CN115223211B - Identification method for converting vein image into fingerprint image - Google Patents

Identification method for converting vein image into fingerprint image Download PDF

Info

Publication number
CN115223211B
CN115223211B CN202211140193.1A CN202211140193A CN115223211B CN 115223211 B CN115223211 B CN 115223211B CN 202211140193 A CN202211140193 A CN 202211140193A CN 115223211 B CN115223211 B CN 115223211B
Authority
CN
China
Prior art keywords
fingerprint
finger vein
image
vein
contour map
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.)
Active
Application number
CN202211140193.1A
Other languages
Chinese (zh)
Other versions
CN115223211A (en
Inventor
李学双
王丽
赵国栋
辛传贤
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Shandong Shengdian Century Technology Co ltd
Original Assignee
Shandong Shengdian Century Technology Co ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Shandong Shengdian Century Technology Co ltd filed Critical Shandong Shengdian Century Technology Co ltd
Priority to CN202211140193.1A priority Critical patent/CN115223211B/en
Publication of CN115223211A publication Critical patent/CN115223211A/en
Application granted granted Critical
Publication of CN115223211B publication Critical patent/CN115223211B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

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
    • G06V10/30—Noise filtering
    • 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
    • G06V10/36—Applying a local operator, i.e. means to operate on image points situated in the vicinity of a given point; Non-linear local filtering operations, e.g. median filtering
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/40—Extraction of image or video features
    • G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
    • 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

Abstract

The invention relates to an identification method for converting a vein image into a fingerprint image, which belongs to the technical field of biological feature identification and comprises the following steps: s1, collecting a finger vein image, wherein fingerprint textures exist in the finger vein image, and sequentially carrying out binarization and edge detection on the finger vein image to obtain a finger vein image contour map containing the fingerprint textures; s2, dividing the finger vein image contour map into a fingerprint contour map and a finger vein contour map based on a linear regression method, and defining the fingerprint texture of the corresponding position covered by the finger vein in the finger vein contour map as the finger vein texture; s3, enhancing the fingerprint contour map by adopting a fingerprint enhancement algorithm; s4, repairing the finger vein texture based on the finger vein intersection points; and S5, fusing the fingerprint contour map after the enhancement treatment and the repaired finger vein texture to form a fingerprint image to be identified. The invention can effectively improve the safety and the accuracy of vein identification by converting the finger vein image into the fingerprint image.

Description

Identification method for converting vein image into fingerprint image
Technical Field
The invention belongs to the technical field of biological feature identification, and particularly relates to an identification method for converting a vein image into a fingerprint image.
Background
Vein technology is widely used as biometric technology by virtue of its privacy and security, but with the commercialization of vein technology, users have increasingly demanded vein identification technology. Firstly, in the application process, the safety of the method needs to be ensured, the prosthesis attack is avoided, the existing scheme generally adopts multi-mode identity card authentication, but the mode can greatly increase the equipment cost; in addition, it is necessary to further improve the accuracy of vein identification and reduce the false identification rate.
To solve the above problems, some biometric identification combines finger vein identification with fingerprint identification, for example, a composite identification method, device and system based on fingerprint and finger vein disclosed in the patent of chinese invention with the granted publication number CN105678233B, the method includes: respectively acquiring a digital fingerprint image and a digital finger vein image through an image sensor; generating an image characteristic value according to a preset rule according to the digital fingerprint image and the digital finger vein image; and comparing the image characteristic value with a pre-stored target characteristic value, and judging whether the image characteristic value is matched with the target characteristic value.
Although the identification rate can be improved to a certain extent by the identification method, the method needs to collect the finger vein image and the fingerprint image respectively and process and fuse the finger vein image and the fingerprint image respectively, the calculation process is complex, the identification efficiency is poor, and the cost of equipment is high due to the fact that two sets of collection equipment are needed.
Disclosure of Invention
The invention provides an identification method for converting a vein image into a fingerprint image, which solves the problems that the existing vein identification cannot carry out living body detection, has low identification rate and the like on the premise of not increasing the cost of acquisition equipment.
