CN101145198A - Certainty coding method and system in fingerprint identification - Google Patents
Certainty coding method and system in fingerprint identification Download PDFInfo
- Publication number
- CN101145198A CN101145198A CNA200710122087XA CN200710122087A CN101145198A CN 101145198 A CN101145198 A CN 101145198A CN A200710122087X A CNA200710122087X A CN A200710122087XA CN 200710122087 A CN200710122087 A CN 200710122087A CN 101145198 A CN101145198 A CN 101145198A
- Authority
- CN
- China
- Prior art keywords
- fingerprint
- point
- image
- pixel
- know
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Landscapes
- Collating Specific Patterns (AREA)
Abstract
The present invention belongs to the computer biometrics technology and relates to a method and a system which adopt certainty coding to accurately match fingerprints during the process of fingerprint comparison. The present invention has the accurate matching by using characteristic of the certainty coding that a characteristic area can be selected from the fingerprint image and the characteristic value corresponding to the characteristic area, such as the total length of fingerprint ridge line, the entropy of the main direction of the direction field of the ridge line and the direction field of the pixel point, to conduct matching work. The certainty coding method and the fingerprint identification system built by the method overcome the defects, such as low accuracy rate and potential safety problems which are caused by adopting fuzzy matching algorithm in the traditional point fingerprint identification system.
Description
Technical field
The invention belongs to computer biometrics technology, particularly a kind ofly adopt deterministic encoding to carry out the method for fingerprint recognition and adopt the constructed fingerprint recognition system of this method.
Background technology
Existing fingerprint identification method is based on the Method of Fuzzy Matching of dot pattern.The bifurcation of definition fingerprint ridge line and the end points of fingerprint ridge line are the unique point of fingerprint.At " Azriel Rosenfeld.Point pattern matching by relaxation " (Pattern Recognition, 1980,12 (5): 269~275) in the literary composition, author Ranade and Rosenfeld have proposed the relaxed algorithm of some coupling: definition point set L1=P1...Pm and point set L2=Q1...Qm.(Pi Qj), defines a relative transformation Tr ij of two point sets for a pair of coupling.According to the matching degree of two concentrated all the other points of point under the TRij conversion, calculate the fiduciary level of TRij.If exist a conversion TRO that point set L1 and point set L2 are reasonably mated, then when certain TRij ≈ TR0, the fiduciary level of TRij is big and fiduciary levels other conversion couplings are less, and fiduciary level is with iterative computation.The major defect of this method is: the extraction of (1) feature point set depends on collecting device, gathers the performance of environment and image Preprocessing Algorithm, and therefore, the error that feature point set extracts has a strong impact on the accuracy of fingerprint recognition system; (2) be the mapping relations of one-to-many between the feature point set of fingerprint image and its feature of description.The fingerprint image of the same finger that collects under the different situations is not quite similar, and the feature point set of extraction also is not quite similar, and therefore, recognition system adopts the pattern of fuzzy matching, and system mistake reject rate (FRR) is higher with false acceptance rate (FAR), has potential safety hazard; (3) owing to adopt exhaustive Method of Fuzzy Matching, therefore, operand is big, and identification is slow, is not suitable for applied environment in real time such as fingerprint access control, password authentification;
Summary of the invention
The objective of the invention is to propose a kind of method that adopts deterministic encoding to carry out fingerprint recognition.This method adopts the deterministic encoding feature to realize the accurate coupling of fingerprint recognition process.Utilize this method only can in fingerprint image, extract and the individual relevant determinacy coding characteristic (fingerprinting DNA) of fingerprint, and utilize the deterministic encoding feature to realize that the fingerprint recognition process accurately mates, thereby overcome tradition low based on the accuracy that only adopts fuzzy matching algorithm to cause in the fingerprint recognition system of dot pattern, have a shortcoming such as potential safety hazard.
Method of the present invention is characterized in that in the process of fingerprint comparison, adopts deterministic encoding feature realization fingerprint recognition process accurately to mate.Saidly utilizing the deterministic encoding feature accurately to mate to be by in fingerprint image, choosing the characteristic area of a certain given shape, and obtain the pairing eigenwert of this characteristic area---the principal direction of fingerprint ridge line length overall, the crestal line field of direction and the direction entropy of pixel mate.This method in DSP successively according to the following steps:
(1) in the fingerprint recognition process, by the fingerprint image acquisition equipment that links to each other with central processing unit (CPU) through bus, gather a fingerprint as template fingerprint, be stored in the internal storage, and by the image display display image.
