CN101887517A - Immune cloned finger venous image characteristic extraction method based on linear weighted function - Google Patents
Immune cloned finger venous image characteristic extraction method based on linear weighted function Download PDFInfo
- Publication number
- CN101887517A CN101887517A CN 201010200518 CN201010200518A CN101887517A CN 101887517 A CN101887517 A CN 101887517A CN 201010200518 CN201010200518 CN 201010200518 CN 201010200518 A CN201010200518 A CN 201010200518A CN 101887517 A CN101887517 A CN 101887517A
- Authority
- CN
- China
- Prior art keywords
- antibody
- value
- function
- affinity
- img
- 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
- 238000000605 extraction Methods 0.000 title claims abstract description 21
- 238000000034 method Methods 0.000 claims abstract description 43
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 32
- 210000003462 vein Anatomy 0.000 claims description 40
- 230000003044 adaptive effect Effects 0.000 claims description 14
- 230000008859 change Effects 0.000 claims description 11
- 238000013377 clone selection method Methods 0.000 claims description 9
- 230000035772 mutation Effects 0.000 claims description 5
- 238000004364 calculation method Methods 0.000 claims description 4
- 238000004519 manufacturing process Methods 0.000 claims description 3
- 101100269328 Caenorhabditis elegans aff-1 gene Proteins 0.000 claims description 2
- 238000013016 damping Methods 0.000 claims description 2
- 238000009499 grossing Methods 0.000 claims description 2
- 230000002093 peripheral effect Effects 0.000 claims description 2
- 108090000623 proteins and genes Proteins 0.000 claims description 2
- 230000012010 growth Effects 0.000 abstract description 4
- 230000000694 effects Effects 0.000 abstract description 3
- 238000004088 simulation Methods 0.000 abstract description 3
- 230000002708 enhancing effect Effects 0.000 abstract 1
- 230000006870 function Effects 0.000 description 39
- 238000010586 diagram Methods 0.000 description 9
- 230000008569 process Effects 0.000 description 9
- 239000000284 extract Substances 0.000 description 6
- 230000007246 mechanism Effects 0.000 description 6
- 238000005516 engineering process Methods 0.000 description 4
- 238000011160 research Methods 0.000 description 3
- 101000633607 Bos taurus Thrombospondin-2 Proteins 0.000 description 2
- 210000000987 immune system Anatomy 0.000 description 2
- 238000002386 leaching Methods 0.000 description 2
- 230000006386 memory function Effects 0.000 description 2
- 230000000452 restraining effect Effects 0.000 description 2
- 210000004204 blood vessel Anatomy 0.000 description 1
- 230000009668 clonal growth Effects 0.000 description 1
- 238000010835 comparative analysis Methods 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 230000028993 immune response Effects 0.000 description 1
- 238000003018 immunoassay Methods 0.000 description 1
- 230000016784 immunoglobulin production Effects 0.000 description 1
- 230000005764 inhibitory process Effects 0.000 description 1
- 238000003909 pattern recognition Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000012797 qualification Methods 0.000 description 1
- 230000008521 reorganization Effects 0.000 description 1
- 238000005728 strengthening Methods 0.000 description 1
- 230000004083 survival effect Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Images
Landscapes
- Peptides Or Proteins (AREA)
Abstract
The invention discloses an immune cloned finger venous image characteristic extraction method based on a linear weighted function (LWF), aiming at the problems that noise is easily misjudged into venous information and the worse venous information is misjudged into the noise for characteristic extraction in a low-quality venous image. In an algorithm of the method, an initial antibody is generated by adopting a self-adaptive threshold method, a weight value is obtained by adopting a curve fitting mode, and balancing denoising and enhancing boundaries are achieved by carrying out linear weighting on a venous region. An affinity function and an antibody concentration function are constructed, and the growth of the venous information is promoted and the interference of the noise is inhibited according to the affinity function and an antibody concentration value. Simulation results indicate that compared with other algorithms, the algorithm has the characteristics of strong authenticity, accuracy, continuity, more abundant detail characteristics and the like and particularly has strong noise resisting capability and good extraction effect for a low-quality venous image.
