n}1)=\sqrt{{\left({\mathrm{\Σ}}_{j=1}^{m*n}cos\right(2d_j\left)\right)}^2+{\left({\mathrm{\Σ}}_{j=1}^{m*n}sin\right(d_j\left)\right)}^2};$

Wherein:

d_j=

θ[j]-

θ 1 [j], (j=1,2 ..., m*n；θ [j]) it is the local ridge orientation angle of fingerprint image jth block；

$K2=(F_{m*n},F_{m*n}1)=1-\frac{{\mathrm{\Σ}}_{j=1}^{m*n}|\frac{1}{f\left|j\right|}-\frac{1}{f1\left|j\right|}|}{{\mathrm{\Σ}}_{j=1}^{m*n}|\frac{1}{f\left|j\right|}+\frac{1}{f1\left|j\right|}|};$

$K3=(D,D1)=1-\frac{D-D1}{max(D,D1)};$

$K4=(\mathrm{\Δ},\mathrm{\Δ}1)=1-\frac{{\mathrm{\Σ}}_{i=1}^3|\mathrm{\ω}i-\mathrm{\ω}1i|}{{\mathrm{\Σ}}_{i=1}^3\mathrm{\ω}i+\mathrm{\ω}1i};$

Step 2: according to four above-mentioned Similarity value, calculates the first high speed angle value between S and S1, draws affinity score P:

P=(S, S1)=μ₁K1+μ₂K1+μ₃K3+μ₄K4；

Step 3: fingerprint to be checked is entered with all fingerprints in data base by retrieval module according to index factor comparison rule Line retrieval contrasts, and after complete fingerprint database of traversal, has obtained a series of similarity score.Finally according to set Threshold score, selects a part of fingerprint that mark is the highest, thus reduces the scope that fingerprint contrast identifies.

Use above technical scheme, present invention produces following beneficial effect:

1, accuracy rate is high: the fingerprint verification system that the present invention provides, is checking and the coupling of feature based value.Both ensure that The efficiency of system response, in turn ensure that the accuracy of matching result.

2, speed is fast: existing fingerprint verification system, along with the expansion of data message, and fingerprint number under a lot of application scenarios Scale according to storehouse is increasing, when system needs to process jumbo fingerprint database, if using traditional man-to-man Fingerprint recognition pattern, then will consume considerable time.And the fingerprint authentication of the present invention and matching system, feature based value calculates Similar value, then retrieves, and substantially increases the efficiency of fingerprint recognition under identical capacity data storehouse so that the phase of whole system Answer speed faster.

3, intelligent: the checking system of the present invention, whole proof procedure can be to need to register with Intelligent Recognition user Need for retrieving.Whole process is processed automatically by system, generates result.And processing procedure uses the intelligence after improving Energy searching algorithm, is greatly improved the accuracy of retrieval.

4, amount of storage is big: the fingerprint verification system of the present invention, finger print data storage is transferred to cloud, is greatly improved fingerprint The data volume of storage.Along with information age, being greatly promoted of data volume, traditional local data base has been difficult to meet huge number Storage according to amount.Cloud based on distributed storage then can solve this problem the most cleverly.

Accompanying drawing explanation

Fig. 1 is a kind of based on cloud intelligent fingerprint verification system and the system structure signal of method of the present invention Figure.

Detailed description of the invention

All features disclosed in this specification, or disclosed all methods or during step, except mutually exclusive Feature and/or step beyond, all can combine by any way.

Any feature disclosed in this specification (including any accessory claim, summary), unless specifically stated otherwise, By other equivalences or there is the alternative features of similar purpose replaced.I.e., unless specifically stated otherwise, each feature is a series of An example in equivalence or similar characteristics.

The embodiment of the present invention 1 provides a kind of intelligent fingerprint verification system based on cloud, system structure such as Fig. 1 Shown in:

A kind of intelligent fingerprint verification system, it is characterised in that described system includes: finger print acquisition module, image are located in advance Reason module, characteristic extracting module, mode decision module, result display module, data transmission module, high in the clouds data transmission module, Registering modules, fingerprint database and retrieval module；Described finger print acquisition module signal is connected to image pre-processing module；Described figure As pretreatment module signal is connected to characteristic extracting module；Described characteristic extracting module signal is connected to mode decision module；Institute State mode decision module by signal and be connected to data transmission module；Described data transmission module signal respectively is connected to result display mould Block and high in the clouds data transmission module；Described high in the clouds data transmission module signal respectively is connected to Registering modules and retrieval module；Institute State Registering modules signal and be connected to fingerprint database；Described retrieval module by signal is connected to fingerprint database.