(2) the fingerprint image acquisition equipment by linking to each other with central processing unit (CPU) through bus is gathered a fingerprint as fingerprint to be known, is stored in the internal storage, and by the image display display image.
(3), the pixel value of each pixel is normalized to [0,1] interval at template fingerprint and wait to know in the image of fingerprint, wherein pixel value is that 0 point is a black color dots, pixel value is that 1 point is a white point, obtains template fingerprint and waits to know the normalized image of fingerprint, and computing formula is
(4) respectively at the said template fingerprint of step (3) and wait to know in the normalized image of fingerprint, fingerprint ridge line is refined as the streakline that has only a pixel width, thinning method is for said normalized image, do not change under the prerequisite of fingerprint ridge line length satisfied, it with pixel value 0 pixel, its pixel value changes to 1, obtains template fingerprint and the refinement figure that waits to know fingerprint;
(5) deterministic encoding Feature Extraction and coupling:
(5.1) choose the masterplate fingerprint respectively and wait to know the crestal line end points of fingerprint as unique point to be selected;
(5.2) respectively template fingerprint refinement figure is followed the tracks of and discrete sampling with the streakline of waiting to know all unique point to be selected places among the fingerprint thinning figure, every setting value D
0Individual pixel is sampled a bit, D
0Span is 8-12, and the coordinate of recording pixel point unified is set every streak line sampling N point, and the N span is 3-5, with unique point this as the 0th sampled point, if the streakline curtailment is then abandoned sampling with sampling N point;
(5.3) treat the unique point arbitrary to be selected known in the fingerprint and the unique point arbitrary to be selected in the template fingerprint, calculate the degree of fitting function of 2 streak lines at 2 unique point places by following formula
Wherein,
d
I, 0---wait to know among the fingerprint thinning figure distance between the i sampled point and the 0th sampled point (unique point) on the said sampling streakline, the i span is 1 to N-1;
D
I, 0---the distance between i sampled point and the 0th sampled point (unique point) on the said sampling streakline among the template fingerprint refinement figure, the i span is 1 to N-1;
If F
dGreater than a certain setting value threshold value F
0, assert tentatively that then the pairing unique point to be selected of this two streak line is a reference pair of points, keep this to unique point, carry out (5.4), if can not satisfy F
dGreater than setting value threshold value F
0, then give up this to unique point, repeat (5.3) said F
0Span 0.01-0.1;
(5.4) said wantonly 2 unique points to be selected of waiting to know in image and the template image in (5.1) are repeated (5.3), obtain reference point collection;
(5.5) carrying out the fingerprint image posture according to the following steps corrects;
(5.5.1) reference point that obtains in (5.4) is to concentrating, and is optional a pair of right as reference point;
(5.5.2) calculating waits to know the translation parameters (t of image with respect to template image according to the following steps
x, t
v) and anglec of rotation θ:
Wherein,
d
Kx, d
Ky---wait to know the horizontal ordinate and the ordinate of k sampled point of this reference point place streakline in the fingerprint image, the k span is 0 to N-1
D
Kx, D
Xy---the horizontal ordinate and the ordinate of k sampled point of this reference point place streakline in the template fingerprint image, k span are 0 to N-1
t
x, t
yBe respectively and wait to know fingerprint image, wait to know fingerprint and treating in the reference point known tangential direction poor of the tangential direction of fingerprint feature point place crestal line and template fingerprint unique point place crestal line for this with respect to the anglec of rotation θ of template fingerprint with respect to template fingerprint image transverse translation parameter and longitudinal translation parameter;
(5.5.3) with all pixel coordinates of masterplate fingerprint (x, y) by following formula be converted into the pixel coordinate (u, v), computing formula is:
Wherein, ρ is the image coefficient of dilatation, ρ=1; t
x, t
yBe respectively that step (5.5.2) is resulting waits to know fingerprint image with respect to template fingerprint image transverse translation parameter and longitudinal translation parameter; θ waits to know the anglec of rotation of fingerprint with respect to template fingerprint;
(5.6) accurately mate according to the following steps;
(5.6.1) for (5.5.1) said reference pair of points, respectively with the reference point of template image and the reference point of waiting to know image be the center, at (4) said template fingerprint and wait to know among the refinement figure of fingerprint, do a characteristic area of the same area by setting shape;
(5.6.2) respectively at (5.6.1) said template fingerprint refinement figure and wait to know in the characteristic area of fingerprint thinning figure, calculate following deterministic encoding eigenwert, comprise the principal direction of fingerprint ridge line total length, the crestal line field of direction and 3 parameters of direction entropy of pixel, wherein:
The fingerprint ridge line total length is 0 pixel number for this corresponding characteristic area interior pixel value;