In order to solve the technical problems, the technical scheme provided by the invention is as follows:
the invention relates to an identification method for converting a vein image into a fingerprint image, which comprises the following steps:
s1, collecting a finger vein image, wherein fingerprint textures exist in the finger vein image, and sequentially carrying out binarization and edge detection on the finger vein image to obtain a finger vein image contour map containing the fingerprint textures;
s2, dividing the finger vein image contour map into a fingerprint contour map and a finger vein contour map based on a linear regression method, and defining the fingerprint texture of the corresponding position covered by the finger vein in the finger vein contour map as the finger vein texture;
s3, enhancing the fingerprint contour map by adopting a fingerprint enhancement algorithm;
s4, searching fingerprint finger vein intersections in the finger vein contour map, and repairing finger vein textures on the basis of the fingerprint finger vein intersections;
and S5, fusing the fingerprint contour map after the enhancement treatment and the repaired finger vein texture to form a fingerprint image to be identified, and finishing identity identification based on the fingerprint image to be identified.
Preferably, the step S1 further includes performing thinning processing on the finger vein image contour map, so that contour lines in the finger vein image contour map are all single-pixel curves. The thinning processing of the finger vein image contour map can avoid the influence of the contour line with the pixel thickness more than 1 on the later curvature calculation.
Preferably, the step S2 of dividing the finger vein image profile map into the fingerprint profile map and the finger vein profile map based on a linear regression method specifically includes the steps of:
s2.1, respectively calculating the curvature of each contour point on each contour line in the finger vein image contour map, wherein the calculation formula is as follows:
Figure DEST_PATH_IMAGE001
in the formula, the first step is that,
Figure 44612DEST_PATH_IMAGE002
is a contour point at
Figure DEST_PATH_IMAGE003
The first derivative of the direction is,
Figure 91197DEST_PATH_IMAGE004
is a contour point at
Figure DEST_PATH_IMAGE005
The first derivative of the direction is,
Figure 567089DEST_PATH_IMAGE006
is a contour point at
Figure 341010DEST_PATH_IMAGE003
The second derivative of the direction is the derivative of,
Figure DEST_PATH_IMAGE007
is a contour point at
Figure 347144DEST_PATH_IMAGE008
The second derivative of the direction;
s2.2, counting that the curvature value on each contour line in the finger vein image contour map is lower than
Figure DEST_PATH_IMAGE009
The number of contour points of (a) is in proportion to the total number of contour points of the contour line
Figure 859902DEST_PATH_IMAGE010
(ii) a When ratio of
Figure 256380DEST_PATH_IMAGE010
Greater than a set proportional threshold
Figure DEST_PATH_IMAGE011
If so, judging that the contour line belongs to the fingerprint contour map; when ratio of
Figure 207893DEST_PATH_IMAGE010
Less than a set proportional threshold
Figure 6216DEST_PATH_IMAGE012
And if so, judging that the contour line belongs to the finger vein contour map.
Preferably, the step S3 of enhancing the fingerprint profile adopts a composite enhancement algorithm combining gaussian filtering, mean filtering and guided filtering, and the method specifically includes the following steps:
s3.1, smoothing the fingerprint contour map divided in the step S2 by adopting Gaussian filtering;
s3.2, smoothing the fingerprint contour map subjected to Gaussian filtering by adopting mean filtering;
s3.3, smoothing the fingerprint contour map divided in the step S2 by adopting guide filtering;
and S3.4, respectively obtaining detail images after different filtering processes based on the images obtained in the steps S3.1-S3.3, and enhancing the fingerprint outline image through the detail images.
Preferably, the enhancement of the fingerprint profile map by the detail image in step S3.4 is represented as:
Figure DEST_PATH_IMAGE013
in the formula, the first step is that,
Figure 955455DEST_PATH_IMAGE014
for the enhanced image of the final fingerprint profile,
Figure 291758DEST_PATH_IMAGE015
for the fingerprint profile map divided in step S2,
Figure DEST_PATH_IMAGE016
for the image processed in step S3.1,
Figure 220531DEST_PATH_IMAGE017
for the image processed at step S3.2,
Figure DEST_PATH_IMAGE018
for the image processed in step S3.3,
Figure 529870DEST_PATH_IMAGE019
the adjustment coefficient of the detail image of the fingerprint contour map after Gaussian filtering processing,
Figure 100002_DEST_PATH_IMAGE020
the adjustment coefficient of the detail image of the fingerprint contour map after the average value filtering processing is represented,
Figure 885896DEST_PATH_IMAGE021
the adjustment coefficient of the detail image of the fingerprint contour map after the guide filtering processing,
Figure 489922DEST_PATH_IMAGE022
。
gaussian filter can be within a specified filter radius, when pixels in the neighborhood of an image are smoothed, pixels at different positions in the neighborhood are endowed with different weights, and more overall gray level distribution characteristics of the image are reserved; the average filtering can replace each pixel value of the original image by the average value of each pixel of the filtering area within the designated filtering radius, and the image is smoothed by reducing the sharp change of the gray value of the image; the guiding filter can be within a specified filter radius, and the guiding graph and the original graph are in a linear relation. When the guide map is steep, the output image should change as the guide map changes, and when the guide map is gentle, the output image should be close to the input image. The fingerprint profile is enhanced through the combination of three filtering modes, so that the detail images in the fingerprint profile can be enhanced to different degrees, the gradient value can be stretched through multiple times of smooth noise reduction processing, gradient amplification can be realized, the amplification ratio of noise is smaller than that of fine lines, and the influence of the noise on the identification result is reduced.