Described finger print acquisition module, for gathering the fingerprint image data of user, locates to image in advance by view data transmission Reason module；Described image pre-processing module, carries out pretreatment processing for fingerprint image data, the view data after processing is sent out Deliver to characteristic extracting module；Described characteristic extracting module, for carrying out feature extraction to the view data after processing；Described pattern Judge module, is used for judging that this user is to need new user registration to need for carrying out fingerprint retrieval；Described data transmission module, It is connected for the signal between connection model judge module with high in the clouds data transmission module, and connects result display module and cloud Signal between end data transport module connects；Described result display module, is used for showing that high in the clouds data transmission module sends back The object information come；Described Registering modules, for registering the new finger print information of typing to fingerprint database；Described retrieval module, For retrieving the finger print information of coupling in fingerprint database.

The described image pre-processing module fingerprint image data to collecting carries out the method for pretreatment processing and includes following Step:

Step 1: fingerprint image is carried out image segmentation, extracts the foreground zone of fingerprint image；

Step 2: be standardized processing to fingerprint image by the mean pixel gray scale calculating fingerprint image so that fingerprint In the gray scale of image and the unified critical field of setting contrast to.

Step 3: be filtered image strengthening；

Step 4: image is carried out binaryzation and micronization processes.

Described Finger print characteristic abstract module carries out the method bag of feature extraction to the fingerprint image after pretreatment module process Include following steps:

Step 1: use Poincare Index algorithm to calculate the characteristic information of finger print core point, be designated as: P (x, y, z).

Step 2: when taking the fingerprint minutiae point, first sets up 8 neighborhood pictures for each pixel M in refined image Element district；Wherein M₁-M₈For the adjacent loops around pixel M around pixel；M₁-M₈The value of middle black color dots is set to 1, the value of white point It is set to 0.

Step 3: according to the G-value of the equation below each pixel of calculating:

$G=0.5{\mathrm{\Σ}}_{i=1}^8|M_i-M_{i+1}|,M_9=M_1;$

Step 4: when G-value is 1, can determine that detected M point is crestal line end points, when G-value is 3, then can determine that M point is Crestal line bifurcation；Return after minutiae point being detected this point coordinate (x, y), and returns type T of minutiae point according to the value of G, Read the deflection θ of this point the most again；Thus show that the specific features of a fingerprint minutiae is finally with M (x, y, θ, T) operator Characterize.

The embodiment of the present invention 2 provides a kind of intelligent fingerprint authentication method based on cloud:

A kind of method based on the intelligent fingerprint verification system one of Claims 1-4 Suo Shu, it is characterised in that institute The method of stating comprises the following steps:

Step 1: after system initialization, system start-up；

Step 2: finger print acquisition module is started working, is acquired fingerprint, is sent by the fingerprint image data collected To fingerprint image preprocessing module；

Step 3: the fingerprint image preprocessing module fingerprint image to collecting carries out pretreatment；

Step 4: Finger print characteristic abstract module starts the fingerprint image after fingerprint pretreatment is carried out feature extraction；To extract After eigenvalue send to mode decision module；

Step 5: mode decision module judges that should carry out new user carries out registration and need for carrying out fingerprint retrieval；If Needs carry out new user's registration, then perform step 6, if needing to carry out fingerprint retrieval, then perform step 7；

Step 6: Registering modules sets up the index factor according to eigenvalue, and is deposited by the fingerprint characteristic value that this index factor pair is answered Store up into fingerprint database；

Step 7: retrieval module sets up the index factor according to this feature value, and retrieves in fingerprint database, generates List of matches, then mates fingerprint characteristic according to list of matches；Matching result is sent to display module.

The described index factor is:WhereinRepresent and fingerprint image is divided After m × n block, the angle matrix being made up of fingerprint image each sub-block local ridge orientation field；F_m*nRepresent the local of fingerprint Ridge frequency matrix；D represent with finger print core point P (x, y, z) centered by, in radii fixus R all minutiae point to core point away from From meansigma methods；Δ then represents from nearest three minutiae point of core point, be denoted as Δ=ω 1, ω 2, ω 3} (ω i=θ i-θ c, And-π ＜ ω i ＜ π).