The principal direction of the crestal line field of direction is tried to achieve according to the following steps:
It at first is the tangential direction of the fingerprint ridge line at each pixel place in the unit calculated characteristics zone with the degree, secondly the tangential direction of the fingerprint ridge line at statistical pixel point place equals the pixel number of a certain angle value, be horizontal ordinate once more with the angle, the pixel number is that ordinate is set up histogram, and the opposite direction of getting the minimum angle direction of pixel number in the histogram at last is the principal direction of the crestal line field of direction;
The direction entropy of pixel is defined as:
P wherein
jBe that spending with 45 degree with 360 serves as to be divided into 8 direction territories at interval, all pixel numbers in j direction territory account for the pixel number purpose ratio in whole 8 direction territories;
(5.6.3) template fingerprint that obtains for (5.6.2) and wait to know the determinacy coding characteristic value of fingerprint is separately converted to the binary result that usefulness " 01 " string list shows with the decimal computation result of each parameter;
(5.6.4) respectively (5.6.3) being obtained template fingerprint and waiting knows encode every binary numeral of 2 binary result of each parameter of the determinacy of fingerprint and compares, if every bit value comparison result of 3 parameters is identical, assert that then two fingerprints are by comparison, the result of output fingerprint matching success, otherwise, change (5.5), select other reference pair of points, compare to (5.6.4) by (5.5);
(5.6.5) pass through comparison corresponding to said two fingerprints, (5.4) said reference point is accurately mated by (5.6) reference point each that concentrate, not by comparison, think that then two fingerprints can not be by contrast, the unsuccessful result of output fingerprint matching as all;
System of the present invention is characterized in that it contains central processing unit, through the fingerprint image acquisition equipment that bus links to each other with central processing unit, image display, keyboard, mouse and internal storage.
Use fingerprint identification method of the present invention,, background noise dissimilar to=30 on 3 width of cloth image/finger x10 finger, image size, picker's sex, the image at age carry out test shows, and the recognition correct rate of this method is FRR=2/60=0.0333333, FAR=1/405=0.00246914.Adopt aforementioned existing fingerprint identification method based on dot pattern, recognition correct rate then is FRR=5/60=0.116667, FAR=2/405=0.00493827.
Superiority of the present invention is: owing to adopted the method for deterministic encoding to carry out fingerprint recognition, promptly at the conventional point pattern matching algorithm after tentatively the match is successful, further utilize the deterministic encoding feature to realize the accurate coupling of fingerprint recognition process, therefore, overcome tradition low based on the accuracy that adopts fuzzy matching algorithm to cause in the fingerprint recognition system of dot pattern, have a shortcoming such as potential safety hazard.The system that this employing deterministic encoding carries out the method for fingerprint recognition and adopts this method to make up can be used for the retrieval of fingerprint recognition, extensive fingerprint base, and the applied environments such as gate control system, password authentification that are used for the higher military restricted zone of level of security, bank, state security department.
Description of drawings
Fig. 1 is applicable to the computer system of the embodiment of the invention.
Fig. 2 deterministic encoding fingerprint recognition system process flow diagram.
Fig. 3 sets shape and gets circular characteristic area synoptic diagram.
Embodiment
The deterministic encoding method mainly may further comprise the steps in the fingerprint recognition that proposes among the present invention, is described in detail as follows in conjunction with the accompanying drawings:
Fig. 1 has described a computer system that is applicable to the embodiment of the invention.This computer system comprises central processing unit (CPU) and some input-output device, as keyboard, mouse, display, can also comprise fingerprint image acquisition equipment etc.Implement software of the present invention and be stored in the internal memory, CPU can internally deposit into line access, and carries out command adapted thereto, to obtain result of implementation of the present invention.
Fig. 2 is a deterministic encoding fingerprint recognition system process flow diagram.Comprise among the figure that original image collection, image binaryzation and refinement, deterministic encoding extract and codes match four parts.
Provide some the relevant definition in the specific embodiment at first earlier, in this embodiment, set the characteristic area of shape and get border circular areas, be called characteristic circle.Fig. 3 is fingerprint characteristic circle zone definitions synoptic diagram.
Characteristic circle: with unique point (fingerprint ridge line bifurcation, end points) is the center of circle, and setting value R is a border circular areas of radius, and the scope of R is a 50-80 length in pixels, is called the characteristic circle of this unique point.