Preferably, in step S4, the sum of absolute values of pixel differences between all two adjacent pixel points in the eight neighborhoods of the finger vein intersection is 6 × 255, and the specific step of step S4 includes:
s4.1, dividing the finger vein contour map into a plurality of block areas, and extracting a direction field of each block area;
s4.2, traversing and searching fingerprint finger vein intersections in the finger vein contour map, and calculating the directions of the fingerprint finger vein intersections based on the vein direction fields of the fingerprint finger vein intersections and the direction fields of the block areas where the fingerprint finger vein intersections are located;
and S4.3, repairing the vein texture based on the direction of the fingerprint vein intersection.
Preferably, the specific steps in step S4.1 include:
s4.1.1, estimating the direction of each pixel point i in the finger vein contour map, comprising the following steps: placing the pixel point in
Figure DEST_PATH_IMAGE023
In the neighborhood window, the trend of the pixel point is divided into 8 directions, and the included angle between two adjacent directions is
Figure 550282DEST_PATH_IMAGE024
Separately calculating said
Figure 556153DEST_PATH_IMAGE023
Dividing the gray average values into four groups according to the direction perpendicular to each other, calculating the difference value of the two gray average values in each group, wherein the two directions with the largest absolute value of the difference value are the prediction directions of the pixel point, respectively calculating the difference value of the gray average values in the two prediction directions and the gray value of the pixel point, and taking the direction with the smaller absolute value of the difference value as the final direction field of the pixel point
Figure DEST_PATH_IMAGE025
；
S4.1.2, divide the contour map of finger vein into
Figure 348659DEST_PATH_IMAGE026
The block area of the size is calculated according to the step S4.1, and the direction fields of all pixel points in the block area are calculated according to the direction fields
Figure 456292DEST_PATH_IMAGE025
The direction with the highest ratio is taken as the direction of the block area
Figure DEST_PATH_IMAGE027
. Preferably. In step S4.2, the calculation formula of the vein direction field of the fingerprint finger vein intersection is:
Figure 225403DEST_PATH_IMAGE028
in the formula, the first step is that,
Figure DEST_PATH_IMAGE029
representing the magnitude of the field in the direction of the intersection of the finger veins,
Figure 852825DEST_PATH_IMAGE030
the coordinates of the point are represented by,
Figure DEST_PATH_IMAGE031
representing the derivative of the fingerprint finger vein intersection in the x-direction,
Figure 320627DEST_PATH_IMAGE032
indicating finger vein intersection in fingerprint
Figure DEST_PATH_IMAGE033
The derivative of the direction.