Description
Technical field
The present invention relates to the Feature Extraction Technology of finger vena picture.
Background technology
Current society, biometrics identification technology more and more is subjected to people's attention.The diversity of biological characteristic has caused the diversity of recognition technology, and the vein pattern recognition technology is also arisen at the historic moment as one of them.2000, researchists such as M.Kono under the subsidy of HIT, developed the finger vena near infrared recognition system that is used for personnel identity identification, and have been applied to personal identification.For different living things feature recognitions, need take the feature extraction algorithm that adapts.Finger vein features is no exception, because different people's veinprints varies in size, the feature of tiny vein blood vessel is not easy to extract; Adding finger vena picture that present harvester obtains, to have a target and background discrimination not obvious, and noise and venous information coexistence want therefrom to obtain accurately quite difficulty of venous information.And inferior quality vein picture feature extracts and easily noise to be misjudged into venous information, and relatively poor venous information is misjudged into noise.In addition, owing to be subjected to environment and Temperature Influence, the finger vena blood vessel expands and shrinks and also makes troubles to feature extraction, is the much-talked-about topic of current domestic and international research.Along with the research of vein pattern extraction algorithm deepens continuously and develops, various new feature extraction algorithms also arise at the historic moment.
At present, have studying aspect the vein paddy shape feature extraction algorithm: direction extraction algorithm, direction template operator, repeated line tracking, local threshold method, adaptive threshold method and the improved linear method that combines of following the tracks of.Analyze said method as can be known,, when adopting the method in zone to extract feature, judge whether current point is vein pattern information, need to adopt point in the neighborhood to describe or represent for the unconspicuous vein picture of target and background discrimination.Immune clone algorithm has reacted a kind of like this mechanism just.Come the limit search scope with initial antibodies, affinity function and antibody concentration retrain vein adjacent domain feature, use the memory function of Immune Clone Selection and carry out the extraction of vein pattern.
Nineteen fifty-nine, famous immunologist Burnet proposed clonal selective theory.Immune Clone Selection is to have used for reference the occurring in nature vegetative propagation.It is the pattern of immune a kind of natural selection.In immune system, the process that is produced antibody by immune response is an immune learning process.And immune system has embodied biological characteristics such as study, memory, antibody diversity in this process.
Summary of the invention
The present invention is directed to the deficiency that existing method exists, the present invention proposes a kind of finger vein features extracting method based on linear weighted function (LWF) immune clone.This method adopts the adaptive threshold method to produce initial antibodies, adopt the mode of curve fitting to obtain weights, and by venosomes being carried out the balanced denoising of linear weighted function and strengthening the border, structure affinity function and antibody concentration function, rely on affinity function and antibody concentration value, promote the growth of venous information and suppress interference of noise.
The present invention is achieved through the following technical solutions:
A kind of based on the linear weighted function immune cloned finger venous image characteristic extraction method, it at first produces initial antibodies to the former figure of finger vena picture with the adaptive threshold method, in order to determine search volume and scope, first generation antibody (initial antibodies) A that obtains
1(i0, coordinate figure j0) return among the former figure, and (i0 j0) carries out the affinity value and calculates, and by given affinity value decision method, produces new antibodies at the coordinate points img of correspondence; Secondly, judge whether to have in the first generation antibody and do not satisfy affinity and antibody concentration; If exist, then carry out this antibody point and remove operation, draw second generation antibody A by clone operations
2(i0, j0); Then, on the basis of second generation antibody, make a variation, affinity, antibody concentration calculate, and draws antibody A of new generation by clone operations
b(i0, j0); At last, obtain optimum antibody.
In the described in front calculating process, comprised the computing of generation, variation the choosing of mode, linear weighted function function (LWF) ranking operation, affinity function and the antibody concentration function of initial antibodies again.
Initial antibodies all has certain influence to quality and the algorithm the convergence speed of finally separating, and in order to help improving speed of convergence and antibody quality, the present invention utilizes priori to obtain quality initial antibodies preferably, can effectively provide tutorial message for search.And adopt the adaptive threshold method to obtain initial antibodies.
