CN103310196A - Finger vein recognition method by interested areas and directional elements - Google Patents
Finger vein recognition method by interested areas and directional elements Download PDFInfo
- Publication number
- CN103310196A CN103310196A CN2013102324509A CN201310232450A CN103310196A CN 103310196 A CN103310196 A CN 103310196A CN 2013102324509 A CN2013102324509 A CN 2013102324509A CN 201310232450 A CN201310232450 A CN 201310232450A CN 103310196 A CN103310196 A CN 103310196A
- Authority
- CN
- China
- Prior art keywords
- image
- finger
- interest
- area
- vein
- 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
- Image Analysis (AREA)
- Collating Specific Patterns (AREA)
Abstract
The invention discloses a finger vein recognition method by interested areas and directional elements. Existing vein recognition algorithms can be divided into two types, namely local feature utilization and global feature utilization, but the problems of long time consumption, large computing capacity and low image quality cannot be solved simultaneously. The method includes firstly reading a finger vein image, extracting an interested area by the interested area extraction method based on rotation correction, then constructing vectors by adopting gradient magnitude and direction of pixel points on finger vein lines to represent direction feature of the vein lines in the interested area, combining the direction feature and membership degree of image sub-blocks, creating feature vectors of the vein image, finally during the finger vein matching process, measuring similarity of different linear feature vectors by adopting cross-correlation coefficient, and figuring out matching results. The method is used for finger vein recognition.
Description
Technical field:
The present invention relates to the finger vein identification method of a kind of area-of-interest and direction element.
Background technology:
Vein identification technology is to utilize the distribution situation of human body venae subcutaneae blood vessel to carry out the identity discriminating, except possessing uniqueness, ubiquity and stability, also be not vulnerable to the external environment influence, have plurality of advantages such as precision height, speed be fast, contactless, become a kind of effective means of identification.Existing vein recognizer can be divided into uses local feature and global characteristics two classes: the method for available technology adopting local feature point coupling is carried out vein identification, and effect is better, but owing to adopt exhaustive coordinate matching operation, consuming time longer; And the method that adopts the pixel comparison is carried out the vein coupling, exists the bigger problem of operand equally; Have with the end points of veinprint as unique point, also have the people then the bifurcation of veinprint to be mated as unique point, these two kinds of algorithm identified speed are very fast, but it is bigger influenced by picture quality.Use in the method for global characteristics, some people also with seven of vein image not bending moment merge as feature to be identified; Perhaps use the method for principal component analysis (PCA) to realize purpose to vein pattern dimensionality reduction coupling, in order to have overcome the not high shortcoming of azygos vein feature discrimination, also having combines principal component analysis (PCA) with ridgelet transform carries out vein identification, and recognition effect is good, but calculated amount is bigger.
Summary of the invention:
The finger vein identification method that the purpose of this invention is to provide a kind of area-of-interest and direction element.
Above-mentioned purpose realizes by following technical scheme:
The finger vein identification method of a kind of area-of-interest and direction element, at first the finger venous image utilization of reading is extracted area-of-interest based on the area-of-interest exacting method of rotation correction, then, employing refers to the gradient magnitude of pixel on the veinprint and the direction character that the directional structure vectorical structure vector characterizes described area-of-interest veinprint, and with the described direction character structure that combines with the degree of membership of image subblock, produce the proper vector of vein image; Refer to the vein matching stage at last, adopting cross-correlation coefficient to weigh the similarity of different linear feature vectors, drawing matching result.
The finger vein identification method of described area-of-interest and direction element, described area-of-interest exacting method based on rotation correction, first finger vein image is rotated correction before area-of-interest extracts, according to each position of finger the difference of near infrared light penetration capacity is extracted the area-of-interest of image on the basis of rotation correction, concrete grammar is:
(1) rotation correction:
Barycenter by the display foreground zone is rotated correction, and after taking out finger areas, calculating target image is the barycenter of finger areas image
, its computing formula is as follows:
Wherein,
In the presentation video
The horizontal ordinate of individual pixel,
In the presentation video
The ordinate of individual element,
Presentation video wide,
The height of presentation video,
The zone that belongs to finger in the presentation video;
After obtaining image centroid, find the middle point coordinate of last row place straight-line segment of finger contours image
O, tie point
CAnd the point
O, calculated line
With the horizontal direction line
Angle be the anglec of rotation
:
Wherein
,
Be respectively a little
CAnd the point
OHorizontal ordinate value; When
, namely
The time, image is turned clockwise; When
, namely
The time, image is rotated counterclockwise; When
, namely
The time, image is not rotated operation;
(2) area-of-interest is determined:
At first, projection is carried out to vertical direction in the entire image zone, the summation of the gray-scale value of every row pixel in the computed image
:
In the vertical direction projection,
Pixel range in find the zone of mean value maximum, the mid point that this is regional
As the articulation point of cutting apart, then the cut-off rule in interesting image regions vertical direction left side is taken as
, according to the left side cut-off rule
Determine the right side cut-off rule
For
In the formula
dRepresent two vertical parallel lines
,
Between distance, i.e. the transverse width of interesting image regions; Again with the internal tangent at the upper and lower edge of finger contours
,
With
,
The rectangular area of intersecting a sealing of formation is the area-of-interest of the finger vena that extracts.
