CN107958210A - A kind of palm grain identification method based on decimal system map locating table - Google Patents
A kind of palm grain identification method based on decimal system map locating table Download PDFInfo
- Publication number
- CN107958210A CN107958210A CN201711157787.2A CN201711157787A CN107958210A CN 107958210 A CN107958210 A CN 107958210A CN 201711157787 A CN201711157787 A CN 201711157787A CN 107958210 A CN107958210 A CN 107958210A
- Authority
- CN
- China
- Prior art keywords
- mrow
- array
- row
- msup
- inquiry table
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/12—Fingerprints or palmprints
- G06V40/1347—Preprocessing; Feature extraction
-
- 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
- G06V10/25—Determination of region of interest [ROI] or a volume of interest [VOI]
-
- 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
- G06V40/1365—Matching; Classification
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Human Computer Interaction (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Collating Specific Patterns (AREA)
Abstract
The invention discloses a kind of palm grain identification method based on decimal system map locating table, some original palmprint images are converted into 01 eigenmatrixes first, every trade of going forward side by side vector is integrated either column vector and is integrated then to being mapped to the palm print characteristics array battle array formed in two-value inquiry table corresponding to original palmprint image after 01 array rows after integration or 01 arrays row progress piecemeal and decimal system conversion；Form palm print characteristics array battle array to be identified during identification in the same way to palmprint image to be identified, then utilize and match palm print characteristics array battle array to be identified into row distance with the palm print characteristics array battle array of original palmprint image apart from matching algorithm, decision-making is authenticated according to matching fraction.This method can improve the security of personal recognition and the identification accuracy of personal recognition and calculate recognition efficiency or speed.
Description
Technical field
The invention belongs to palmprint recognition technology field, more particularly to a kind of personal recognition based on decimal system map locating table
Method.
Background technology
With internet in recent years+theory become more and more popular, the security of personal information also becomes more and more important, tradition
Identification authentication mode, such as key, password etc., exists and the danger such as is easily lost, is easily stolen, being easily cracked, in order to improve peace
Quan Xing, biological identification technology have been stepped among the life of the mankind.It is compared to traditional recognition method and is held using individual is relied on
The mark for having either memory carry out identification biological identification technology using the intrinsic physiological characteristic of the mankind or behavioural characteristic into
Row identification, has stability strong, distinctiveness is high, gathers the advantages that easy.
Personal recognition is a kind of newer biometrics identification technology proposed in recent years.Contain abundant master in palmmprint
Line, wrinkle, tiny texture, ridge tip, bifurcation etc. may serve to carry out identification, it is clear that personal recognition is as identity
Identification and authentication mode has certain convenience and reliability.But palmprint image is scarce resource, a people only has two
Palm, the palm of most people have symmetry, and with advancing age, the basic texture structure of palmmprint simultaneously no longer occurs
Change.In view of the generality of authentication scene now, the palmmprint biological characteristic of same user's same characteristic features is likely to preserve
Be shared in multiple databases.If palmmprint biological characteristic of the user in a database is successfully stolen, then this user
The palmmprint biological characteristic being stored in other databases is all no longer safe, and the individual privacy contained in palmmprint is probably let out
Dew, it also is adapted for being used for other Verification Systems again.Therefore the personal recognition authentication techniques of research safety are very important.
The content of the invention
The technical problem to be solved in the present invention is, there is provided a kind of palm grain identification method based on decimal system map locating table,
Improve the security of personal recognition and the identification accuracy of personal recognition and calculate recognition efficiency or speed.