Preferably, in step S4.2, the calculation formula of the direction of the intersection of the finger vein of the fingerprint is:
Figure 463027DEST_PATH_IMAGE034
in the formula, the first step is that,
Figure DEST_PATH_IMAGE035
the direction of the intersection of the finger veins is the fingerprint,
Figure 301407DEST_PATH_IMAGE036
and
Figure DEST_PATH_IMAGE037
respectively in the direction of the block area
Figure 580073DEST_PATH_IMAGE038
Vein direction field of intersection point of finger vein and fingerprint
Figure DEST_PATH_IMAGE039
A weighting coefficient of
Figure 291546DEST_PATH_IMAGE040
。
Preferably, the specific steps in step S4.3 are:
s4.3.1, setting a matching threshold value
Figure DEST_PATH_IMAGE041
；
S4.3.2, calculating one side of the edge of the contour line of the finger veinAny fingerprint finger vein cross point
Figure 685356DEST_PATH_IMAGE042
And any fingerprint finger vein intersection point on the other side of the edge of the finger vein contour line
Figure DEST_PATH_IMAGE043
Difference in direction of
Figure 297734DEST_PATH_IMAGE044
When is coming into contact with
Figure DEST_PATH_IMAGE045
Judging the matching of the two points;
s4.3.3, if the number of the fingerprint finger vein intersections which can not be successfully matched in any intersection point on one side of the edge of the finger vein contour line is more than 30 percent of the total number of the fingerprint finger vein intersections on the corresponding side, judging that the matching threshold value set in the step S5.1 is matched
Figure 392424DEST_PATH_IMAGE041
Invalid; otherwise, if the number of the fingerprint finger vein intersections which are not successfully matched in any side intersection of the edge of the finger vein contour line is less than or equal to 30% of the total number of the fingerprint finger vein intersections on the corresponding side, the matching threshold set in the step S5.1 is determined
Figure 884585DEST_PATH_IMAGE041
Valid, record the matching threshold
Figure 673680DEST_PATH_IMAGE041
The lower fingerprint refers to the group number of vein intersections successfully matched;
s4.3.4. Increasing the matching threshold gradually
Figure 276700DEST_PATH_IMAGE046
Until the threshold is matched, steps S5.2 and S5.3 are repeated
Figure 497335DEST_PATH_IMAGE046
Increasing to 30 deg., will match the valid threshold andcounting the corresponding pairing groups, and taking the matching threshold with the most pairing groups as a reference matching threshold;
and S4.3.5, connecting the fingerprint finger vein intersections matched with each other under the reference matching threshold value to simulate the finger vein texture.
The step S5 of completing identity recognition based on the fingerprint image to be recognized specifically comprises the following steps: and acquiring a texture image with fingerprint characteristics in a fingerprint image normalization processing mode, and identifying the texture image.
Compared with the prior art, the technical scheme provided by the invention has the following beneficial effects:
1. the identification method for converting the vein image into the fingerprint image can convert the vein image into the fingerprint image with identity identifiability, can effectively prevent prosthesis attack and improve the safety of vein identification; when the prosthesis attacks, the vein gray level image is generally printed on the prosthesis, but fingerprint textures are not printed at the same time, so that the fingerprint image is extracted from the collected finger vein image in the finger vein identification process, and if the fingerprint image with the identity identifiability is not extracted, the prosthesis attack can be determined, the registration is not performed, and the safety performance of the vein identification is improved.
2. The invention relates to an identification method for converting vein images into fingerprint images, which is a multi-mode identity authentication mode that collected vein images are converted into fingerprint images with identity identifiability and finger vein identity authentication and fingerprint identity authentication can be carried out simultaneously.
3. The identification method for converting the vein image into the fingerprint image adopts a composite enhancement algorithm combining Gaussian filtering, mean filtering and guided filtering for the fingerprint contour map, enhances the fingerprint contour map by combining three filtering modes, can strengthen the detail image in the fingerprint texture part to different degrees, can stretch the gradient value by multiple times of smooth noise reduction processing, can realize gradient amplification, can ensure that the amplification ratio of noise is smaller than that of fine lines, and reduces the influence of the noise on the identification result.
Drawings
FIG. 1 is a display of a fingerprint texture portion in an acquired vein image;
fig. 2 is a diagram of the fingerprint texture part after being enhanced by combining three filtering methods.
Detailed Description
For further understanding of the present invention, the present invention will be described in detail with reference to examples, which are provided for illustration of the present invention but are not intended to limit the scope of the present invention.
The invention relates to an identification method for converting a vein image into a fingerprint image, which comprises the following steps:
s1, collecting a finger vein image, wherein fingerprint textures exist in the finger vein image, and sequentially carrying out binarization and edge detection on the finger vein image to obtain a finger vein image contour map containing the fingerprint textures.
The specific steps in this example are: collecting a finger vein image, wherein fingerprint textures exist in the finger vein image, preprocessing the finger vein image, removing noise brought by equipment in the finger vein image and influence brought by finger pressure, then sequentially carrying out binarization processing on the finger vein image, and carrying out edge detection through a canny operator to obtain a finger vein image contour map containing the fingerprint textures; since the thickness of the pixel of the partial contour point is more than 1, which will have adverse effect on the later curvature calculation, the step also comprises the step of thinning the partial contour in the finger vein image contour map, so that the contour lines in the finger vein image contour map are all single-pixel curves.