The method that described retrieval module carries out fingerprint retrieval comprises the following steps:

Step 1: assume that the fingerprint index factor to be checked isThen Calculate every similarity score:

$K1=({\∂}_{m*n},{\∂}_{m*n}1)=\sqrt{{\left({\mathrm{\Σ}}_{j=1}^{m*n}cos\right(2d_j\left)\right)}^2+{\left({\mathrm{\Σ}}_{j=1}^{m*n}sin\right(d_j\left)\right)}^2};$

Wherein:

d_j=

θ[j]-

θ 1 [j], (j=1,2 ..., m*n；θ [j]) it is the local ridge orientation angle of fingerprint image jth block；

$K2=(F_{m*n},F_{m*n}1)=1-\frac{{\mathrm{\Σ}}_{j=1}^{m*n}|\frac{1}{f\left|j\right|}-\frac{1}{f1\left|j\right|}|}{{\mathrm{\Σ}}_{j=1}^{m*n}|\frac{1}{f\left|j\right|}+\frac{1}{f1\left|j\right|}|};$

$K3=(D,D1)=1-\frac{D-D1}{max(D,D1)};$

$K4=(\mathrm{\Δ},\mathrm{\Δ}1)=1-\frac{{\mathrm{\Σ}}_{i=1}^3|\mathrm{\ω}i-\mathrm{\ω}1i|}{{\mathrm{\Σ}}_{i=1}^3\mathrm{\ω}i+\mathrm{\ω}1i};$

Step 2: according to four above-mentioned Similarity value, calculates the first high speed angle value between S and S1, draws affinity score P:

P=(S, S1)=μ₁K1+μ₂K1+μ₃K3+μ₄K4；

Step 3: fingerprint to be checked is entered with all fingerprints in data base by retrieval module according to index factor comparison rule Line retrieval contrasts, and after complete fingerprint database of traversal, has obtained a series of similarity score.Finally according to set Threshold score, selects a part of fingerprint that mark is the highest, thus reduces the scope that fingerprint contrast identifies.

Providing a kind of intelligent fingerprint verification system based on cloud and method in the embodiment of the present invention 3, system is tied Structure is as shown in Figure 1:

A kind of intelligent fingerprint verification system, it is characterised in that described system includes: finger print acquisition module, image are located in advance Reason module, characteristic extracting module, mode decision module, result display module, data transmission module, high in the clouds data transmission module, Registering modules, fingerprint database and retrieval module；Described finger print acquisition module signal is connected to image pre-processing module；Described figure As pretreatment module signal is connected to characteristic extracting module；Described characteristic extracting module signal is connected to mode decision module；Institute State mode decision module by signal and be connected to data transmission module；Described data transmission module signal respectively is connected to result display mould Block and high in the clouds data transmission module；Described high in the clouds data transmission module signal respectively is connected to Registering modules and retrieval module；Institute State Registering modules signal and be connected to fingerprint database；Described retrieval module by signal is connected to fingerprint database.

Described finger print acquisition module, for gathering the fingerprint image data of user, locates to image in advance by view data transmission Reason module；Described image pre-processing module, carries out pretreatment processing for fingerprint image data, the view data after processing is sent out Deliver to characteristic extracting module；Described characteristic extracting module, for carrying out feature extraction to the view data after processing；Described pattern Judge module, is used for judging that this user is to need new user registration to need for carrying out fingerprint retrieval；Described data transmission module, It is connected for the signal between connection model judge module with high in the clouds data transmission module, and connects result display module and cloud Signal between end data transport module connects；Described result display module, is used for showing that high in the clouds data transmission module sends back The object information come；Described Registering modules, for registering the new finger print information of typing to fingerprint database；Described retrieval module, For retrieving the finger print information of coupling in fingerprint database.

The described image pre-processing module fingerprint image data to collecting carries out the method for pretreatment processing and includes following Step:

Step 1: fingerprint image is carried out image segmentation, extracts the foreground zone of fingerprint image；

Step 2: be standardized processing to fingerprint image by the mean pixel gray scale calculating fingerprint image so that fingerprint In the gray scale of image and the unified critical field of setting contrast to.

Step 3: be filtered image strengthening；

Step 4: image is carried out binaryzation and micronization processes.

Described Finger print characteristic abstract module carries out the method bag of feature extraction to the fingerprint image after pretreatment module process Include following steps:

Step 1: use Poincare Index algorithm to calculate the characteristic information of finger print core point, be designated as: P (x, y, z).