In order to solve the above technical problems, the technical solution adopted in the present invention is：
A kind of palm grain identification method based on decimal system map locating table, comprises the following steps：
S1：Some original palmprint images are obtained, the ROI image of each original palmprint image is extracted, the ROI image is turned
Change 01 eigenmatrixes into；
S2：For each 01 eigenmatrix, its all row vector is integrated into 01 arrays in a row or is owned
Column vector is integrated into 01 array of a row；
S3：The step S2 each 01 array rows produced or 01 arrays row such as are subjected at piecemeal, the code of each piecemeal
Word length is ω, and each piecemeal array is converted into decimal number；
S4：If being 01 array row piecemeals in step S3, a two-value inquiry table is randomly generated according to piecemeal size, is looked into
The line number for asking table is equal to 2ω, columns ω, is denoted as the first inquiry table；If 01 array row piecemeals, then according to piecemeal size with
Machine produces a two-value inquiry table, and the line number of inquiry table is ω, and columns is 2 ω, is denoted as the second inquiry table；First inquiry table
Line number and the second inquiry table columns since zero；
S5：D (d≤ω) row are randomly selected in the first inquiry table, for any block count obtained by 01 array rows
Group, corresponds to decimal number according to it and corresponding line is chosen in the first inquiry table, and is chosen in the row and belong to d that foregoing d is arranged
Bit number；
Alternatively,
D (d≤ω) is randomly selected in the second inquiry table OK, any block count arranged for same 01 array
Group, corresponds to decimal number according to it and chooses respective column in the first inquiry table, and chooses in the row and to belong to d of foregoing d rows
Bit number；
As all bit arrays selected by all piecemeals that same 01 array row or 01 arrays arrange into being used as this
The palm print characteristics array battle array of 01 array rows or the original palmprint image in 01 array row sources；
S6：For palmprint image to be identified, its palm print characteristics array battle array is obtained by step S1 to S5；
S7：The palm print characteristics array battle array of palmprint image to be identified and the palm print characteristics array battle array of original palmprint image are carried out
Apart from match cognization.
In the above-mentioned technical solutions, inquiry table, which randomly generates characteristic, can ensure that it is applied in different identification scenes,
As different piecemeal sizes and different poll bits obtained from different selection modes, it is special that different palmmprints can be produced
Levy array battle array.The palm print characteristics array battle array that this method produces disclosure satisfy that irreversibility, and diversity, defeasibility, are conducive to carry
Security, accuracy and the recognition speed of height identification.
Certification is identified using palm print characteristics array battle array in this method, no longer directly uses palmmprint main line, wrinkle, tiny
The palmmprint biological characteristic such as texture, ridge tip, bifurcation.Even if palm print characteristics array battle array is stolen, its irreversibility ensure that, nothing
Method reduces original palmprint image by palm print characteristics array battle array.Its diversity ensure that the palm print characteristics array battle array that can be generated
Quantity it is enough, even same original palmmprint is applied to multiple identification scenes, can also ensure different identification scenes
Different either random palm print characteristics array battle arrays can be assigned to, relevance is not produced between different identification scenes.I.e.
The palm print characteristics array battle array in multiple identification scenes or database is stolen, original can not be reduced by way of cross-matched
Beginning palmmprint.Even if its defeasibility ensure that palm print characteristics, array battle array is stolen, can be big by varying inquiry table or change piecemeal
The modes such as small or change poll bits produce new safe palm print characteristics array battle array and are used for personal recognition certification.
The matching of this method applications distances carries out personal recognition, and distance matching is suitable for identification in real time and matches, and identification is accurate
Degree is high, and recognition speed is fast, so as to ensure that the recognition performance of this method.
As an improvement, in above-mentioned step S1, following Gabor filter is constructed first：
Wherein, x'=(x-x0)cosθ+(y-y0) sin θ, y'=- (x-x0)sinθ+(y-y0) cos θ, (x0,y0) represent filter
Ripple device central point, ω represent radial frequency, and θ represents filter angles,δ represents the half range of frequency response
Bandwidth；
The rules of competition is defined as：
arg minj(I(x,y)*ψR(x,y,w,θj))
Wherein, I is ROI image, ψRIt is the real part of Gabor filter, θjIt is filter angles, j={ 0,1,2,3,4,5 }
Represent what wave filter was chosenSix angles.
Then the contention code eigenmatrix of generation is changed into corresponding binary feature matrix.
Gabor wavelet and the visual stimulus response of simple cell in human visual system are closely similar.It is in extraction target
Local space and frequency-domain information in terms of there is good characteristic.Edge sensitive of the Gabor wavelet for image, using the teaching of the invention it is possible to provide
Good set direction and scale selection characteristic, and it is insensitive for illumination variation, using the teaching of the invention it is possible to provide it is good to illumination variation
Adaptability.
As an improvement, the distance is Hamming distance, represent as follows：
Wherein H represents final Hamming distance, and m, n represent the line number and columns of palm print characteristics array battle array, and F represents that palmmprint is special
Levy array battle array.Complexity when match cognization calculates can be effectively reduced with Hamming distance matching.
To sum up, the present invention can either improve the security of personal recognition, and can improve the identification accuracy of personal recognition
And calculate recognition efficiency or speed.