The fingerprint ridge line total length is 1 pixel number for this corresponding characteristic area interior pixel value in the characteristic circle;
The principal direction of the crestal line field of direction is tried to achieve according to the following steps in the characteristic circle:
It at first is the tangential direction of the fingerprint ridge line at each pixel place in the unit calculated characteristics zone with the degree, secondly the tangential direction of the fingerprint ridge line at statistical pixel point place equals the pixel number of a certain angle value, be horizontal ordinate once more with the angle, the pixel number is that ordinate is set up histogram, and the opposite direction of getting the minimum angle direction of pixel number in the histogram at last is the principal direction of the crestal line field of direction;
The direction entropy of characteristic circle interior pixel point is defined as:
P wherein
jBe that spending with 45 degree with 360 serves as to be divided into 8 direction territories at interval, all pixel numbers in j direction territory account for the pixel number purpose ratio in whole 8 direction territories;
1, fingerprint image is handled and feature point extraction
Before the characteristic information of deterministic encoding method extracts in fingerprint recognition, at first carry out fingerprint image preprocessing by binaryzation and refinement, obtain the refinement figure of fingerprint image, end points and bifurcation information secondly take the fingerprint on the refinement figure of fingerprint image.
1) original image binaryzation and refinement
At template fingerprint and wait to know in the image of fingerprint, the pixel value of each pixel is normalized to [0,1] interval, wherein pixel value is that 0 point is a black color dots, and pixel value is that 1 point is a white point, obtains template fingerprint and waits to know the normalized image of fingerprint, and computing formula is
Respectively at template fingerprint and wait to know in the normalized image of fingerprint, fingerprint ridge line is refined as the streakline that has only a pixel width, thinning method is for said normalized image, do not change under the prerequisite of fingerprint ridge line length satisfied, it with pixel value 0 pixel, its pixel value changes to 1, obtains template fingerprint and the refinement figure that waits to know fingerprint;
2) unique point that takes the fingerprint
(x y), makes array N for arbitrary pixel
9(x comprises for describing that y) (x is y) with 3 * 3 point sets of its 8 neighborhood territory pixel point, as if (x, pixel value y) are 0, then according to N
9(x, value judging point y) (x, type y): if N
9(x, y) in, remove point (x, y) outside, have and have only 1 black color dots, then (x y) is end points to point; If N
9(x, y) in, remove point (x, y) outside, have and have only 3 black color dots, then (x y) is bifurcation to point; All the other situations are not considered.
2, deterministic encoding feature extraction and coupling
1) unique point chooses
If fingerprint feature point satisfies two conditions, then as the sampling unique point: it is the center that P point is in fingerprint ridge line end points and the fingerprint image with the P point, with setting value R
0For there not being the further feature point in the zone of radius.Not having further feature point in certain zone around the requirement sampling unique point then is because of such unique point reliability height, R
0Value can control the sampling unique point quantity.
2) streakline discrete sampling
Template and the streakline of waiting to know all sampling unique point places in the image are followed the tracks of, every D
0Individual pixel is sampled a bit, i.e. the coordinate of recording pixel point.Unified set every streak line sampling N point, with unique point this as the 0th sampled point.If streakline length is not enough, then abandon sampling, do not keep any information of the streakline of following the tracks of.
What 3) reference point was right tentatively determines
The degree of fitting function of two fingerprint ridge lines of definition is:
Wherein,
d
I, 0---wait to know in the image distance between the i sampled point and the 0th sampled point (unique point) on a certain sampling streakline
D
I, 0---the distance between i sampled point and the 0th sampled point (unique point) on the same sampling streakline in the template image
Treat and know in the image arbitrary sampling unique point in the arbitrary sampling unique point and template image, calculate their degree of fitting function.If F
dGreater than setting value threshold value F
0, assert tentatively that then the pairing unique point of this two streak line is a reference pair of points.
4) the right further screening of reference point
Right to each to the reference point of preliminary identification, waiting to know search 3 unique points nearest in image and the template image,, form two subpatterns that respectively comprise 4 unique points respectively together with himself with it.Investigate the similarity of these two subpatterns,, then this is deleted reference point if similarity degree is not high; Otherwise kept.
5) images of gestures is corrected
The reference point centering that remains, optional a pair of as reference point, calculate and wait to know translation, the rotation parameter of image with respect to template image:
Wherein,
d
Ix, d
Iy---wait to know the horizontal ordinate and the ordinate of each sampled point of i of this unique point place streakline in the fingerprint image.