Choosing of variation mode, the change point scope definition is around antibody point.
The present invention adopts linear weighted function function (LWF) for vein picture to be processed, and venosomes is carried out the balanced denoising of linear weighted function and strengthens the border.Determining of the coefficient of linear weighted function function adopted the mode of curve fitting.
The calculating of affinity function and antibody concentration function, combine the relevant knowledge in the immune clone algorithm, by affinity function and antibody concentration value, promote the growth of venous information and suppress interference of noise, be the reservation of vein pattern information and the inhibition and the removal of interference noise.
Binding immunoassay clone mechanism of the present invention, this method adopts improved adaptive threshold method to produce initial antibodies, and by venosomes latent structure affinity function, in the neighborhood of candidate solution, produce variation and separate, the mechanism in Immune Clone Selection is issued to optimum antibody at last.Optimum antibody IMU is crossed the immune clone computing and is obtained, and has the characteristics of the nature survival of the fittest, carries out Immune Clone Selection for vein pattern antibody, eliminates selection for noise antibody.
The immunogene operation comprises and intersects and variation, and immunology thinks that affinity generation ripe and antibody diversity mainly relies on the high frequency variation of antibody, but not intersects or reorganization.Therefore, the present invention has adopted mutation operation.The Immune Clone Selection operation has been compressed the size of population effectively by according to qualifications local.Because the choice mechanism of clone's operator itself has memory function, can guarantee that the algorithm convergence with probability 1 is to optimum solution.
This method is different from traditional algorithm, it has used for reference the agamic immune clone principle in the organic sphere, by the variation of adjacent domain point, the calculating of affinity function and antibody concentration value has effectively suppressed noise information and has promoted the clonal growth of venous information to breed.Simulation result shows: the vein pattern information that this algorithm extracts is better than other algorithm, has effectively distinguished vein and noise, and the vein pattern details of extraction is abundanter, has very strong adaptivity, anti-interference, practicality.
Description of drawings
Fig. 1 algorithm concrete steps process flow diagram
Fig. 2 is initial antibodies and Xin Dai antibody production process block diagram
Fig. 3 is the analogous diagram before and after the match
The death of Fig. 4 noise antibody
The clone of Fig. 5 vein pattern antibody
Fig. 6 is an immune clone algorithm leaching process analogous diagram
Fig. 7 is different extraction algorithms and this paper algorithm effects comparison diagram
Embodiment
Below in conjunction with accompanying drawing and instantiation enforcement of the present invention is elaborated.
Be illustrated in figure 1 as the inventive method concrete steps process flow diagram, of the present inventionly comprise following concrete steps:
Step 5 is calculated its antibody concentration functional value C
V, find out in given domain value range, there is how much antibody, when max (Aff), wherein, C
VBe the antibody concentration of immune clone computing, Tn is given thresholding, and value is 0.8, and (i0 j0) is the coordinate points of the antibody in this domain value range to img, carries out the Immune Clone Selection operation; Obtain antibody A of new generation through step 4 and step 5
b, a new generation is that each of later generation of (b＞1) is for antibody;
There is new antibodies in to its neighborhood in step 6 in for antibody and antibody point itself that satisfy in affinity value and the antibody concentration domain value range carries out clone operations at each, remove do not satisfy condition each for the antibody point, in the next generation, will no longer operate;
Step 7 is A
bIn the new antibody gene that obtains make a variation, repeating step three is to step 6, b=b+1, b represent antibody algebraically, the initial value of b is 1.Work as A
bIn antibody when not changing program stop, promptly obtain optimum antibody.
1. initial antibodies production method
Initial antibodies all has certain influence to quality and the algorithm the convergence speed of finally separating.In order to help improving speed of convergence and antibody quality, the present invention utilizes priori to obtain quality initial antibodies preferably, can effectively provide tutorial message for search.