Brief description of the drawings
Fig. 1 is that decimal system map locating represents to be intended in the specific embodiment of the invention.
Fig. 2 is the ROC curve figure of emulation experiment in the specific embodiment of the invention.
Embodiment
The palm grain identification method based on decimal system map locating table, comprises the following steps：
S1：Some original palmprint images, these original palmmprints are obtained using suitable method (such as take pictures, the mode such as scan)
Image comes from several body.Under different scenes, selected several body may be different, and this several body is conduct
Follow-up identification reference.
Extract the ROI image of each original palmprint image., can be in the following way when carrying out ROI image extraction：
For any original palmprint image, which is subjected to binary conversion treatment and obtains bianry image, two
Palm external periphery outline is extracted in value image and detects the angle point (finger formed between forefinger, middle finger between nameless, little finger of toe
Recess between finger is angle point), using the line of two angle points as the longitudinal axis, hang down from the midpoint of two angle points to the longitudinal axis
Line, using vertical line as transverse axis, using the intersection point of the longitudinal axis and transverse axis as coordinate origin, coordinate origin forms new seat with the longitudinal axis, transverse axis
Mark system, under new coordinate system, size is intercepted on original palmprint image and is schemed for the palmmprint region of 128 × 128 pixels as ROI
Picture.
Each ROI image is converted into 01 eigenmatrixes.
Specifically, for each ROI image, in the following manner can be taken to convert thereof into 01 eigenmatrixes：
(1) following Gabor filter is constructed first：
Wherein, x'=(x-x0)cosθ+(y-y0) sin θ, y'=- (x-x0)sinθ+(y-y0) cos θ, (x0,y0) represent filter
Ripple device central point, ω represent radial frequency, and θ represents filter angles,δ represents the half range of frequency response
Bandwidth；
The rules of competition is defined as：
arg minj(I(x,y)*ψR(x,y,w,θj))
Wherein, I is pretreated image, ψRIt is the real part of Gabor filter, θjIt is filter angles, j=0,1,2,
3,4,5 } represent what wave filter was chosenSix angles；
Then the contention code eigenmatrix of generation is changed into corresponding binary feature matrix, i.e. 01 eigenmatrixes.
Numerical value inside contention code eigenmatrix is all the value of 1-6, converts it into the binary numeral of 3 and forms two-value
Feature, such as 3 change into 011,6 and change into 110.
(2) accidental projection mode can also be used.An accidental projection matrix is produced, by accidental projection matrix premultiplication ROI
Image, and be normalized；Then an accidental projection matrix is regenerated, and is normalized；Then by first random throwing
The normalization result of shadow matrix premultiplication ROI image size compared with the normalization result of second accidental projection matrix is by value, will
The value that the former is more than on the position of the latter is set to zero, and the value on remaining position is set to 1, or the former is more than on the position of the latter
Value be set to 1, the value on remaining position is set to 0, so as to obtain 01 encoder matrixs.
By step S1, each original palmprint image (or ROI image) is corresponding to produce 01 eigenmatrixes, and institute
The mode used is identical, otherwise all or all using accidental projection mode, or all made using wave filter, contention code mode
With other suitable methods.
S2：For each 01 eigenmatrix, its all row vector is integrated into 01 arrays in a row or is owned
Column vector is integrated into 01 array of a row, otherwise so that a 01 array row (i.e. a line are obtained by each 01 eigenmatrix
01 arrays), otherwise a 01 arrays row (i.e. 01 array of a row) is obtained, and all 01 eigenmatrixes uniformly obtain zero
One array row or 01 arrays row.
S3：The step S2 each 01 array rows produced or 01 arrays row such as are subjected at piecemeal, the code of each piecemeal
Word length is ω, and each piecemeal array is converted into decimal number；Such as ω=3, if the code character after some piecemeal
(it may be the code character after row piecemeal or the code character after row piecemeal, be collectively expressed as 10 herein 1), then to its turn for 101
It is melted into decimal number 5.
S4：If being 01 array row piecemeals in step S3, a two-value inquiry table is randomly generated according to piecemeal size, is looked into
The line number for asking table is equal to 2ω, columns ω, is denoted as the first inquiry table；If 01 array row piecemeals, then according to piecemeal size with
Machine produces a two-value inquiry table, and the line number of inquiry table is ω, and columns is 2 ω, is denoted as the second inquiry table；The row of first inquiry table
Number and the columns of the second inquiry table are since zero.Such as ω=4, produce 16 rows, 4 row the first inquiry tables or
Second inquiry tables of 4 rows, 16 row, the number in the first inquiry table and the second inquiry table is 0 or 1, and the line number of the first inquiry table is 0
Columns to 15, second inquiry tables is 0 to 15.