D
Ix, D
Iy---the horizontal ordinate and the ordinate of each sampled point of i of this unique point place streakline in the template fingerprint image.
t
x, t
yBe respectively and wait to know fingerprint image, wait to know fingerprint and try to achieve by calculating this difference the direction of reference point with respect to the anglec of rotation θ of template fingerprint with respect to template fingerprint image transverse translation, longitudinal translation and the anglec of rotation.Masterplate fingerprint image and fingerprint image to be matched are corrected to same rectangular coordinate system, and computing formula is:
Wherein, (x y) is the rectangular coordinate of template fingerprint pixel, (u, v) for waiting to know the rectangular coordinate of fingerprint pixel, ρ is the image coefficient of dilatation, θ waits to know the anglec of rotation of fingerprint with respect to template fingerprint.
6) accurately mate
With the template image that remains and any reference pair of points of waiting to know in the image is the center of circle, setting value R
0Be radius constitutive characteristic circle.Direction entropy in field of direction principal direction and the circle in the interior streakline total length of circle of difference calculated characteristics circle, the garden.The decimal computation result is converted into scale-of-two " 01 " string list shows, and two scale-of-two deterministic encoding features that obtain are carried out the comparison of each numerical digit numerical value.Identical as each numerical digit numerical value, assert that then two fingerprints are by comparison.Otherwise, change the 5th) and the step, use other reference point to comparing.By comparison, think then that two fingerprints contrasts can not pass through as all.
Claims (2)
1. the method for a fingerprint recognition is characterized in that this fingerprint identification method finishes successively according to the following steps in DSP:
Step (1) is gathered a fingerprint as template fingerprint;
Step (2) is gathered a fingerprint to be identified as fingerprint to be known;
Step (3) at template fingerprint and wait to know in the image of fingerprint, normalizes to [0 with the pixel value of each pixel, 1] interval, wherein pixel value is that 0 point is a black color dots, pixel value is that 1 point is a white point, obtain template fingerprint and the normalized image of waiting to know fingerprint, computing formula is
Said setting value threshold value T span is 0.5-0.7;
Step (4), respectively at the said template fingerprint of step (3) and wait to know in the normalized image of fingerprint, fingerprint ridge line is refined as the streakline that has only a pixel width, thinning method is for said normalized image, do not change under the prerequisite of fingerprint ridge line length satisfied, with pixel value is 0 pixel, and its pixel value changes to 1, obtains template fingerprint and waits to know the refinement figure of fingerprint;
Step (5), deterministic encoding Feature Extraction and coupling:
Step (5.1) is chosen the masterplate fingerprint respectively and is waited to know the crestal line end points of fingerprint as unique point to be selected;
Step (5.2) is followed the tracks of and discrete sampling with the streakline of waiting to know all unique point to be selected places among the fingerprint thinning figure template fingerprint refinement figure respectively, every setting value D
0Individual pixel is sampled a bit, D
0Span is 8-12, and the coordinate of recording pixel point unified is set every streak line sampling N point, and the N span is 3-5, with unique point this as the 0th sampled point, if the streakline curtailment is then abandoned sampling with sampling N point;
Step (5.3) is treated the unique point arbitrary to be selected known in the fingerprint and the unique point arbitrary to be selected in the template fingerprint, calculates the degree of fitting function of 2 streak lines at 2 unique point places by following formula
Wherein,
d
J, 0---wait to know among the fingerprint thinning figure distance between the i sampled point and the 0th sampled point (unique point) on the said sampling streakline, the i span is 1 to/N-1;
D
J, 0---the distance between i sampled point and the 0th sampled point (unique point) on the said sampling streakline among the template fingerprint refinement figure, the i span is 1 to N-1;
If F
dGreater than a certain setting value threshold value F
0, assert tentatively that then the pairing unique point to be selected of this two streak line is a reference pair of points, keep this to unique point, carry out step (5.4), if can not satisfy F
dGreater than setting value threshold value F
0, then give up this to unique point, repeating step (5.3), said F
0Span 0.01-0.1;