S2, dividing the finger vein image contour map into a fingerprint contour map and a finger vein contour map part based on a linear regression method, and defining the fingerprint texture of the corresponding position covered by the finger vein in the finger vein contour map as the finger vein texture, wherein the specific steps are as follows:
s2.1, for any point on any contour line
Figure DEST_PATH_IMAGE047
，
Figure 707867DEST_PATH_IMAGE048
At the front of the point are
Figure DEST_PATH_IMAGE049
A dot is followed by
Figure 138586DEST_PATH_IMAGE050
A point having a total length of
Figure 100002_DEST_PATH_IMAGE051
Partial derivative thereof
Figure 92767DEST_PATH_IMAGE052
，
Figure 100002_DEST_PATH_IMAGE053
Respectively as follows:
Figure 964646DEST_PATH_IMAGE054
Figure 100002_DEST_PATH_IMAGE055
according to the method and the device, in the process of obtaining the partial derivative of a certain pixel point, the adjacent front and rear pixel points are not directly selected, but a plurality of pixel points are respectively selected in front of and behind the pixel point, so that the problem that the calculation accuracy is not high due to the fact that a certain contour point on a contour line deviates can be effectively solved.
Respectively calculating the curvature of each contour point on each contour line in the finger vein image contour map, wherein the calculation formula is as follows:
Figure 611659DEST_PATH_IMAGE056
in the formula, the first step is that,
Figure 145320DEST_PATH_IMAGE002
is a contour point at
Figure 824564DEST_PATH_IMAGE003
The first derivative of the direction is,
Figure 255676DEST_PATH_IMAGE004
is a contour point at
Figure 260541DEST_PATH_IMAGE005
The first derivative of the direction is,
Figure 10060DEST_PATH_IMAGE006
is a contour point at
Figure 492994DEST_PATH_IMAGE003
The second derivative of the direction is the derivative of,
Figure 100002_DEST_PATH_IMAGE057
is a contour point at
Figure 840930DEST_PATH_IMAGE008
A second derivative of direction;
S2.2.
Figure 385DEST_PATH_IMAGE058
the value can reflect the bending degree of the curve, the value is closer to 1, the curve is straighter, the contour line of the finger vein is generally a straight line, and the contour line of the fingerprint is generally a bent curve, so that the contour line in the contour map of the finger vein image can be divided into the fingerprint contour or the finger vein contour through the curvature, and further, the fingerprint contour map and the finger vein contour map are divided. The specific method in the embodiment is as follows: the curve value on each contour line in the contour map of the statistical finger vein image is lower than
Figure 987932DEST_PATH_IMAGE058
The number of contour points of (a) is in proportion to the total number of contour points of the contour line
Figure 25290DEST_PATH_IMAGE010
(ii) a When in proportion
Figure 414683DEST_PATH_IMAGE010
Greater than a set proportional threshold
Figure 213880DEST_PATH_IMAGE011
If yes, judging that the contour line belongs to the fingerprint contour map; when ratio of
Figure 688724DEST_PATH_IMAGE010
Less than a set proportional threshold
Figure 264193DEST_PATH_IMAGE011
If so, judging that the contour line belongs to the finger vein contour map; specific proportional threshold
Figure 508092DEST_PATH_IMAGE011
Sum curvature value
Figure 292507DEST_PATH_IMAGE058
Running batch setting can be carried out according to the quality and other conditions of the collected finger vein images;
s3, enhancing the fingerprint profile map by adopting a fingerprint enhancement algorithm: in the process of collecting the finger veins, the grains of the fingerprints are very shallow, so a combined enhancement algorithm combining Gaussian filtering, mean filtering and guided filtering is adopted in the step, and the specific steps are as follows:
s3.1, adopting the filter radius of
Figure 100002_DEST_PATH_IMAGE059
The Gaussian filtering is carried out on the fingerprint profile map divided in the step S2
Figure 67696DEST_PATH_IMAGE060
Performing smoothing to obtain fingerprint profile
Figure 696124DEST_PATH_IMAGE016
；
S3.2, adopting the filter radius of
Figure 100002_DEST_PATH_IMAGE061
The mean value filtering is carried out on the fingerprint contour map after the Gaussian filtering processing
Figure 371693DEST_PATH_IMAGE016
Performing smoothing to obtain fingerprint profile
Figure 810896DEST_PATH_IMAGE017
；
S3.3. Adopting the filter radius of
Figure 260332DEST_PATH_IMAGE062
The guide filtering is performed on the fingerprint contour map divided in the step S2
Figure 676139DEST_PATH_IMAGE060
Performing smoothing to obtain fingerprint profile
Figure 894630DEST_PATH_IMAGE018
；
S3.4, respectively obtaining detail images after different filtering processes based on the images obtained in the steps S3.1-S3.3, enhancing the fingerprint contour map through the detail images, wherein the enhancement of the fingerprint contour map through the detail images is represented as follows:
Figure 100002_DEST_PATH_IMAGE063
in the formula, the first step is that,
Figure 567051DEST_PATH_IMAGE064
for the enhanced image of the final fingerprint profile,
Figure 487472DEST_PATH_IMAGE015
for dividing in step S2The separated fingerprint outline graph is shown in the figure,
Figure 723281DEST_PATH_IMAGE016
for the image processed in step S3.1,
Figure 547012DEST_PATH_IMAGE017
for the image processed at step S3.2,
Figure 577284DEST_PATH_IMAGE018
for the image processed at step S3.3,
Figure 318756DEST_PATH_IMAGE019
the adjustment coefficient of the detail image of the fingerprint contour map after Gaussian filtering processing,
Figure 577831DEST_PATH_IMAGE020
the adjustment coefficient of the detail image of the fingerprint contour map after the mean value filtering processing,
Figure 770914DEST_PATH_IMAGE021
the adjustment coefficient of the detail image of the fingerprint contour map after the guide filtering processing,
Figure 100002_DEST_PATH_IMAGE065
and the specific numerical root can be set by running according to the quality of the collected finger vein image. Referring to fig. 1 and 2, the texture of the fingerprint texture part after enhancement is clearly clearer than the texture of the fingerprint texture part before enhancement.
S4, finding fingerprint finger vein intersections in the finger vein contour map, and repairing finger vein textures based on the fingerprint finger vein intersections in a specific mode: finding fingerprint finger vein crossing points in a finger vein contour map, wherein the sum of absolute values of pixel differences of all two adjacent pixel points in eight neighborhoods of the fingerprint finger vein crossing points is 6 x 255, and repairing finger vein textures based on the fingerprint finger vein crossing points, and the steps are as follows:
s4.1, dividing the finger vein contour map into a plurality of block areas, and extracting a direction field of the block areas, wherein the specific steps are as follows:
s4.1.1, estimating the direction of each pixel point i in the finger vein contour map, comprising the following steps: estimating the direction of each pixel point i in the finger vein contour map, comprising the following steps: placing the pixel point in
Figure 283673DEST_PATH_IMAGE066
In the neighborhood window, the trend of the pixel point is divided into 8 directions, and the included angle between two adjacent directions is
Figure 945730DEST_PATH_IMAGE024
Separately calculate said
Figure 100002_DEST_PATH_IMAGE067
Dividing the gray average values into four groups according to the direction perpendicular to each other, calculating the difference value of the two gray average values in each group, wherein the two directions with the largest absolute value of the difference value are the prediction directions of the pixel point, respectively calculating the difference value of the gray average values in the two prediction directions and the gray value of the pixel point, and taking the direction with the smaller absolute value of the difference value as the final direction field of the pixel point
Figure 38188DEST_PATH_IMAGE025
；
S4.1.2 divide the contour map of finger vein into
Figure 85779DEST_PATH_IMAGE068
The block area of the size is calculated according to the step S4.1, and the direction fields of all pixel points in the block area are calculated according to the direction fields
Figure 208587DEST_PATH_IMAGE025
The direction with the highest ratio is taken as the direction of the block area
Figure 918792DEST_PATH_IMAGE027
；
S4.2, traversing and searching fingerprint finger vein intersections in the finger vein contour map, and calculating the directions of the fingerprint finger vein intersections based on the vein direction fields of the fingerprint finger vein intersections and the direction fields of the block areas where the fingerprint finger vein intersections are located;
the calculation formula of the vein direction field of the fingerprint finger vein intersection point is as follows:
Figure 100002_DEST_PATH_IMAGE069
in the formula, the first step is that,
Figure 847564DEST_PATH_IMAGE070
representing the magnitude of the field in the direction of the intersection of the finger veins,
Figure 100002_DEST_PATH_IMAGE071
the coordinates of the point are represented by,
Figure 55386DEST_PATH_IMAGE031
representing the derivative of the fingerprint finger vein intersection in the x-direction,
Figure 598363DEST_PATH_IMAGE072
indicating finger vein intersection in fingerprint
Figure 703854DEST_PATH_IMAGE033
A derivative of the direction;
the calculation formula of the direction of the fingerprint finger vein intersection point is as follows:
Figure 100002_DEST_PATH_IMAGE073