Beneficial effect:
1. the present invention has taken full advantage of the directional information of veinprint, has higher accuracy of identification, and has the ability of certain antinoise and image geometry deformation.
The present invention is directed to the characteristics of the vein image that the noncontact mode gathers, design reduces the difficulty of successive image coupling based on the area-of-interest exacting method of rotation correction; And taking full advantage of the vein image lines roughly to the extension of finger two ends, direction changes mild characteristics design proper vector in regional area; Again this feature is combined with the degree of membership of image subblock and construct the proper vector of final vein image, thereby avoid not appear in the identical sub-piece same finger some point that is in the image subblock border in the image that difference is gathered constantly, but appear at error condition in the adjacent sub-piece.
Description of drawings:
Accompanying drawing 1 is method flow diagram of the present invention.
Accompanying drawing 2 is that the rotation correction key point is chosen synoptic diagram.
Accompanying drawing 3 is the finger vein images behind the rotation correction.
Accompanying drawing 4 is pixel column synoptic diagram of choosing in finger venous image.
Accompanying drawing 5 is gray-scale value curve maps of a row pixel.
Accompanying drawing 6 is gray-scale value curve maps of b row pixel.
Accompanying drawing 7 is gray-scale value curve maps of c row pixel.
Accompanying drawing 9 refers to vein ROI areal map.
Accompanying drawing 10 is 9 * 9 neighborhood templates of calculating pixel point direction.
Accompanying drawing 11 refers to the field of direction image of vein image.
Accompanying drawing 12 refers to the fuzzy piecemeal synoptic diagram of vein image.
Accompanying drawing 13 refers to vein image a.
Accompanying drawing 14 refers to vein image b.
Accompanying drawing 15 refers to vein image c.
Accompanying drawing 16 is legal coupling and the illegal matching distance scatter chart that mates.
Accompanying drawing 17 is fuzzy division methods translation test result figure.
Accompanying drawing 18 is fuzzy division methods rotary test figure as a result.
Embodiment:
Embodiment 1:
The finger vein identification method of a kind of area-of-interest and direction element, at first the finger venous image utilization of reading is extracted area-of-interest based on the area-of-interest exacting method of rotation correction, then, employing refers to the gradient magnitude of pixel on the veinprint and the direction character that the directional structure vectorical structure vector characterizes described area-of-interest veinprint, and with the described direction character structure that combines with the degree of membership of image subblock, produce the proper vector of vein image; Refer to the vein matching stage at last, adopting cross-correlation coefficient to weigh the similarity of different linear feature vectors, drawing matching result.
Embodiment 2:
Finger vein identification method according to embodiment 1 described area-of-interest and direction element, described area-of-interest exacting method based on rotation correction, first finger vein image is rotated correction before area-of-interest extracts, according to each position of finger the difference of near infrared light penetration capacity is extracted the area-of-interest of image on the basis of rotation correction, concrete grammar is:
(1) rotation correction:
Barycenter by the display foreground zone is rotated correction, and after taking out finger areas, calculating target image is the barycenter of finger areas image
, its computing formula is as follows:
Wherein,
In the presentation video
The horizontal ordinate of individual pixel,
In the presentation video
The ordinate of individual element,
Presentation video wide,
The height of presentation video,
The zone that belongs to finger in the presentation video;
After obtaining image centroid, find the middle point coordinate of last row place straight-line segment of finger contours image
O, tie point
CAnd the point
O, calculated line
With the horizontal direction line
Angle be the anglec of rotation
:
Wherein
,
Be respectively a little
CAnd the point
OHorizontal ordinate value; When
, namely
The time, image is turned clockwise; When
, namely
The time, image is rotated counterclockwise; When
, namely
The time, image is not rotated operation;
(2) area-of-interest is determined:
At first, projection is carried out to vertical direction in the entire image zone, the summation of the gray-scale value of every row pixel in the computed image
:
Wherein,
Be on the image
iRow
jThe pixel that lists,
HHeight for image;
In the vertical direction projection,
Pixel range in find the zone of mean value maximum, the mid point that this is regional
As the articulation point of cutting apart, then the cut-off rule in interesting image regions vertical direction left side is taken as
, according to the left side cut-off rule
Determine the right side cut-off rule
For
In the formula
dRepresent two vertical parallel lines
,
Between distance, i.e. the transverse width of interesting image regions; Again with the internal tangent at the upper and lower edge of finger contours
,
With
,
The rectangular area of intersecting a sealing of formation is the area-of-interest of the finger vena that extracts.