Step (5.4) to said wantonly 2 unique point repeating steps to be selected (5.3) of waiting to know in image and the template image in the step (5.1), obtains reference point to collection;
Step (5.5) is carried out the fingerprint image posture according to the following steps and is corrected;
Step (5.5.1), the reference point that obtains in step (5.4) is to concentrating, and is optional a pair of right as reference point;
Step (5.5.2) is calculated according to the following steps and is waited to know the translation parameters (t of image with respect to template image
x, t
y) and anglec of rotation θ:
Wherein,
d
Kx, d
Ky---wait to know the horizontal ordinate and the ordinate of k sampled point of this reference point place streakline in the fingerprint image, the k span is 0 to N-1
D
Kx, D
Ky---the horizontal ordinate and the ordinate of k sampled point of this reference point place streakline in the template fingerprint image, k span are 0 to N-1
t
x, t
yBe respectively and wait to know fingerprint image, wait to know fingerprint and treating in the reference point known tangential direction poor of the tangential direction of fingerprint feature point place crestal line and template fingerprint unique point place crestal line for this with respect to the anglec of rotation θ of template fingerprint with respect to template fingerprint image transverse translation parameter and longitudinal translation parameter;
Step (5.5.3), with all pixel coordinates of masterplate fingerprint (x, y) by following formula be converted into the pixel coordinate (u, v), computing formula is:
Wherein, ρ is the image coefficient of dilatation, ρ=1; t
x, t
yBe respectively that step (5.5.2) is resulting waits to know fingerprint image with respect to template fingerprint image transverse translation parameter and longitudinal translation parameter; θ waits to know the anglec of rotation of fingerprint with respect to template fingerprint;
Step (5.6) is accurately mated according to the following steps;
Step (5.6.1), for the said reference pair of points of step (5.5.1), respectively with the reference point of template image and the reference point of waiting to know image be the center, at the said template fingerprint of step (4) with wait to know among the refinement figure of fingerprint, do a characteristic area of the same area by setting shape;
Step (5.6.2), respectively at the said template fingerprint refinement of step (5.6.1) figure and wait to know in the characteristic area of fingerprint thinning figure, calculate following deterministic encoding eigenwert, comprise the principal direction of fingerprint ridge line total length, the crestal line field of direction and 3 parameters of direction entropy of pixel, wherein:
The fingerprint ridge line total length is 0 pixel number for this corresponding characteristic area interior pixel value;
The principal direction of the crestal line field of direction is tried to achieve according to the following steps:
It at first is the tangential direction of the fingerprint ridge line at each pixel place in the unit calculated characteristics zone with the degree, secondly the tangential direction of the fingerprint ridge line at statistical pixel point place equals the pixel number of a certain angle value, be horizontal ordinate once more with the angle, the pixel number is that ordinate is set up histogram, and the opposite direction of getting the minimum angle direction of pixel number in the histogram at last is the principal direction of the crestal line field of direction;
The direction entropy of pixel is defined as:
P wherein
jBe that spending with 45 degree with 360 serves as to be divided into 8 direction territories at interval, all pixel numbers in j direction territory account for the pixel number purpose ratio in whole 8 direction territories;
Step (5.6.3), the template fingerprint that obtains for step (5.6.2) and wait to know the determinacy coding characteristic value of fingerprint is separately converted to the binary result that usefulness " 01 " string list shows with the decimal computation result of each parameter;
Step (5.6.4), respectively step (5.6.3) being obtained template fingerprint and waiting knows encode every binary numeral of 2 binary result of each parameter of the determinacy of fingerprint and compares, if every bit value comparison result of 3 parameters is identical, assert that then two fingerprints are by comparison, the result of output fingerprint matching success, otherwise, step (5.5) changeed, select other reference pair of points, compare to (5.6.4) (5.5) set by step;
Step (5.6.5), corresponding to said two fingerprints by comparison, to the said reference point of step (5.4) to concentrate each to reference point set by step (5.6) accurately mate, as all by comparison, think that then two fingerprints can not be by contrast, the unsuccessful result of output fingerprint matching.