S5：If being 01 array row piecemeals in step S3, d (d≤ω) row are randomly selected in the first inquiry table, for
Any piecemeal array (i.e. each small piecemeal) obtained by 01 array rows, decimal number is corresponded in the first inquiry table according to it
Corresponding line is chosen, and the d bit number (i.e. 0,1 number) for belonging to foregoing d row is chosen in the row.
If for example, ω=4, and randomly selecting to the 2nd row and the 3rd and arranging in the first inquiry table, some piecemeal array pair
The decimal number answered is 5, then the 5th row and the 2nd, 3 row of the first inquiry table is navigated to, so as to obtain two bit numbers.
By aforesaid way, each piecemeal array after 01 array row piecemeals obtains d bit number, is denoted as bit number
Block.
If being 01 array row piecemeals in step S3, d (1≤d≤ω) is randomly selected in the second inquiry table OK, for
Any piecemeal array (i.e. each small piecemeal) arranged by 01 arrays, decimal number is corresponded in the first inquiry table according to it
Respective column is chosen, and the d bit number (i.e. 0,1 number) for belonging to foregoing d rows is chosen in the row.
If for example, ω=4, and randomly selected in the second inquiry table to the 2nd row and the 3rd row, some piecemeal array pair
The decimal number answered is 5, then the 5th row and the 2nd, 3 rows of the first inquiry table is navigated to, so as to obtain two bit numbers.
Equally, by aforesaid way, each piecemeal array after 01 array row piecemeals obtains d bit number, is denoted as ratio
Special several piece.
For any one 01 array row, by as selected by its all piecemeal array to bit several piece looked into by first
Place line number in inquiry table is arranged, and obtains 01 bit array of a line, or perhaps the matrix that a line number is 1.By one
This obtained 01 bit array of a line of a 01 array row also just becomes the original palmmprint as the 01 array row source
The palm print characteristics array battle array of image.
Likewise, for any one 01 array row for, by as selected by its all piecemeal array to bit several piece
Arranged by the place columns in the second inquiry table, obtain 01 bit array of a row, or perhaps the square that a columns is 1
Battle array.This obtained 01 bit array of row is arranged also just as the original as the 01 array row source by 01 arrays
The palm print characteristics array battle array of beginning palmprint image.
It is further described below by an example.
If as shown in Figure 1, generate 01 eigenmatrixes by a certain original palmprint image I, this 01 feature
All row vectors of matrix integrate 01 array in a row.Piecemeal, the code word length of each piecemeal are carried out to this 01 array of a line
Spend for 4, obtain 6 piecemeals, the from the 1st to the 6th block count group corresponds to decimal number 7,2,10,13,9,5 respectively.Randomly generate
One line number is 16, and columns is 4 the first inquiry table, and the line number of the first inquiry table is 0 to 15, and the number in the first inquiry table is 0
Or 1, and the 2nd row and the 3rd row of selected first inquiry table.
Position respectively the first inquiry table the 2nd, 5,7,9,10, the bit number intersected with the 2nd row and the 3rd row in 13 rows,
6 bit several pieces " 01 " " 10 " " 10 " " 10 " " 10 " " 01 " are obtained, this 6 bit several pieces are carried out according to place line number
Arrangement, obtains a line bit array " 011010101001 ", this line bit array becomes original image I
Palm print characteristics array battle array.
A row bit array can be similarly obtained in the case of 01 eigenmatrixes are integrated into 01 array of row, this
Row bit array can also become the palm print characteristics array battle array of original image I.
It can be selected as needed for the selected columns of the first inquiry table and the line number of the second inquiry table, such as on
State and 2 row have been selected in example, it is also an option that 1 arranges either 3 row or 4 row, the bit number number in the block of the bit number obtained from
It can change, other are identical with above-mentioned example.
S6：For palmprint image to be identified, also pass through step S1 and obtain 01 eigenmatrixes, and palmprint image to be identified
The mode for obtaining 01 eigenmatrixes is identical with each then palmprint image.