Embodiment 3:
Finger vein identification method according to embodiment 1 or 2 described area-of-interests and direction element, described area-of-interest should be positioned at all and refer to zone identical on the vein image, and main venous information all should be present in this zone, and for can be at further registration of when coupling, we need extract stable reference element as suitable reference point from refer to vein image, come image is positioned and the extraction of area-of-interest, to reduce effect of non-linear such as the rotation that causes in the sampling process, translation, distortion.
(1) rotation correction
Because contactless acquisition mode has more friendly, therefore finger vein image of the present invention storehouse adopts contactless mode to gather, owing to do not use locating device under this acquisition mode, the picker points putting position and direction is some difference, there be rotation and translation phenomenon in various degree in the vein image that makes different time obtain from same finger, and finger is not recessed such as referring to that uncut jade or finger-joint are bent to form, the feature of auxiliary ROI extracted region such as salient point, therefore, the present invention's proposition first finger vein image before the ROI extracted region is rotated correction, extract the ROI zone of image on the basis of rotation correction, can reduce the difficulty of successive image coupling, increase the robustness of system.
The barycenter of finger areas is the performance index that every width of cloth image all exists, and the calculating of barycenter is of overall importance, and its antijamming capability is stronger, so the barycenter of the present invention by the display foreground zone is rotated correction.After taking out finger areas, calculating target image is the barycenter of finger areas image
, its computing formula is as follows:
Wherein,
In the presentation video
The horizontal ordinate of individual pixel,
In the presentation video
The ordinate of individual element,
Presentation video wide,
The height of presentation video,
The zone that belongs to finger in the presentation video.
After obtaining image centroid, we find the straight-line segment at last row place of finger contours image, and determine the middle point coordinate of this line segment
O, tie point
CAnd the point
OIn alignment
, calculated line
With the horizontal direction line
Angle be the anglec of rotation
, (as shown in Figure 2) is rotated correction with this to image.
Computing formula as follows:
(4)
Wherein
,
Be respectively a little
CAnd the point
OHorizontal ordinate value.When
, namely
The time, image is turned clockwise; When
, namely
The time, image is rotated counterclockwise; When
, namely
The time, image is not rotated operation.Image behind the rotation correction as shown in Figure 3.
(2) area-of-interest is determined
Because the articulations digitorum manus position has cartilaginous tissue, obtaining under the mode of image based near infrared irradiation, articulations digitorum manus has stronger penetration capacity with respect to other position of finger, therefore, finger-joint position brightness ratio is bigger in the whole finger vein image, and namely the pixel value of this parts of images is than the other parts height.
We extract the pixel at a row articulations digitorum manus position from the image array that extracts finger areas, extract two row pixels again from other position, this three row gray values of pixel points is plotted curve map as shown in Figure 4, horizontal ordinate is the line number at the plain place of institute's capture among the figure, and ordinate is this gray values of pixel points.Can draw from figure, that row gray values of pixel points of joint part wants high with respect to non-joint part.Therefore, we can make joint position and then realize that the location of image cuts apart by finding out the higher Lieque of gray-scale value.
At first, projection is carried out to vertical direction in the entire image zone, the summation of the gray-scale value of every row pixel in the computed image
:
In the vertical direction projection,
Pixel range in find the zone of mean value maximum, the mid point that this is regional
As the articulation point of cutting apart, then the cut-off rule in interesting image regions vertical direction left side is taken as
, according to the left side cut-off rule
Determine the right side cut-off rule
For
In the formula
dRepresent two vertical parallel lines
,
Between distance, i.e. the transverse width of interesting image regions; Again with the internal tangent at the upper and lower edge of finger contours
,
With
,
Intersect to form the rectangular area of a sealing as shown in Figure 5, thereby the area-of-interest that extracts finger vena as shown in Figure 6.