2. realize the designed fingerprint recognition system of the accurate matching process of fingerprint according to the determinacy coding characteristic of claim 1, it is characterized in that, it contains central processing unit, through the fingerprint image acquisition equipment that bus links to each other with central processing unit, image display, keyboard, mouse and internal storage.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CNB200710122087XA CN100470579C (en) | 2007-09-21 | 2007-09-21 | Certainty coding method and system in fingerprint identification |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CNB200710122087XA CN100470579C (en) | 2007-09-21 | 2007-09-21 | Certainty coding method and system in fingerprint identification |
Publications (2)
Publication Number | Publication Date |
---|---|
CN101145198A true CN101145198A (en) | 2008-03-19 |
CN100470579C CN100470579C (en) | 2009-03-18 |
Family
ID=39207721
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CNB200710122087XA Expired - Fee Related CN100470579C (en) | 2007-09-21 | 2007-09-21 | Certainty coding method and system in fingerprint identification |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN100470579C (en) |
Cited By (17)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102202142A (en) * | 2010-03-26 | 2011-09-28 | 鸿富锦精密工业(深圳)有限公司 | Burr detection device and burr detection method |
CN102254167A (en) * | 2011-08-16 | 2011-11-23 | 杭州晟元芯片技术有限公司 | Method for reducing false reject rate of fingerprint comparison |
CN101593268B (en) * | 2008-05-30 | 2012-03-07 | 深圳市普罗巴克科技股份有限公司 | Method and system for multi-resolution fingerprint identification |
CN101609506B (en) * | 2008-06-20 | 2012-05-23 | 索尼株式会社 | Method and device for recognizing model object in image |
CN102831403A (en) * | 2012-08-10 | 2012-12-19 | 深圳市奔凯安全技术有限公司 | Identification method based on fingerprint feature points |
CN103077377A (en) * | 2012-12-31 | 2013-05-01 | 清华大学 | Direction field distribution-based fingerprint correction method |
CN104077576A (en) * | 2014-07-03 | 2014-10-01 | 南昌欧菲生物识别技术有限公司 | Fingerprint recognition method and terminal device having fingerprint recognition function |
CN104376307A (en) * | 2014-11-24 | 2015-02-25 | 西南交通大学 | Fingerprint image information coding method |
CN104462896A (en) * | 2014-11-19 | 2015-03-25 | 福建亿榕信息技术有限公司 | Reliable video data acquisition and certification method with security certification function |
CN104992154A (en) * | 2015-06-30 | 2015-10-21 | 广东欧珀移动通信有限公司 | Fingerprint identification method and terminal device |
CN105447436A (en) * | 2014-12-19 | 2016-03-30 | 比亚迪股份有限公司 | Fingerprint identification system, fingerprint identification method and electronic device |
CN105868597A (en) * | 2016-05-31 | 2016-08-17 | 广东欧珀移动通信有限公司 | Fingerprint unlocking method and mobile terminal |
CN106200247A (en) * | 2016-09-21 | 2016-12-07 | 成都创慧科达科技有限公司 | A kind of projector possessing fingerprint identification function |
CN106503609A (en) * | 2015-09-08 | 2017-03-15 | 腾讯科技(深圳)有限公司 | The recognition methods of fingerprint ridge point and device |
CN107251044A (en) * | 2015-02-27 | 2017-10-13 | 高通股份有限公司 | Fingerprint verification system |
WO2018213946A1 (en) * | 2017-05-20 | 2018-11-29 | 深圳信炜科技有限公司 | Image recognition method, image recognition device, electronic device, and computer storage medium |
CN110139212A (en) * | 2019-06-21 | 2019-08-16 | Oppo广东移动通信有限公司 | Location processing method and Related product |
-
2007
- 2007-09-21 CN CNB200710122087XA patent/CN100470579C/en not_active Expired - Fee Related
Cited By (25)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101593268B (en) * | 2008-05-30 | 2012-03-07 | 深圳市普罗巴克科技股份有限公司 | Method and system for multi-resolution fingerprint identification |
CN101609506B (en) * | 2008-06-20 | 2012-05-23 | 索尼株式会社 | Method and device for recognizing model object in image |
CN102202142A (en) * | 2010-03-26 | 2011-09-28 | 鸿富锦精密工业(深圳)有限公司 | Burr detection device and burr detection method |
CN102202142B (en) * | 2010-03-26 | 2014-04-30 | 鸿富锦精密工业(深圳)有限公司 | Burr detection device and burr detection method |
CN102254167B (en) * | 2011-08-16 | 2013-06-19 | 杭州晟元芯片技术有限公司 | Method for reducing false reject rate of fingerprint comparison |
CN102254167A (en) * | 2011-08-16 | 2011-11-23 | 杭州晟元芯片技术有限公司 | Method for reducing false reject rate of fingerprint comparison |
CN102831403B (en) * | 2012-08-10 | 2015-09-23 | 深圳市奔凯安全技术股份有限公司 | A kind of recognition methods based on fingerprint feature point |
CN102831403A (en) * | 2012-08-10 | 2012-12-19 | 深圳市奔凯安全技术有限公司 | Identification method based on fingerprint feature points |
CN103077377A (en) * | 2012-12-31 | 2013-05-01 | 清华大学 | Direction field distribution-based fingerprint correction method |