At first, the vein picture (being former figure) that collects is done the pre-service of gray scale scale merit.Gray scale scale merit computing formula is:
Img＝((img-min(img*255)/(max(img)-min(img) (1)
Then, pretreated vein picture is carried out the adaptive threshold method handle, adaptive threshold method mathematical formulae is:
Wherein, t is the window size that is adopted, and u, v are to be variable in the window of t between size, and value is in [0, t-1].I, j are the pixel coordinate of the former figure of correspondence.What this emulation was adopted is the moving window of 5*5 size.
The initial antibodies A that the adaptive threshold method is produced
1(i j) is:
2. mutation operator
The prerequisite of variation is near antibody point, and the variation mode is:
u＝4-8*rand(1，8) (4)
U is rounded, u is the random number of the interior generation of given range again.Assignment is given m, n, m, n for the random number span that generates in (5,5).Make i0=x
i+ α m; J0=y
i+ an, (Aff), Aff is the affinity value of current point to α=β * exp, and β is the variable of control characteristic function decay, generally gets 0.15, and α is the exponential damping amount.(x wherein
i, y
i) be current antibody point, (i0 j0) is the antibody point after the variation.
The gray-scale value of change point is:
Q(k)＝img(i0+kcosθ，j0+ksinθ) (5)
Wherein, k ∈ [1,2t+1]; Here get t=4.θ is the peripheral direction of change point, and expression is the filled circles zone of k with the radius.Q (k) is the pixel value of this region point.
3. affinity function and antibody concentration
Except the affinity function, in antibody, also have many antibody similar in threshold range.That is to say that the similarity degree that has of how much antibody and this antibody is arranged in threshold range.Antibody concentration not only passing threshold reflects similarity between the antibody, can also reflect the quantity of similar antibody.
Hence one can see that, and the structure of affinity function should carry out linear weighted function (LWF) computing to area pixel in conjunction with the characteristics in vein paddy shape zone.The linear weighted function function is the linear combination of Gaussian function and its second derivative.LWF has very strong denoising and strengthens the ability on border, can reach denoising and strengthen the equilibrium on border by the combination coefficient c0c2 that regulates them.Computing formula is:
L(x/δ)＝c
0h
0(x/δ)+c
2h
2(x/δ) (6)
Wherein, first h
0(x/ δ) is smoothing operator:
Second h
2(x/ δ) strengthens operator for the edge:
L (x/ δ) is the linear weighted function function, and δ is a gaussian coefficient, gets δ=1 here, and x is the computing variable.
Fig. 3 is analogous diagram before and after the match of adopting this method gained.Wherein, the vein pattern measured data y of L (x) under perfect condition, obtaining.h
0(x), h
0(x) be function without match.c
0, c
2Be weights, by to y=c
0h
0(x)+c
2h
2(x) carry out curve fitting, behind over-fitting, draw c
0=1.40, c
2=-3.95, this moment well balance denoising and strengthen the border.
H (k) is the pixel value of change point through the weighting function weighting, L (k) linear weighted function function.K is the computing variable, k ∈ [1,2t+1].Computing formula is:
H(k)＝L(k)*Q(k) (9)
C
vBe the antibody concentration of immune clone computing, function antibody concentration formula is:
Aff
1Be the mean value of pixel difference sum, t the pixel value sum mean value calculation formula of preceding t pixel value and back after the weighting is:
Aff is the affinity function of immune clone computing, and affinity function calculation formula is:
Aff＝1/aff
1 (12)
4. the clone of the death of non-vein antibody and vein pattern antibody
Fig. 4 is the vein pattern that contains noise antibody, judges through affinity and antibody concentration value, and it is carried out the vein pattern figure of the dead operation of antibody point front and back.
Fig. 5 is a vein pattern antibody, judges through affinity value and antibody concentration, and the vein antibody point of neighborhood space carried out the vein pattern figure of clone operations front and back around its antibody point was reached.
5. simulation result comparative analysis
Figure 6 shows that immune clone algorithm leaching process analogous diagram.That laterally describe among the figure is the sample figure of Different Individual; What vertically describe is same individuality, through immune clone algorithm from the generative process of original image-initial antibodies-centre for antibody-optimum antibody.