Embodiment 4:
According to the finger vein identification method of embodiment 1 or 2 or 3 described area-of-interests and direction element, describe based on the finger vein pattern of direction:
(1) directivity of veinprint
From the topological structure of finger vena as can be known, its veinprint extends towards specific direction, has tangible directivity, this directivity can be represented by the direction of each pixel in the image, and the direction of pixel refers to that its gray-scale value keeps continuous direction, can judge according to the intensity profile in the neighborhood of pixel points.
At first the direction of pixel is carried out discretize, design the direction template of calculating pixel point, thereby calculate the direction of each pixel.Template representation shown in Figure 10 is with pixel
Centered by it discrete is turned to 8 directions, template size is 81 pixels.
Because the pixel on the veinprint and the gray scale difference maximum on the same lines vertical direction, and and the gray scale difference minimum between the point on the same direction of same lines, the step of obtaining the pixel direction according to this principle is as follows:
1) at first, obtains the pixel grey scale mean value of each pixel on 8 directions centered by this point
2) again will
Be divided into 4 groups by direction perpendicular to each other, calculate the absolute value of two average value differences in every group respectively
,
Two vertical direction in the direction group of value maximum
With
+ 4 is pixel
Possible veinprint direction;
3) at last will
With
In+4 average gray with
The direction that approaches of gray-scale value be
Ridge orientation
, that is:
Obtain after level and smooth the vein directional diagram as shown in figure 11, the different different discrete direction of greyscale color represent pixel point among the figure, as can be seen from the figure, the directivity of vein image is fairly obvious.
(2) direction character of vein image is described
Although above-mentioned field of direction image can reflect the directivity that refers to veinprint, this method antijamming capability is relatively poor, and since the orientation angle of pixel can for
Between any value, therefore, need a kind ofly to determine that the mode of continuous direction characterizes.The gradient of pixel has been described the size and Orientation of vein image grey scale change, therefore can detect the direction of pixel by the gradient of calculating pixel point.
If
IBe that size is
The finger vein image, at first utilize derivative operator to obtain the image slices vegetarian refreshments
Gray scale is along the partial derivative of level with vertical both direction
With
, utilize the Canny operator to calculate its gradient magnitude then
With the angle amplitude
As follows:
According to the principle of Canny operator extraction image outline, right
Direction on carry out maximum value compression, and gradient magnitude is set to zero less than the maximum point of certain threshold value.Like this,
In the amplitude of remaining veinprint wire-frame image vegetarian refreshments only just.We are with the direction of the pixel of veinprint
Be defined as the gradient direction perpendicular to this point, that is:
Because the gradient direction of veinprint point can be [90 °, + 90 °] between any value, so also desirable [0 ° of direction element of veinprint, 180 °] interval interior any value, be difficult to describe the direction property of veinprint pixel like this, simultaneously for fear of directly direction being quantized, we solve this problem by the ambiguity in definition set, definition 0 °, 45 °, 90 ° and 135 ° of four blur direction element sets in the domain of all direction elements in image, the half-open type membership function of above-mentioned four blur direction is defined as follows:
(13)
For each point
Carry it in above-mentioned four membership functions, having two values at least is 0, and the quadratic sum of above-mentioned four membership functions is 1.Our definable goes out each pixel thus
The characteristic of correspondence vector is:
This eigenvector not only contains the directional information that refers on the veinprint, but also contains the strength information of these lines.The above-mentioned point image that obtains is divided into
The height piece is designated as
, so far, the finger vein pattern that we define based on the direction element is described below:
Embodiment 5:
Finger vein identification method according to embodiment 1 or 2 or 3 or 4 described area-of-interests and direction element, describe based on the finger vein pattern of direction element and fuzzy piecemeal: the above-mentioned vein contour images that will refer to is divided in the method for sub-piece of non-overlapping copies, owing to may produce translation and rotation in the image acquisition process, therefore, the finger vein image that same finger is gathered constantly in difference, some point that is in sub-block boundary may not appear in the identical sub-piece, but appear in the adjacent sub-piece, thereby the accuracy of identification of generation error effect system.
In order to address this problem, we blur division to referring to the vein contour images, namely the piece that when piecemeal each image subblock is adjacent has overlapping pixel, the number of overlapping point is relevant with the length of side of blurred block, just there is not clear and definite border like this between the blurred block, the problem of having avoided clear and definite piecemeal to exist, thus the translation of image is had certain robustness with rotation.Above-mentioned fuzzy division thought as shown in figure 12, the pixel among the figure in the dashed circle is a pixel in the blurred block.