CN103077377B (en) * | 2012-12-31 | 2015-07-29 | 清华大学 | Based on the fingerprint correction method of field of direction distribution |
CN104077576A (en) * | 2014-07-03 | 2014-10-01 | 南昌欧菲生物识别技术有限公司 | Fingerprint recognition method and terminal device having fingerprint recognition function |
CN104462896A (en) * | 2014-11-19 | 2015-03-25 | 福建亿榕信息技术有限公司 | Reliable video data acquisition and certification method with security certification function |
CN104376307B (en) * | 2014-11-24 | 2017-08-04 | 西南交通大学 | A kind of information in fingerprint coding method |
CN104376307A (en) * | 2014-11-24 | 2015-02-25 | 西南交通大学 | Fingerprint image information coding method |
CN105447436A (en) * | 2014-12-19 | 2016-03-30 | 比亚迪股份有限公司 | Fingerprint identification system, fingerprint identification method and electronic device |
WO2016095615A1 (en) * | 2014-12-19 | 2016-06-23 | 比亚迪股份有限公司 | Fingerprint recognition system and fingerprint recognition method and electronic device |
US10366274B2 (en) | 2014-12-19 | 2019-07-30 | Byd Company Limited | Fingerprint identification system, fingerprint identification method, and electronic equipment |
CN107251044A (en) * | 2015-02-27 | 2017-10-13 | 高通股份有限公司 | Fingerprint verification system |
CN104992154A (en) * | 2015-06-30 | 2015-10-21 | 广东欧珀移动通信有限公司 | Fingerprint identification method and terminal device |
CN106503609A (en) * | 2015-09-08 | 2017-03-15 | 腾讯科技(深圳)有限公司 | The recognition methods of fingerprint ridge point and device |
CN106503609B (en) * | 2015-09-08 | 2019-07-16 | 腾讯科技(深圳)有限公司 | The recognition methods of fingerprint ridge point and device |
CN105868597A (en) * | 2016-05-31 | 2016-08-17 | 广东欧珀移动通信有限公司 | Fingerprint unlocking method and mobile terminal |
CN106200247A (en) * | 2016-09-21 | 2016-12-07 | 成都创慧科达科技有限公司 | A kind of projector possessing fingerprint identification function |
WO2018213946A1 (en) * | 2017-05-20 | 2018-11-29 | 深圳信炜科技有限公司 | Image recognition method, image recognition device, electronic device, and computer storage medium |
CN110139212A (en) * | 2019-06-21 | 2019-08-16 | Oppo广东移动通信有限公司 | Location processing method and Related product |
Also Published As
Publication number | Publication date |
---|---|
CN100470579C (en) | 2009-03-18 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN100470579C (en) | Certainty coding method and system in fingerprint identification | |
Peralta et al. | On the use of convolutional neural networks for robust classification of multiple fingerprint captures | |
CN107748877B (en) | Fingerprint image identification method based on minutiae and textural features | |
Bhanu et al. | Fingerprint indexing based on novel features of minutiae triplets | |
Song et al. | Aggregating minutia-centred deep convolutional features for fingerprint indexing | |
CN107403084A (en) | A kind of personal identification method based on gait data | |
CN105512599A (en) | Face identification method and face identification system | |
Almajmaie et al. | Fingerprint recognition system based on modified multi-connect architecture (MMCA) | |
Feng et al. | Fingerprint indexing using ridge invariants | |
Zhao et al. | Fingerprint pre-processing and feature engineering to enhance agricultural products categorization | |
Conti et al. | An advanced technique for user identification using partial fingerprint | |
Chen et al. | A finger vein recognition algorithm based on deep learning | |
Sheng et al. | Consensus fingerprint matching with genetically optimised approach | |
Ge et al. | Deep and discriminative feature learning for fingerprint classification | |
Bhargava et al. | Fingerprint matching using ridge-end and bifurcation points | |
CN105320935A (en) | Multiple-template fingerprint identification method | |
Mandi et al. | Rotation–invariant fingerprint identification system | |
Pokhriyal et al. | MERIT: Minutiae Extraction using Rotation Invariant Thinning | |
Al Taee et al. | A new approach for fingerprint authentication in biometric systems using BRISK algorithm | |
Samai et al. | Oriented Local Binary Pattern (LBP θ): A new scheme for an efficient feature extraction technique | |
Chowdhury et al. | Efficient fingerprint matching based upon minutiae extraction | |
Abdulameer | Developing and valuation for techniques fingerprint recognition system | |
Boutella et al. | Fingerprint identification by wave atoms transform and SVM | |
Chen et al. | A matching algorithm based on local topologic structure | |
Mehrotra et al. | An efficient dual stage approach for iris feature extraction using interest point pairing |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C14 | Grant of patent or utility model | ||
GR01 | Patent grant | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20090318 Termination date: 20140921 |
|
EXPY | Termination of patent right or utility model |