Step 2: when taking the fingerprint minutiae point, first sets up 8 neighborhood pictures for each pixel M in refined image Element district；Wherein M₁-M₈For the adjacent loops around pixel M around pixel；M₁-M₈The value of middle black color dots is set to 1, the value of white point It is set to 0.

Step 3: according to the G-value of the equation below each pixel of calculating:

$G=0.5{\mathrm{\Σ}}_{i=1}^8|M_i-M_{i+1}|,M_9=M_1;$

Step 4: when G-value is 1, can determine that detected M point is crestal line end points, when G-value is 3, then can determine that M point is Crestal line bifurcation；Return after minutiae point being detected this point coordinate (x, y), and returns type T of minutiae point according to the value of G, Read the deflection θ of this point the most again；Thus show that the specific features of a fingerprint minutiae is finally with M (x, y, θ, T) operator Characterize.

A kind of method based on the intelligent fingerprint verification system one of Claims 1-4 Suo Shu, it is characterised in that institute The method of stating comprises the following steps:

Step 1: after system initialization, system start-up；

Step 2: finger print acquisition module is started working, is acquired fingerprint, is sent by the fingerprint image data collected To fingerprint image preprocessing module；

Step 3: the fingerprint image preprocessing module fingerprint image to collecting carries out pretreatment；

Step 4: Finger print characteristic abstract module starts the fingerprint image after fingerprint pretreatment is carried out feature extraction；To extract After eigenvalue send to mode decision module；

Step 5: mode decision module judges that should carry out new user carries out registration and need for carrying out fingerprint retrieval；If Needs carry out new user's registration, then perform step 6, if needing to carry out fingerprint retrieval, then perform step 7；

Step 6: Registering modules sets up the index factor according to eigenvalue, and is deposited by the fingerprint characteristic value that this index factor pair is answered Store up into fingerprint database；

Step 7: retrieval module sets up the index factor according to this feature value, and retrieves in fingerprint database, generates List of matches, then mates fingerprint characteristic according to list of matches；Matching result is sent to display module.

The described index factor is:WhereinRepresent and fingerprint image is divided After m × n block, the angle matrix being made up of fingerprint image each sub-block local ridge orientation field；F_m*nRepresent the local of fingerprint Ridge frequency matrix；D represent with finger print core point P (x, y, z) centered by, in radii fixus R all minutiae point to core point away from From meansigma methods；Δ then represents from nearest three minutiae point of core point, be denoted as Δ=ω 1, ω 2, ω 3} (ω i=θ i-θ c, And-π ＜ ω i ＜ π).

The method that described retrieval module carries out fingerprint retrieval comprises the following steps:

Step 1: assume that the fingerprint index factor to be checked isThen Calculate every similarity score:

$K1=({\∂}_{m*n},{\∂}_{m*n}1)=\sqrt{{\left({\mathrm{\Σ}}_{j=1}^{m*n}cos\right(2d_j\left)\right)}^2+{\left({\mathrm{\Σ}}_{j=1}^{m*n}sin\right(d_j\left)\right)}^2};$

Wherein:

d_j=

θ[j]-

θ 1 [j], (j=1,2 ..., m*n；θ [j]) it is the local ridge orientation angle of fingerprint image jth block；

$K2=(F_{m*n},F_{m*n}1)=1-\frac{{\mathrm{\Σ}}_{j=1}^{m*n}|\frac{1}{f\left|j\right|}-\frac{1}{f1\left|j\right|}|}{{\mathrm{\Σ}}_{j=1}^{m*n}|\frac{1}{f\left|j\right|}+\frac{1}{f1\left|j\right|}|};$

$K3=(D,D1)=1-\frac{D-D1}{max(D,D1)};$

$K4=(\mathrm{\Δ},\mathrm{\Δ}1)=1-\frac{{\mathrm{\Σ}}_{i=1}^3|\mathrm{\ω}i-\mathrm{\ω}1i|}{{\mathrm{\Σ}}_{i=1}^3\mathrm{\ω}i+\mathrm{\ω}1i};$

Step 2: according to four above-mentioned Similarity value, calculates the first high speed angle value between S and S1, draws affinity score P:

P=(S, S1)=μ₁K1+μ₂K1+μ₃K3+μ₄K4；

Step 3: fingerprint to be checked is entered with all fingerprints in data base by retrieval module according to index factor comparison rule Line retrieval contrasts, and after complete fingerprint database of traversal, has obtained a series of similarity score.Finally according to set Threshold score, selects a part of fingerprint that mark is the highest, thus reduces the scope that fingerprint contrast identifies.