Figure 7 shows that different extraction algorithms and this paper algorithm effects comparison diagram.Have the sample figure of 4 Different Individual in the test.Comprise this method, shared 3 kinds of algorithms contrast.They are respectively document [2] (Xiang Yu, Wenming Yang.A Novel Finger Vein Pattern ExtractionApproach for Near-Infrared Image[J], Image and Signal Processing, 2009.CISP ' 09.2nd International Congress on 17-19Oct.2009Page (s): improved linear tracking 1-5Digital ObjectIdentifier 10.1109/CISP.2009.5304440), document [3] (YuChengbo, Qing Huafeng.A Research on Extracting Low Quality Human FingerVein Pattern Characteristics[J], Bioinformatics and Biomedical Engineering, 2008.ICBBE 2008.The 2nd International Conference on 16-18May 2008Page (s): direction template operator 1876-1879Digital Object Identifier 10.1109/ICBBE.2008.798) and the algorithm of this paper.As seen from the figure, found more noise information among the result that document [2,3] extracts samples pictures 1, document [3] is especially obvious, and the algorithm good restraining that this paper provides interference of noise.Though initial antibodies contains noise information in Fig. 6 sample Fig. 1,, eliminated noise information through suppressing mechanism; During to the extraction of sample Fig. 2, document [2,3] extracts insufficient on the information of some vein pattern, discontinuous part occurred, and the algorithm of this paper has well kept venous information, makes it more continuous, smooth.The problem that in Fig. 6 sample Fig. 2 initial antibodies, occurs document [2,3] equally, but, well kept passages through which vital energy circulates information through promotion mechanism to venous information; From sample Fig. 3,4 results that extract as can be known, the vein pattern that this paper algorithm extracts is more continuous, smooth, good restraining noise information and promoted the growth of venous information, and abundanter on the vein pattern details.
Claims (5)
1. finger vein features extracting method based on the immune clone of linear weighted function, described method at first, former figure to the finger vena picture produces initial antibodies with the adaptive threshold method, in order to determine search volume and scope, is the first generation antibody that obtains initial antibodies A
1(i0, coordinate figure j0) return among the former figure, and (i0 j0) carries out the affinity value and calculates, and by given affinity value decision method, produces new antibodies at the coordinate points img of correspondence;
Secondly, judge whether to have in the first generation antibody and do not satisfy affinity and antibody concentration; If exist, then carry out this antibody point and remove operation, draw second generation antibody A by clone operations
2(i0, j0);
Then, on the basis of second generation antibody, make a variation, affinity, antibody concentration calculate, and draws antibody A of new generation by clone operations
b(i0, j0);
At last, obtain optimum antibody.
2. the finger vein features extracting method of the immune clone based on linear weighted function according to claim 1, it specifically may further comprise the steps:
Step 1 produces initial antibodies A with the adaptive threshold method
1, setup parameter is determined the scope of search volume;
Step 2 is to the initial antibodies A that produces
1Carry out the mutation operation in the neighborhood, obtain change point img (i0, j0);
Step 3, to being the center with the change point, k is the interior point of the circle of radius, img (i0+kcos θ, j0+ksin θ) is weighted, and calculates its affinity value Aff and antibody concentration value C again
v
Step 4 is found out optimum antibody max (Aff) and is carried out the Immune Clone Selection operation;
Step 5 is calculated its antibody concentration functional value C
v, find out in given domain value range, there is how much antibody, as (x
i, y
i) time, wherein, C
vBe the antibody concentration of immune clone computing, Tn is given thresholding, and value is 0.8, and (i0 j0) is the coordinate points of the antibody in this domain value range to img, carries out the Immune Clone Selection operation; Obtain antibody A of new generation through step 4 and step 5
b, a new generation is that each of later generation of (b＞1) is for antibody;