Determine that to the distance size at fuzzy sub-piece center this pixel belongs to the degree of membership of this sub-piece according to each pixel.The degree of membership calculation criterion is as follows:
Wherein
,
For each blurs the horizontal ordinate of the central point of sub-piece,
The length of side for sub-piece.
To each fuzzy sub-piece
, define a four-dimensional vector
,
Be the gross energy of having a few in this piece:
Wherein,
Be finger vein image size,
Be pixel
Gradient magnitude,
For the point
The direction membership function, wherein
,
Membership function for blurred block.Then the vector of all piece correspondences is formed one in the image
The proper vector of dimension.
Here in order to reduce the influence of illumination variation, right
Carry out following normalized:
By following formula as can be known, proper vector
Can reflection refer to the intensity of veinprint on each four different directions in zone of vein image, namely
The strength information and the structural information that have comprised veinprint simultaneously.
Embodiment 6:
Finger vein identification method according to embodiment 1 or 2 or 3 or 4 or 5 described area-of-interests and direction element adopts the cross-correlation coefficient between each proper vector to weigh the similarity degree that correspondence refers to vein image as matching degree.If
With
Be two and refer to the corresponding proper vector of vein image, then their cross-correlation coefficient is:
Wherein
,
With
,
It is respectively vector
With
The average of component and variance separately,
Value be in interval [1 ,+1], to change.
Obviously, if two width of cloth images
With
Come from same individual, then
Value bigger, otherwise
Value much smaller than 1.In Figure 13, Figure 14, three width of cloth vein images shown in Figure 15, preceding two width of cloth are the different vein images of gathering constantly of same people, the 3rd width of cloth image is another people's vein image, facies relationship numerical value between them is as shown in table 1, wherein the facies relationship numerical value between the non-homogeneous image is less, facies relationship numerical value between the homology different images is bigger, and the facies relationship numerical value between the same width of cloth image is 1.
Table 1
Embodiment 7:
Finger vein identification method according to embodiment 1 or 2 or 3 or 4 or 5 or 6 described area-of-interests and direction element, interpretation: the invention provides three groups of validity of testing to verify said method, experimental data base has been gathered the forefinger vein image of 300 samples, each sample collection image 5 width of cloth wherein, original image size is 320 * 240, after the area-of-interest exacting method that the present invention proposes was handled, the image size normalization was 124 * 64 pixels.
(1) matching performance analysis experiment
In order to analyze the matching performance of the inventive method, at first finger vein image samples all in the image library is carried out single sample matches operation, carry out coupling altogether 1500 * 1499=224850 time, wherein 300 * 5 * 4=6000 time is legal coupling, all the other are illegal coupling, and these two kinds of matching distance distribution curves as shown in figure 16.Wherein, transverse axis is represented the cross correlation numerical value of two width of cloth images, the longitudinal axis is the shared number percent of the sample of corresponding cross correlation numerical value, real, dashed curve is represented illegal coupling and legal coupling respectively, as can be seen from the figure, these two curve intersections are few, all have a tangible crest and this two crests have certain distance, show that this method can distinguish the vein image of different samples effectively.
(2) distinct methods identification control experiment
Compare for the recognition performance of verifying the inventive method and with method that document [1] (method one), document [2] (method two) adopt, 5 width of cloth of each sample refer to that vein image selects an image construction authentication image to be identified storehouse at random from database, totally 300 width of cloth images in the storehouse, remaining 4 width of cloth image construction template image storehouses of each sample, totally 300 * 4=1200 width of cloth image in the storehouse, utilize above-mentioned image library respectively these three kinds of methods to be carried out 1:1 authentication experiment and 1:n identification experiment, experimental result is shown in table 2 and table 3.
Table 2
Table 3
(3) experiment is analyzed in the rotation translation
Though through the image after the above-mentioned area-of-interest exacting method processing, be subjected to the influence of conversion such as translation and rotation less, but can not eliminate this influence fully, therefore, we carry out even translation respectively with the image in the database in 1 ~ 5 pixel coverage, in 1 ° ~+5 ° angular range, carry out Rotating with Uniform, form translation test library and rotary test storehouse test analysis is carried out in the translation of method and rotation susceptibility.To experimental result such as Figure 17 and shown in Figure 180 that the image in translation and the image rotating storehouse uses the inventive method to test, horizontal ordinate is the number of pixels of rotation among the figure, the discrimination of correspondence when ordinate is counted for the rotation different pixels.