If 01 eigenmatrixes of original palmprint image have all carried out row vector integration, then for palmprint image to be identified
01 eigenmatrixes also carry out row vector integration, and the palm print characteristics number of palmprint image to be identified is obtained by step S2 to S5
Group, wherein palmprint image to be identified uses first inquiry table identical with original palmprint image and selected same column.
S7：By the palm print characteristics array battle array of the palmprint image to be identified palm print characteristics number with each original palmprint image respectively
Group battle array carries out Hamming distance match cognization and classification, and each individual is as a kind of in wherein step S1.
Represent as follows：
Wherein H represents final Hamming distance, and m, n represent the line number and columns of palm print characteristics array battle array, and F represents that palmmprint is special
Levy array battle array.
When with Hamming distance matching, we set a threshold value, if palm print characteristics array battle array to be identified is to each
The distance of the palm print characteristics array battle array of obtained original palmprint image is rejected up to or over the threshold value, then identification,
That is identification certification does not pass through.If some or all of distance is less than the threshold value, judge to wait to know on the basis of being identified by
Other person belongs to the minimum one kind of distance.If thus if we have learned that the body of these original palmprint image source individuals
Part, then when classification is identified, it can not only determine to be identified individual identify whether by that can also be identified by
On the basis of determine individual identity (which kind of belongs to) to be identified.
If building identifying system using this method, then only palm print characteristics array need to be stored in the database of identifying system
Battle array, without storing palmprint image.Even if palm print characteristics array battle array is stolen, its irreversibility ensure that, can not be special by palmmprint
Levy array battle array and reduce original palmprint image.Its diversity ensure that the quantity for the palm print characteristics array battle array that can be generated is enough
It is more, even same original palmmprint is applied to multiple identification scenes, it can also ensure that different identification scenes can be assigned to
Different either random palm print characteristics array battle arrays, do not produce relevance between different identification scenes.Even if multiple identifications
Palm print characteristics array battle array in scene or database is stolen, and original palmmprint can not be reduced by way of cross-matched.Its
Even if defeasibility ensure that palm print characteristics, array battle array is stolen, and can either change piecemeal size or change by varying inquiry table
The modes such as poll bits produce new safe palm print characteristics array battle array and are used for personal recognition certification.
This method has also carried out emulation experiment, and emulation experiment is carried out using palm print database disclosed in The Hong Kong Polytechnic University
's.In the database altogether comprising it is 600 big it is small be 384 × 284 palmprint image, pick up from 100 people, everyone 6.Everyone
6 palmprint images to pick up from two different periods, time interval be two months.
In an experiment, 01 eigenmatrix sizes of generation are 32 × 64, and every trade of going forward side by side vector is integrated, then in piecemeal
ω=8, the d=4 of selection, the inquiry table size randomly generated are 28OK, 8 row, and two row are randomly selected, the palmmprint ultimately generated
The size of feature array battle array is a line, 512 bits.The image of the same palm of same person carries out matching as in class in an experiment
Matching, carries out matching and is matched between class between not employment or between the different palms of same people.Experiment carries out altogether
179700 matchings, wherein 1500 times are matching in class, match for 178200 times between class, the false acceptance rate (FAR) of experiment and
The curve map (ROC curve) of real receptance (GAR) is as shown in Figure 2.As can be seen from the figure ROC curve is straight line, is said
Bright discrimination is 100%, has obtained good recognition effect.
Claims (3)
1. a kind of palm grain identification method based on decimal system map locating table, comprises the following steps：
S1：Some original palmprint images are obtained, the ROI image of each original palmprint image is extracted, the ROI image is converted into
01 eigenmatrixes；
S2：For each 01 eigenmatrix, by its all row vector integrate 01 arrays in a row or by its it is all arrange to
Amount is integrated into 01 array of a row；
S3：The step S2 each 01 array rows produced or 01 arrays row are carried out waiting piecemeal, the code word of each piecemeal is grown
Spend for ω, and each piecemeal array is converted into decimal number；
S4：If being 01 array row piecemeals in step S3, a two-value inquiry table, inquiry table are randomly generated according to piecemeal size
Line number be equal to 2ω, columns ω, is denoted as the first inquiry table；If 01 array row piecemeals, then produced at random according to piecemeal size
A raw two-value inquiry table, the line number of inquiry table are ω, columns 2ω, it is denoted as the second inquiry table；The row of first inquiry table
Number and the columns of the second inquiry table are since zero；
S5：D (d≤ω) row are randomly selected in the first inquiry table, for any piecemeal array obtained by 01 array rows, root
Decimal number is corresponded to according to it corresponding line is chosen in the first inquiry table, and d bit for belonging to foregoing d row is chosen in the row
Number；
Alternatively,
D (d≤ω) is randomly selected in the second inquiry table OK, any piecemeal array arranged for same 01 array, root
Decimal number is corresponded to according to it respective column is chosen in the first inquiry table, and d bit for belonging to foregoing d rows is chosen in the row
Number；
All bit arrays selected by all piecemeals arranged as same 01 array row or 01 arrays into as this 01
The palm print characteristics array battle array of array row or the original palmprint image in 01 array row sources；
S6：For palmprint image to be identified, its palm print characteristics array battle array is obtained by step S1 to S5；
S7：By the palm print characteristics array battle array of the palm print characteristics array battle array of palmprint image to be identified and original palmprint image into row distance
Match cognization.