Use above technical scheme, present invention produces following beneficial effect:

The fingerprint verification system that the present invention provides, is checking and the coupling of feature based value.Both ensure that what system responded Efficiency, in turn ensure that the accuracy of matching result.

Existing fingerprint verification system, along with the expansion of data message, the rule of fingerprint database under a lot of application scenarios Mould is increasing, when system needs to process jumbo fingerprint database, if using traditional man-to-man fingerprint recognition Pattern, then will consume considerable time.And the fingerprint authentication of the present invention and matching system, feature based value calculates similar value, Then retrieve, substantially increase the efficiency of fingerprint recognition under identical capacity data storehouse so that the corresponding speed of whole system Faster.

The checking system of the present invention, whole proof procedure can with Intelligent Recognition user be need to carry out registration need for into Line retrieval.Whole process is processed automatically by system, generates result.And processing procedure uses the intelligent retrieval after improving to calculate Method, is greatly improved the accuracy of retrieval.

The fingerprint verification system of the present invention, is transferred to cloud by finger print data storage, is greatly improved the data of fingerprint storage Amount.Along with information age, being greatly promoted of data volume, traditional local data base has been difficult to meet depositing of huge data volume Storage.Cloud based on distributed storage then can solve this problem the most cleverly.

The invention is not limited in aforesaid detailed description of the invention.The present invention expands to any disclose in this manual New feature or any new combination, and the arbitrary new method that discloses or the step of process or any new combination.

Claims (7)

1. an intelligent fingerprint verification system based on cloud, it is characterised in that described system includes: fingerprint collecting mould Block, image pre-processing module, characteristic extracting module, mode decision module, result display module, data transmission module, high in the clouds number According to transport module, Registering modules, fingerprint database and retrieval module；Described finger print acquisition module signal is connected to Image semantic classification Module；Described image pre-processing module signal is connected to characteristic extracting module；Described characteristic extracting module signal is connected to pattern Judge module；Described mode decision module by signal is connected to data transmission module；Described data transmission module signal respectively connects In result display module and high in the clouds data transmission module；Described high in the clouds data transmission module signal respectively be connected to Registering modules and Retrieval module；Described Registering modules signal is connected to fingerprint database；Described retrieval module by signal is connected to fingerprint database.

2. intelligent fingerprint verification system based on cloud as claimed in claim 1, it is characterised in that described fingerprint collecting Module, for gathering the fingerprint image data of user, sends view data to image pre-processing module；Described Image semantic classification Module, carries out pretreatment processing for fingerprint image data, sends the view data after processing to characteristic extracting module；Described Characteristic extracting module, for carrying out feature extraction to the view data after processing；Described mode decision module, is used for judging this use Family is to need new user registration to need for carrying out fingerprint retrieval；Described data transmission module, for connection model judge module It is connected with the signal between the data transmission module of high in the clouds, and between connection result display module and high in the clouds data transmission module Signal connects；Described result display module, for showing that high in the clouds data transmission module sends the object information of returning；Described registration Module, for registering the new finger print information of typing to fingerprint database；Described retrieval module, for retrieving in fingerprint database The finger print information of coupling.

3. intelligent fingerprint verification system based on cloud as claimed in claim 2, it is characterised in that described image is located in advance The reason module fingerprint image data to collecting carries out the method for pretreatment processing and comprises the following steps:

Step 1: fingerprint image is carried out image segmentation, extracts the foreground zone of fingerprint image；

Step 2: be standardized processing to fingerprint image by the mean pixel gray scale calculating fingerprint image so that fingerprint image Gray scale and the unified critical field of setting contrast in.

Step 3: be filtered image strengthening；

Step 4: image is carried out binaryzation and micronization processes.

4. intelligent fingerprint verification system based on cloud as claimed in claim 3, it is characterised in that described fingerprint characteristic The method that extraction module carries out feature extraction to the fingerprint image after pretreatment module process comprises the following steps:

Step 1: use Poincare Index algorithm to calculate the characteristic information of finger print core point, be designated as: P (x, y, z).

Step 2: when taking the fingerprint minutiae point, first sets up 8 neighborhood territory pixels for each pixel M in refined image District；Wherein M₁-M₈For the adjacent loops around pixel M around pixel；M₁-M₈The value of middle black color dots is set to 1, and the value of white point sets It is 0.