From experimental result as can be seen, when referring to that vein image translation pixel count is between 1 ~ 5, the reject rate of the image after the translation and original image differ and are not very big, and the translation pixel count is 4 o'clock, discrimination still can reach 90%, and the anti-translation capability that shows the inventive method is 4 pixels; And the image anglec of rotation is between 1 ~+3 ° the time, reject rate and the original image of postrotational image are more or less the same, during with 4 pixels of image rotation, the discrimination of system becomes 86% from 90%, change greatlyyer, and during with 5 pixels of image rotation, the discrimination of system is less than 80%, discrimination declines to a great extent, and the anti-rotation performance that shows the inventive method is 3 pixels.
Claims (2)
1. the finger vein identification method of an area-of-interest and direction element, it is characterized in that: at first the finger venous image utilization of reading is extracted area-of-interest based on the area-of-interest exacting method of rotation correction, then, employing refers to the gradient magnitude of pixel on the veinprint and the direction character that the directional structure vectorical structure vector characterizes described area-of-interest veinprint, and with the described direction character structure that combines with the degree of membership of image subblock, produce the proper vector of vein image; Refer to the vein matching stage at last, adopting cross-correlation coefficient to weigh the similarity of different linear feature vectors, drawing matching result.
2. according to the finger vein identification method of the described area-of-interest of claim 1 and direction element, it is characterized in that: described area-of-interest exacting method based on rotation correction, first finger vein image is rotated correction before area-of-interest extracts, according to each position of finger the difference of near infrared light penetration capacity is extracted the area-of-interest of image on the basis of rotation correction, concrete grammar is:
(1) rotation correction:
Barycenter by the display foreground zone is rotated correction, and after taking out finger areas, calculating target image is the barycenter of finger areas image
, its computing formula is as follows:
Wherein,
In the presentation video
The horizontal ordinate of individual pixel,
In the presentation video
The ordinate of individual element,
Presentation video wide,
The height of presentation video,
The zone that belongs to finger in the presentation video;
After obtaining image centroid, find the middle point coordinate of last row place straight-line segment of finger contours image
O, tie point
CAnd the point
O, calculated line
With the horizontal direction line
Angle be the anglec of rotation
:
Wherein
,
Be respectively a little
CAnd the point
OHorizontal ordinate value; When
, namely
The time, image is turned clockwise; When
, namely
The time, image is rotated counterclockwise; When
, namely
The time, image is not rotated operation;
(2) area-of-interest is determined:
At first, projection is carried out to vertical direction in the entire image zone, the summation of the gray-scale value of every row pixel in the computed image
:
In the vertical direction projection,
Pixel range in find the zone of mean value maximum, the mid point that this is regional
As the articulation point of cutting apart, then the cut-off rule in interesting image regions vertical direction left side is taken as
, according to the left side cut-off rule
Determine the right side cut-off rule
For
In the formula
dRepresent two vertical parallel lines
,
Between distance, i.e. the transverse width of interesting image regions; Again with the internal tangent at the upper and lower edge of finger contours
,
With
,
The rectangular area of intersecting a sealing of formation is the area-of-interest of the finger vena that extracts.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310232450.9A CN103310196B (en) | 2013-06-13 | 2013-06-13 | The finger vein identification method of area-of-interest and direction element |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310232450.9A CN103310196B (en) | 2013-06-13 | 2013-06-13 | The finger vein identification method of area-of-interest and direction element |
Publications (2)
Publication Number | Publication Date |
---|---|
CN103310196A true CN103310196A (en) | 2013-09-18 |
CN103310196B CN103310196B (en) | 2017-12-01 |
Family
ID=49135392