There is new antibodies in to its neighborhood in step 6 in for antibody and antibody point itself that satisfy in affinity value and the antibody concentration domain value range carries out clone operations at each, remove do not satisfy condition each for the antibody point, in the next generation, will no longer operate;
Step 7 is A
bIn the new antibody gene that obtains make a variation, repeating step three is to step 6, b=b+1, b represent antibody algebraically, the initial value of b is 1, works as A
bIn antibody when not changing program stop, promptly obtain optimum antibody;
3. the finger vein features extraction algorithm of the immune clone based on linear weighted function according to claim 2, it is characterized in that: the production method of described initial antibodies is as follows:
At first, be that former figure does the pre-service of gray scale scale merit to the vein picture that collects, gray scale scale merit computing formula is:
Img＝((img-min(img))*255)/(max(img)-min(img)) (1)
Then, pretreated vein picture is carried out the adaptive threshold method handle, adaptive threshold method mathematical formulae is:
Wherein, t is the window size that is adopted, and u, v are to be variable in the window of t between size, and value is in [0, t-1]; I, j are the pixel coordinate of the former figure of correspondence; What this emulation was adopted is the moving window of 5*5 size;
The initial antibodies A that the adaptive threshold method is produced
1(i j) is:
4. the finger vein features extraction algorithm of the immune clone based on linear weighted function according to claim 2, it is characterized in that: the method for described mutation operation is as follows:
The prerequisite of variation is near antibody point, and the variation mode is:
u＝4-8*rand(1，8) (4)
U is rounded, u is the random number of the interior generation of given range again; Assignment is given m, n, m, n for the random number span that generates in (5,5); Make i0=x
i+ α m; J0=y
i+ α n, (Aff), Aff is the affinity value of current point to α=β * exp, and β is the variable of control characteristic function decay, generally gets 0.15, and α is exponential damping amount, wherein (x
i, y
i) be current antibody point, (i0 j0) is the antibody point after the variation;
The gray-scale value of change point is:
Q(k)＝img(i0+kcosθ，j0+ksinθ) (5)
Wherein, k ∈ [1,2t+1]; Here get t=4; θ is the peripheral direction of change point, and expression is the filled circles zone of k with the radius; Q (k) is the pixel value of this region point.
5. the finger vein features extraction algorithm of the immune clone based on linear weighted function according to claim 2, it is characterized in that: the computing formula of affinity function and antibody concentration is as follows:
L(x/δ)＝c
0h
0(x/δ)+c
2h
2(x/δ) (6)
Wherein, first h
0(x/ δ) is smoothing operator:
Second h
2(x/ δ) strengthens operator for the edge:
L (x/ δ) is the linear weighted function function, and δ is a gaussian coefficient, gets δ=1 here, and x is the computing variable;
H (k) is the pixel value of change point through the weighting function weighting, L (k) linear weighted function function.K is the computing variable, k ∈ [1,2t+1]; Computing formula is:
H(k)＝L(k)*Q(k) (9)
C
VBe the antibody concentration of immune clone computing, function antibody concentration formula is:
Aff
1Be the mean value of pixel difference sum, t the pixel value sum mean value calculation formula of preceding t pixel value and back after the weighting is:
Aff is the affinity function of immune clone computing, and affinity function calculation formula is:
Aff＝1/aff
1 (12)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2010102005181A CN101887517B (en) | 2010-06-13 | 2010-06-13 | Immune cloned finger venous image characteristic extraction method based on linear weighted function |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2010102005181A CN101887517B (en) | 2010-06-13 | 2010-06-13 | Immune cloned finger venous image characteristic extraction method based on linear weighted function |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101887517A true CN101887517A (en) | 2010-11-17 |
| CN101887517B CN101887517B (en) | 2012-09-26 |
Family
ID=43073432