Step 3: according to the G-value of the equation below each pixel of calculating:

M₉=M₁；

Step 4: when G-value is 1, can determine that detected M point is crestal line end points, when G-value is 3, then can determine that M point is crestal line Bifurcation；(x y), and returns type T of minutiae point, then according to the value of G to return the coordinate of this point after minutiae point being detected Read the deflection θ of this point again；Thus show that the specific features of a fingerprint minutiae finally carrys out table with M (x, y, θ, T) operator Levy.

5. a method based on based on cloud the intelligent fingerprint verification system one of Claims 1-4 Suo Shu, it is special Levy and be, said method comprising the steps of:

Step 1: after system initialization, system start-up；

Step 2: finger print acquisition module is started working, is acquired fingerprint, is sent by the fingerprint image data collected and extremely refers to Print image pretreatment module；

Step 3: the fingerprint image preprocessing module fingerprint image to collecting carries out pretreatment；

Step 4: Finger print characteristic abstract module starts the fingerprint image after fingerprint pretreatment is carried out feature extraction；After extracting Eigenvalue sends to mode decision module；

Step 5: mode decision module judges that should carry out new user carries out registration and need for carrying out fingerprint retrieval；The need to Carrying out new user's registration, then perform step 6, if needing to carry out fingerprint retrieval, then performing step 7；

Step 6: Registering modules sets up the index factor according to eigenvalue, and is stored into by the fingerprint characteristic value that this index factor pair is answered Fingerprint database；

Step 7: retrieval module sets up the index factor according to this feature value, and retrieves in fingerprint database, generates coupling List, then mates fingerprint characteristic according to list of matches；Matching result is sent to display module.

6. intelligent fingerprint authentication method based on cloud as claimed in claim 5, it is characterised in that the described index factor For:WhereinRepresent after fingerprint image is divided into m × n block, by fingerprint The angle matrix of image each sub-block local ridge orientation field composition；F_m*nRepresent the local ridge frequency matrix of fingerprint；D represents With finger print core point P (x, y, z) centered by, in radii fixus R, all minutiae point are to the meansigma methods of core point distance；Δ then represents From three minutiae point that core point is nearest, it is denoted as Δ={ ω 1, ω 2, ω 3} (ω i=θ i-θ c, and-π < ω i < π).

7. based on cloud the intelligent fingerprint authentication method as described in claim 5 or 6, it is characterised in that described retrieval Module carries out the method for fingerprint retrieval and comprises the following steps:

Step 1: assume that the fingerprint index factor to be checked isThen calculate Every similarity score:

$K1=({\&part;}_{m*n},{\&part;}_{m*n}1)=\sqrt{{\left({\mathrm{\&Sigma;}}_{j=1}^{m*n}cos\right(2d_j\left)\right)}^2+{\left({\mathrm{\&Sigma;}}_{j=1}^{m*n}sin\right(d_j\left)\right)}^2};$

Wherein:

d_j=

θ[j]-

θ 1 [j], (j=1,2 ..., m*n；θ [j]) it is the local ridge orientation angle of fingerprint image jth block；

$K2=(F_{m*n},F_{m*n}1)=1-\frac{{\mathrm{\&Sigma;}}_{j=1}^{m*n}|\frac{1}{f\left|j\right|}-\frac{1}{f1\left|j\right|}|}{{\mathrm{\&Sigma;}}_{j=1}^{m*n}|\frac{1}{f\left|j\right|}+\frac{1}{f1\left|j\right|}|};$

$K3=(D,D1)=1-\frac{D-D1}{max(D,D1)};$

$K4=(\mathrm{\&Delta;},\mathrm{\&Delta;}1)=1-\frac{{\mathrm{\&Sigma;}}_{i=1}^3|\mathrm{\&omega;}i-\mathrm{\&omega;}1i|}{{\mathrm{\&Sigma;}}_{i=1}^3\mathrm{\&omega;}i+\mathrm{\&omega;}1i};$

Step 2: according to four above-mentioned Similarity value, calculates the first high speed angle value between S and S1, draws affinity score P:

P=(S, S1)=μ₁K1+μ₂K1+μ₃K3+μ₄K4；

Step 3: fingerprint to be checked is examined with all fingerprints in data base by retrieval module according to index factor comparison rule Rope contrasts, and after complete fingerprint database of traversal, has obtained a series of similarity score.Finally according to the threshold value set Mark, selects a part of fingerprint that mark is the highest, thus reduces the scope that fingerprint contrast identifies.