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201310232450.9A Expired - Fee Related CN103310196B (en) | 2013-06-13 | 2013-06-13 | The finger vein identification method of area-of-interest and direction element |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN103310196B (en) |
Cited By (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105512629A (en) * | 2015-12-04 | 2016-04-20 | 南方医科大学 | Method for positioning bone and joint in human finger-vein image |
CN105512656A (en) * | 2014-09-22 | 2016-04-20 | 郭进锋 | Palm vein image collection method |
CN105975950A (en) * | 2016-05-27 | 2016-09-28 | 国创科视科技股份有限公司 | Rapid comparison method for three-dimensional finger vein similarity |
WO2017059591A1 (en) * | 2015-10-10 | 2017-04-13 | 厦门中控生物识别信息技术有限公司 | Finger vein identification method and device |
WO2018032861A1 (en) * | 2016-08-17 | 2018-02-22 | 广州广电运通金融电子股份有限公司 | Finger vein recognition method and device |
TWI622001B (en) * | 2017-07-24 | 2018-04-21 | 國立高雄第一科技大學 | Finger vein identification method |
CN108830158A (en) * | 2018-05-16 | 2018-11-16 | 天津大学 | The vein area-of-interest exacting method that finger contours and gradient distribution blend |
CN109829493A (en) * | 2019-01-25 | 2019-05-31 | 黑龙江大学 | Fingerprint and finger hand vein recognition fusion method based on feature hybrid matrix |
CN110647890A (en) * | 2019-08-28 | 2020-01-03 | 惠州市德赛西威智能交通技术研究院有限公司 | High-performance image feature extraction and matching method, system and storage medium |
CN111345668A (en) * | 2018-12-21 | 2020-06-30 | 苏州西顿家用自动化有限公司 | Automatic feeding mechanism of automatic cooker |
CN111639559A (en) * | 2020-05-15 | 2020-09-08 | 圣点世纪科技股份有限公司 | Finger vein machine learning identification method and device based on topographic concave-convex characteristics |
CN111797828A (en) * | 2020-06-19 | 2020-10-20 | 智慧眼科技股份有限公司 | Method and device for acquiring ROI (region of interest) of finger vein image and related equipment |
CN114841958A (en) * | 2022-05-04 | 2022-08-02 | 哈尔滨理工大学 | Method for automatically extracting image of brain subcutaneous tissue patch |
CN114898413A (en) * | 2022-07-15 | 2022-08-12 | 山东圣点世纪科技有限公司 | Vein identification method based on image contour direction field under complex background |
CN117746076A (en) * | 2024-02-19 | 2024-03-22 | 成都航空职业技术学院 | Equipment image matching method based on machine vision |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20060147094A1 (en) * | 2003-09-08 | 2006-07-06 | Woong-Tuk Yoo | Pupil detection method and shape descriptor extraction method for a iris recognition, iris feature extraction apparatus and method, and iris recognition system and method using its |
CN102663393A (en) * | 2012-03-02 | 2012-09-12 | 哈尔滨工程大学 | Method for extracting region of interest of finger vein image based on correction of rotation |
-
2013
- 2013-06-13 CN CN201310232450.9A patent/CN103310196B/en not_active Expired - Fee Related
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20060147094A1 (en) * | 2003-09-08 | 2006-07-06 | Woong-Tuk Yoo | Pupil detection method and shape descriptor extraction method for a iris recognition, iris feature extraction apparatus and method, and iris recognition system and method using its |
CN102663393A (en) * | 2012-03-02 | 2012-09-12 | 哈尔滨工程大学 | Method for extracting region of interest of finger vein image based on correction of rotation |
Non-Patent Citations (1)
Title |
---|
竺乐庆: "基于手部特征的多模态生物识别算法研究与系统实现", 《中国博士学位论文全文数据库信息科技辑》 * |
Cited By (23)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105512656A (en) * | 2014-09-22 | 2016-04-20 | 郭进锋 | Palm vein image collection method |
US10762366B2 (en) | 2015-10-10 | 2020-09-01 | Zkteco Co., Ltd. | Finger vein identification method and device |
WO2017059591A1 (en) * | 2015-10-10 | 2017-04-13 | 厦门中控生物识别信息技术有限公司 | Finger vein identification method and device |
CN105512629B (en) * | 2015-12-04 | 2019-02-05 | 南方医科大学 | A kind of human finger vein image Bones and joints localization method |
CN105512629A (en) * | 2015-12-04 | 2016-04-20 | 南方医科大学 | Method for positioning bone and joint in human finger-vein image |
CN105975950A (en) * | 2016-05-27 | 2016-09-28 | 国创科视科技股份有限公司 | Rapid comparison method for three-dimensional finger vein similarity |
WO2018032861A1 (en) * | 2016-08-17 | 2018-02-22 | 广州广电运通金融电子股份有限公司 | Finger vein recognition method and device |
CN107766776A (en) * | 2016-08-17 | 2018-03-06 | 广州广电运通金融电子股份有限公司 | A kind of finger vein identification method and device |
TWI622001B (en) * | 2017-07-24 | 2018-04-21 | 國立高雄第一科技大學 | Finger vein identification method |
CN108830158A (en) * | 2018-05-16 | 2018-11-16 | 天津大学 | The vein area-of-interest exacting method that finger contours and gradient distribution blend |
CN111345668A (en) * | 2018-12-21 | 2020-06-30 | 苏州西顿家用自动化有限公司 | Automatic feeding mechanism of automatic cooker |