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2010102005181A Active CN101887517B (en) | 2010-06-13 | 2010-06-13 | Immune cloned finger venous image characteristic extraction method based on linear weighted function |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101887517B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103778423A (en) * | 2013-11-25 | 2014-05-07 | 中国人民解放军国防科学技术大学 | Finger vein line restoration method based on gray scale neighbor and variance constraint |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1873658A (en) * | 2006-06-29 | 2006-12-06 | 武汉大学 | Method for selecting features of artificial immunity in remote sensing images |
| CN101271572A (en) * | 2008-03-28 | 2008-09-24 | 西安电子科技大学 | Image segmentation method based on immunity clone selection clustering |
| CN101625725A (en) * | 2008-07-11 | 2010-01-13 | 西安电子科技大学 | Artificial immunization non-supervision image classification method based on manifold distance |
| US20100098304A1 (en) * | 2008-10-17 | 2010-04-22 | Hitachi, Ltd. | Finger vein authentication device |
| CN101699514A (en) * | 2009-10-30 | 2010-04-28 | 西安电子科技大学 | Immune clone quantum clustering-based SAR image segmenting method |
-
2010
- 2010-06-13 CN CN2010102005181A patent/CN101887517B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1873658A (en) * | 2006-06-29 | 2006-12-06 | 武汉大学 | Method for selecting features of artificial immunity in remote sensing images |
| CN101271572A (en) * | 2008-03-28 | 2008-09-24 | 西安电子科技大学 | Image segmentation method based on immunity clone selection clustering |
| CN101625725A (en) * | 2008-07-11 | 2010-01-13 | 西安电子科技大学 | Artificial immunization non-supervision image classification method based on manifold distance |
| US20100098304A1 (en) * | 2008-10-17 | 2010-04-22 | Hitachi, Ltd. | Finger vein authentication device |
| CN101699514A (en) * | 2009-10-30 | 2010-04-28 | 西安电子科技大学 | Immune clone quantum clustering-based SAR image segmenting method |
Non-Patent Citations (2)
| Title |
|---|
| 《IEEE》 20080627 余成波等 手指静脉图像特征提取算法的研究 全文 1-5 , 2 * |
| 《IEEE》 20091031 E.C.Lee 等 Restoration method of skin scattering blurred vein image for finger vein recognition 全文 1-5 第45卷, 第21期 2 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103778423A (en) * | 2013-11-25 | 2014-05-07 | 中国人民解放军国防科学技术大学 | Finger vein line restoration method based on gray scale neighbor and variance constraint |
| CN103778423B (en) * | 2013-11-25 | 2016-09-21 | 中国人民解放军国防科学技术大学 | Finger vein grain restorative procedure based on gray scale neighbour and Variance Constraints |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101887517B (en) | 2012-09-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109829436B (en) | Multi-face tracking method based on depth appearance characteristics and self-adaptive aggregation network | |
| CN102682302B (en) | Human body posture identification method based on multi-characteristic fusion of key frame | |
| CN110765873B (en) | Facial expression recognition method and device based on expression intensity label distribution | |
| CN107240122A (en) | Video target tracking method based on space and time continuous correlation filtering | |
| Arif et al. | Automated body parts estimation and detection using salient maps and Gaussian matrix model | |
| CN106326857A (en) | Gender identification method and gender identification device based on face image | |
| CN111597991A (en) | Rehabilitation detection method based on channel state information and BilSTM-Attention | |
| CN112149557B (en) | Person identity tracking method and system based on face recognition | |
| CN114469120B (en) | Multi-scale Dtw-BiLstm-Gan electrocardiosignal generation method based on similarity threshold migration | |
| CN110991397B (en) | Travel direction determining method and related equipment | |
| CN111931654A (en) | Intelligent monitoring method, system and device for personnel tracking | |
| CN105930808A (en) | Moving object tracking method based on vector boosting template updating | |
| CN103105924A (en) | Man-machine interaction method and device | |
| CN111144165A (en) | Gait information identification method, system and storage medium | |
| CN110110727A (en) | The image partition method post-processed based on condition random field and Bayes | |
| CN114842238A (en) | Embedded mammary gland ultrasonic image identification method | |
| KR20170025535A (en) | Method of modeling a video-based interactive activity using the skeleton posture datset | |