in the formula, the first step is that,
Figure 200432DEST_PATH_IMAGE074
the direction of the intersection of the finger veins is the fingerprint,
Figure 100002_DEST_PATH_IMAGE075
and
Figure 770085DEST_PATH_IMAGE076
respectively the direction of the block area
Figure 687225DEST_PATH_IMAGE038
Vein direction field of intersection point of finger vein and fingerprint
Figure 100002_DEST_PATH_IMAGE077
A weighting coefficient of
Figure 106443DEST_PATH_IMAGE078
；
S4.3, finger vein textures are repaired based on the direction of the fingerprint finger vein intersection, and the method specifically comprises the following steps: the method comprises the following specific steps:
s4.3.1 setting a matching threshold
Figure 100002_DEST_PATH_IMAGE079
；
S4.3.2. Calculating the intersection point of any finger vein on one side of the edge of the finger vein contour line
Figure 377018DEST_PATH_IMAGE042
And any fingerprint finger vein intersection point on the other side of the edge of the finger vein contour line
Figure 502975DEST_PATH_IMAGE080
Difference in direction of
Figure 100002_DEST_PATH_IMAGE081
When is coming into contact with
Figure 466383DEST_PATH_IMAGE082
Then, the two points are judged to be matched, and the initial matching threshold value is
Figure 100002_DEST_PATH_IMAGE083
The setting is as small as possible, for example, the setting is started from 1 degree, and the direction difference between the two finger vein intersections is smaller as the two finger vein intersections are matched;
s4.3.3. If the number of the finger vein intersections that can not be successfully matched in any intersection point on one side of the edge of the finger vein contour line is greater than 30% of the total number of the finger vein intersections on the corresponding side, it is determined that the matching set in step S5.1 is successfulMatching threshold value
Figure 441073DEST_PATH_IMAGE079
Invalid; otherwise, if the number of the fingerprint finger vein intersections which are not successfully matched in any side intersection of the edge of the finger vein contour line is less than or equal to 30% of the total number of the fingerprint finger vein intersections on the corresponding side, the matching threshold set in the step S5.1 is determined
Figure 187443DEST_PATH_IMAGE079
Valid, record the matching threshold
Figure 918639DEST_PATH_IMAGE079
The lower fingerprint refers to the group number of vein intersections successfully matched;
s4.3.4. Gradually increasing the matching threshold value according to 1 degree of increase every time
Figure 754746DEST_PATH_IMAGE079
Until the threshold is matched, steps S5.2 and S5.3 are repeated
Figure 40233DEST_PATH_IMAGE079
Increasing the matching threshold to 30 degrees, counting the effective matching threshold and the corresponding number of the matched groups, and taking the matching threshold with the largest number of the matched groups as a reference matching threshold;
s4.3.5, fingerprint finger vein cross points matched with each other under a reference matching threshold value are connected, and finger vein textures are simulated.
And S5, fusing the fingerprint contour map after the enhancement treatment and the repaired finger vein texture to form a fingerprint image to be identified, and finishing identity identification based on the fingerprint image to be identified.
In order to verify that the fingerprint image acquired by the method has identity identifiability, the size normalization of the acquired fingerprint image is performed, a common fingerprint chip is used for testing, the fingerprint texture definition is improved by more than 50% through the experimental comparison of single features and multiple features, the fingerprint texture detail feature is improved by more than 50%, the stability of the fingerprint texture is improved by more than 50%, the noise of the fingerprint texture image is reduced by more than 50%, and the algorithm time is reduced by more than 50%.
On one hand, false body attack detection can be carried out through whether the fingerprint image is obtained or not, and if the fingerprint image cannot be extracted from the collected finger vein image, false body attack can be regarded as not to be registered or verified; on the other hand, the fingerprint finger vein multi-mode identity authentication can be performed, and the accuracy of vein identification is improved under the condition that the equipment cost is not increased.
The present invention has been described in detail with reference to the embodiments, but the description is only for the preferred embodiments of the present invention and should not be construed as limiting the scope of the present invention. All equivalent changes and modifications made within the scope of the present invention shall fall within the scope of the present invention.