CN109829493A (en) * | 2019-01-25 | 2019-05-31 | 黑龙江大学 | Fingerprint and finger hand vein recognition fusion method based on feature hybrid matrix |
CN110647890A (en) * | 2019-08-28 | 2020-01-03 | 惠州市德赛西威智能交通技术研究院有限公司 | High-performance image feature extraction and matching method, system and storage medium |
CN110647890B (en) * | 2019-08-28 | 2022-05-27 | 惠州市德赛西威智能交通技术研究院有限公司 | High-performance image feature extraction and matching method, system and storage medium |
CN111639559A (en) * | 2020-05-15 | 2020-09-08 | 圣点世纪科技股份有限公司 | Finger vein machine learning identification method and device based on topographic concave-convex characteristics |
WO2021227331A1 (en) * | 2020-05-15 | 2021-11-18 | 圣点世纪科技股份有限公司 | Finger vein machine learning recognition method and apparatus based on terrain concave and convex chjaracteristic |
CN111639559B (en) * | 2020-05-15 | 2022-07-05 | 圣点世纪科技股份有限公司 | Finger vein machine learning identification method and device based on topographic concave-convex characteristics |
CN111797828A (en) * | 2020-06-19 | 2020-10-20 | 智慧眼科技股份有限公司 | Method and device for acquiring ROI (region of interest) of finger vein image and related equipment |
CN114841958A (en) * | 2022-05-04 | 2022-08-02 | 哈尔滨理工大学 | Method for automatically extracting image of brain subcutaneous tissue patch |
CN114898413A (en) * | 2022-07-15 | 2022-08-12 | 山东圣点世纪科技有限公司 | Vein identification method based on image contour direction field under complex background |
CN114898413B (en) * | 2022-07-15 | 2022-09-13 | 山东圣点世纪科技有限公司 | Vein identification method based on image contour direction field under complex background |
CN117746076A (en) * | 2024-02-19 | 2024-03-22 | 成都航空职业技术学院 | Equipment image matching method based on machine vision |
CN117746076B (en) * | 2024-02-19 | 2024-04-26 | 成都航空职业技术学院 | Equipment image matching method based on machine vision |
Also Published As
Publication number | Publication date |
---|---|
CN103310196B (en) | 2017-12-01 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN103310196A (en) | Finger vein recognition method by interested areas and directional elements | |
Lin et al. | Line segment extraction for large scale unorganized point clouds | |
CN102663393B (en) | Method for extracting region of interest of finger vein image based on correction of rotation | |
Zhao et al. | Adaptive fingerprint pore modeling and extraction | |
Guerreiro et al. | Connectivity-enforcing Hough transform for the robust extraction of line segments | |
CN107748877A (en) | A kind of Fingerprint recognition method based on minutiae point and textural characteristics | |
Kumar Mishra et al. | A review of optical imagery and airborne lidar data registration methods | |
CN108009472B (en) | Finger back joint print recognition method based on convolutional neural network and Bayes classifier | |
CN109409292A (en) | The heterologous image matching method extracted based on fining characteristic optimization | |
CN103136525B (en) | A kind of special-shaped Extended target high-precision locating method utilizing Generalized Hough Transform | |
CN104112115A (en) | Three-dimensional face detection and identification technology | |
CN103984920B (en) | Three-dimensional face identification method based on sparse representation and multiple feature points | |
CN106023187A (en) | Image registration method based on SIFT feature and angle relative distance | |
Oldal et al. | Hand geometry and palmprint-based authentication using image processing | |
CN104123547A (en) | Improved directional filter and flexible matching based recognition method | |
Wang | Automatic extraction of building outline from high resolution aerial imagery | |
CN103336964A (en) | SIFT image matching method based on module value difference mirror image invariant property | |
Han et al. | Harris corner detection algorithm at sub-pixel level and its application | |
CN113516096B (en) | Finger vein ROI (region of interest) region extraction method and device | |
You et al. | Finger vein recognition method based on gaussian low-pass filter and direction detection | |
CN106056599B (en) | A kind of object recognition algorithm and device based on Object Depth data | |
Yan et al. | An improved matching algorithm for feature points matching | |
Zhang et al. | Satellite Cloud Image Registration by Combining Curvature Shape Representation with Particle Swarm Optimization. | |
Zhao et al. | Oil tanks extraction from high resolution imagery using a directional and weighted hough voting method | |
Yang et al. | A New Algorithm for Line Detection Based on Relative Maximum Point |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant | ||
CF01 | Termination of patent right due to non-payment of annual fee | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20171201 Termination date: 20200613 |