| CN110458022A (en) | It is a kind of based on domain adapt to can autonomous learning object detection method | |
| CN103279944A (en) | Image division method based on biogeography optimization | |
| CN116311497A (en) | Tunnel worker abnormal behavior detection method and system based on machine vision | |
| CN106250814B (en) | A kind of finger venous image recognition methods based on hypersphere granulation quotient space model | |
| CN111144167A (en) | Gait information identification optimization method, system and storage medium | |
| CN113327269A (en) | Unmarked cervical vertebra movement detection method | |
| CN101887517B (en) | Immune cloned finger venous image characteristic extraction method based on linear weighted function | |
| CN104331700A (en) | Track-energy-diffusion-diagram-based group behavior identification method |
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 | ||
| C41 | Transfer of patent application or patent right or utility model | ||
| CB03 | Change of inventor or designer information |
Inventor after: Zhang Hui Inventor after: Ding Cheng Inventor after: Zhao Guodong Inventor before: Yu Chengbo Inventor before: Zhou Zhaomin Inventor before: Li Hongbing Inventor before: Tang Haiyan |
|
| COR | Change of bibliographic data | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20151014 Address after: 030006, Taiyuan, Shanxi province high tech Zone, South Central Street, No. 529, clear control innovation base, block C, 9 Patentee after: Shanxi Top Glory Tech Co.,Ltd. Address before: 400050 Chongqing city Yangjiaping Xingsheng Road No. 4 Patentee before: Chongqing University of Technology |
|
| C56 | Change in the name or address of the patentee | ||
| CP01 | Change in the name or title of a patent holder |
Address after: 030006, Taiyuan, Shanxi province high tech Zone, South Central Street, No. 529, clear control innovation base, block C, 9 Patentee after: SHANXI TOP GLORY CENTURY TECHNOLOGY CO.,LTD. Address before: 030006, Taiyuan, Shanxi province high tech Zone, South Central Street, No. 529, clear control innovation base, block C, 9 Patentee before: Shanxi Top Glory Tech Co.,Ltd. |
|
| CP03 | Change of name, title or address |
Address after: 030006 Room 0906, 9th Floor, Block C, Cleaning and Control Innovation Base, 529 South Central Street, Taiyuan Xuefu Park, Shanxi Comprehensive Reform Demonstration Area, Taiyuan City, Shanxi Province Patentee after: Holy Point Century Technology Co.,Ltd. Address before: 030006 9th Floor, Block C, Cleaning and Control Innovation Base, 529 Nanzhonghuan Street, Taiyuan High-tech Zone, Shanxi Province Patentee before: SHANXI TOP GLORY CENTURY TECHNOLOGY CO.,LTD. |
|
| CP03 | Change of name, title or address | ||
| EE01 | Entry into force of recordation of patent licensing contract |
Application publication date: 20101117 Assignee: Zhongguancun Technology Leasing Co.,Ltd. Assignor: Holy Point Century Technology Co.,Ltd. Contract record no.: X2021980014114 Denomination of invention: Feature extraction method of immune clonal finger vein image based on linear weighting Granted publication date: 20120926 License type: Exclusive License Record date: 20211206 |
|
| EE01 | Entry into force of recordation of patent licensing contract | ||
| PE01 | Entry into force of the registration of the contract for pledge of patent right |
Denomination of invention: Feature extraction method of immune clonal finger vein image based on linear weighting Effective date of registration: 20211208 Granted publication date: 20120926 Pledgee: Zhongguancun Technology Leasing Co.,Ltd. Pledgor: Holy Point Century Technology Co.,Ltd. Registration number: Y2021980014394 |
|
| PE01 | Entry into force of the registration of the contract for pledge of patent right | ||
| EC01 | Cancellation of recordation of patent licensing contract |
Assignee: Zhongguancun Technology Leasing Co.,Ltd. Assignor: Holy Point Century Technology Co.,Ltd. Contract record no.: X2021980014114 Date of cancellation: 20231222 |
|
| EC01 | Cancellation of recordation of patent licensing contract | ||
| PC01 | Cancellation of the registration of the contract for pledge of patent right |
Date of cancellation: 20231222 Granted publication date: 20120926 Pledgee: Zhongguancun Technology Leasing Co.,Ltd. Pledgor: Holy Point Century Technology Co.,Ltd. Registration number: Y2021980014394 |
|
| PC01 | Cancellation of the registration of the contract for pledge of patent right |