2. the palm grain identification method according to claim 1 based on decimal system map locating table, it is characterised in that：In step
In S1, following Gabor filter is constructed first：
<mrow>
<mi>&psi;</mi>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>,</mo>
<mi>&omega;</mi>
<mo>,</mo>
<mi>&theta;</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mfrac>
<mi>&omega;</mi>
<mrow>
<msqrt>
<mrow>
<mn>2</mn>
<mi>&pi;</mi>
</mrow>
</msqrt>
<mi>K</mi>
</mrow>
</mfrac>
<msup>
<mi>e</mi>
<mrow>
<mo>-</mo>
<mfrac>
<msup>
<mi>&omega;</mi>
<mn>2</mn>
</msup>
<mrow>
<mn>8</mn>
<msup>
<mi>K</mi>
<mn>2</mn>
</msup>
</mrow>
</mfrac>
<mrow>
<mo>(</mo>
<mn>4</mn>
<msup>
<mi>x</mi>
<mrow>
<mo>&prime;</mo>
<mn>2</mn>
</mrow>
</msup>
<mo>+</mo>
<msup>
<mi>y</mi>
<mrow>
<mo>&prime;</mo>
<mn>2</mn>
</mrow>
</msup>
<mo>)</mo>
</mrow>
</mrow>
</msup>
<mrow>
<mo>(</mo>
<msup>
<mi>e</mi>
<mrow>
<msup>
<mi>i&omega;x</mi>
<mo>&prime;</mo>
</msup>
</mrow>
</msup>
<mo>-</mo>
<msup>
<mi>e</mi>
<mrow>
<mo>-</mo>
<mfrac>
<msup>
<mi>K</mi>
<mn>2</mn>
</msup>
<mn>2</mn>
</mfrac>
</mrow>
</msup>
<mo>)</mo>
</mrow>
</mrow>
Wherein, x'=(x-x0)cosθ+(y-y0) sin θ, y'=- (x-x0)sinθ+(y-y0) cos θ, (x0,y0) represent wave filter
Central point, ω represent radial frequency, and θ represents filter angles,δ represents half web of frequency response
It is wide；
The rules of competition is defined as：
arg minj(I(x,y)*ψR(x,y,w,θj))
Wherein, I is ROI image, ψRIt is the real part of Gabor filter, θjIt is filter angles, j={ 0,1,2,3,4, } 5 is represented
What wave filter was chosenSix angles.
3. the palm grain identification method according to claim 1 or 2 based on decimal system map locating table, it is characterised in that：Institute
It is Hamming distance to state distance, is represented as follows：
<mrow>
<mi>H</mi>
<mo>=</mo>
<mfrac>
<mrow>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>i</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>m</mi>
</munderover>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>j</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>n</mi>
</munderover>
<msubsup>
<mi>F</mi>
<mrow>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
</mrow>
<mrow>
<mo>&prime;</mo>
<mn>1</mn>
</mrow>
</msubsup>
<mo>&CirclePlus;</mo>
<msubsup>
<mi>F</mi>
<mrow>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
</mrow>
<mrow>
<mo>&prime;</mo>
<mn>2</mn>
</mrow>
</msubsup>
</mrow>
<mrow>
<mi>m</mi>
<mo>&times;</mo>
<mi>n</mi>
</mrow>
</mfrac>
</mrow>
Wherein H represents final Hamming distance, and m, n represent the line number and columns of palm print characteristics array battle array, and F represents palm print characteristics number
Group battle array.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711157787.2A CN107958210A (en) | 2017-11-20 | 2017-11-20 | A kind of palm grain identification method based on decimal system map locating table |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711157787.2A CN107958210A (en) | 2017-11-20 | 2017-11-20 | A kind of palm grain identification method based on decimal system map locating table |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107958210A true CN107958210A (en) | 2018-04-24 |
Family
ID=61965052
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711157787.2A Pending CN107958210A (en) | 2017-11-20 | 2017-11-20 | A kind of palm grain identification method based on decimal system map locating table |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107958210A (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104268533A (en) * | 2014-10-10 | 2015-01-07 | 南昌航空大学 | Non-contact revocable palm print bimodal authentication method |
| CN106446867A (en) * | 2016-10-13 | 2017-02-22 | 济南大学 | Double-factor palmprint identification method based on random projection encryption |
-
2017
- 2017-11-20 CN CN201711157787.2A patent/CN107958210A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104268533A (en) * | 2014-10-10 | 2015-01-07 | 南昌航空大学 | Non-contact revocable palm print bimodal authentication method |
| CN106446867A (en) * | 2016-10-13 | 2017-02-22 | 济南大学 | Double-factor palmprint identification method based on random projection encryption |
Non-Patent Citations (5)
| Title |
|---|
| ADAMS WAI-KIN KONG等: "《Competitive coding scheme for palmprint verification》", 《PROCEEDINGS OF THE 17TH INTERNATIONAL CONFERENCE ON PATTERN RECOGNITION (ICPR’04)》 * |
| RUDRESH DWIVEDI等: "《A privacy-preserving cancelable iris template generation scheme using decimal encoding and A privacy-preserving cancelable iris template generation scheme using decimal encoding and look-up table mapping》", 《COMPUTERS AND SECURITY》 * |
| 杨懿竣: "《指纹安全认证的不可逆变换技术研究》", 《中国博士学位论文全文数据库 信息科技辑》 * |
| 郭秀梅: "《掌纹识别算法的研究》", 《中国博士学位论文全文数据库 信息科技辑》 * |
| 陈晓红: "《数据降维的广义相关分析研究》", 《中国博士学位论文全文数据库 信息科技辑》 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Lin et al. | A CNN-based framework for comparison of contactless to contact-based fingerprints | |
| Moujahdi et al. | Fingerprint shell: Secure representation of fingerprint template | |
| Yang et al. | A fingerprint recognition scheme based on assembling invariant moments for cloud computing communications | |
| CN106446867B (en) | A kind of double factor palm grain identification method based on accidental projection encryption | |
| Alam et al. | An alignment-free cancelable fingerprint template for bio-cryptosystems | |
| CN105528572A (en) | Fingerprint identification method | |
| Zhang et al. | Graph fusion for finger multimodal biometrics | |
| US11138409B1 (en) | Biometric recognition and security system | |
| Ali et al. | Polynomial Vault: A secure and robust fingerprint based authentication | |
| Barman et al. | An efficient fingerprint matching approach based on minutiae to minutiae distance using indexing with effectively lower time complexity | |
| CN112686191B (en) | Living body anti-counterfeiting method, system, terminal and medium based on three-dimensional information of human face | |
| Attia et al. | Deep learning-driven palmprint and finger knuckle pattern-based multimodal Person recognition system | |
| CN107958211A (en) | A kind of palm grain identification method based on matrix conversion | |
| Lakshmanan | Efficient person authentication based on multi‐level fusion of ear scores | |
| Ali et al. | Fingerprint shell with impregnable features | |
| CN108960003A (en) | Based on Gabor and the palm print characteristics of chaotic maps generate and authentication method | |
| Mallikarjuna Reddy et al. | Generating cancelable fingerprint template using triangular structures | |
| Wu et al. | Fingerprint bio‐key generation based on a deep neural network | |
| Liu et al. | Random local region descriptor (RLRD): A new method for fixed-length feature representation of fingerprint image and its application to template protection | |
| Selwal et al. | A scheme for template security at feature fusion level in multimodal biometric system | |
| de Diego et al. | Face verification with a kernel fusion method | |
| Farmanbar et al. | A hybrid approach for person identification using palmprint and face biometrics | |
| Dillak et al. | A novel approach for iris recognition | |
| Ganesan et al. | Multimodal hybrid features in 3D ear recognition | |
| Wang et al. | Human ear image recognition method using PCA and Fisherface complementary double feature extraction |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180424 |