Claims (9)

1. An identification method for converting a vein image into a fingerprint image, characterized by: which comprises the following steps:
s1, collecting a finger vein image, wherein fingerprint textures exist in the finger vein image, and sequentially carrying out binarization and edge detection on the finger vein image to obtain a finger vein image contour map containing the fingerprint textures;
s2, dividing the finger vein image contour map into a fingerprint contour map and a finger vein contour map based on a linear regression method, defining the fingerprint texture of the corresponding position covered by the finger vein in the finger vein contour map as the finger vein texture,
the specific steps of dividing the finger vein image contour map into a fingerprint contour map and a finger vein contour map based on a linear regression method are as follows:
s2.1, respectively calculating the curvature of each contour point on each contour line in the finger vein image contour map, wherein the calculation formula is as follows:
Figure DEST_PATH_IMAGE002
in the formula, the content of the active carbon is shown in the specification,
Figure DEST_PATH_IMAGE004
is a contour point at
Figure DEST_PATH_IMAGE006
The first derivative of the direction is,
Figure DEST_PATH_IMAGE008
is a contour point at
Figure DEST_PATH_IMAGE010
The first derivative of the direction of the light beam,
Figure DEST_PATH_IMAGE012
is a contour point at
Figure 883038DEST_PATH_IMAGE006
The second derivative of the direction of the light,
Figure DEST_PATH_IMAGE014
is a contour point at
Figure DEST_PATH_IMAGE015
The second derivative of the direction;
s2.2, counting that the curvature value on each contour line in the finger vein image contour map is lower than that of each contour line
Figure DEST_PATH_IMAGE017
The number of contour points of (a) accounts for the total number of contour points of the contour line
Figure DEST_PATH_IMAGE019
(ii) a When ratio of
Figure DEST_PATH_IMAGE020
Greater than a set proportional threshold
Figure DEST_PATH_IMAGE022
If so, judging that the contour line belongs to the fingerprint contour map; when ratio of
Figure 49446DEST_PATH_IMAGE020
Less than a set proportional threshold
Figure 688237DEST_PATH_IMAGE022
If so, judging that the contour line belongs to the finger vein contour map;
s3, enhancing the fingerprint contour map by adopting a fingerprint enhancement algorithm;
s4, finding fingerprint finger vein intersections in the finger vein contour map, and repairing finger vein textures based on the fingerprint finger vein intersections;
and S5, fusing the fingerprint contour map after the enhancement treatment and the repaired finger vein texture to form a fingerprint image to be identified, and finishing identity identification based on the fingerprint image to be identified.
2. An identification method for converting a vein image into a fingerprint image according to claim 1, wherein: the step S1 further comprises thinning the finger vein image contour map, so that contour lines in the finger vein image contour map are all single-pixel curves.
3. An identification method for converting a vein image into a fingerprint image according to claim 1, wherein: in the step S3, a composite enhancement algorithm combining gaussian filtering, mean filtering and guided filtering is adopted for enhancing the fingerprint profile, and the method specifically includes the following steps:
s3.1, smoothing the fingerprint contour map divided in the step S2 by adopting Gaussian filtering;
s3.2, smoothing the fingerprint contour map subjected to Gaussian filtering by adopting mean filtering;
s3.3, smoothing the fingerprint contour map divided in the step S2 by adopting guide filtering;
and S3.4, respectively obtaining detail images after different filtering processes based on the images obtained in the steps S3.1-S3.3, and enhancing the fingerprint outline image through the detail images.
4. A recognition method for converting a vein image into a fingerprint image according to claim 3, characterized in that: in the step S3.4, enhancing the fingerprint contour map through the detail image is represented as follows:
Figure DEST_PATH_IMAGE024
in the formula, the first step is that,
Figure DEST_PATH_IMAGE026
for the enhanced image of the final fingerprint profile,
Figure DEST_PATH_IMAGE028
for the fingerprint profile map divided in step S2,
Figure DEST_PATH_IMAGE030
for the image processed in step S3.1,
Figure DEST_PATH_IMAGE032
for the image processed in step S3.2,
Figure DEST_PATH_IMAGE034
for the image processed in step S3.3,
Figure DEST_PATH_IMAGE036
the adjustment coefficient of the detail image of the fingerprint contour map after Gaussian filtering processing,
Figure DEST_PATH_IMAGE038
the adjustment coefficient of the detail image of the fingerprint contour map after the mean value filtering processing,
Figure DEST_PATH_IMAGE040
the adjustment coefficient of the detail image of the fingerprint contour map after the guide filtering processing,
Figure DEST_PATH_IMAGE042
。
5. an identification method for converting a vein image into a fingerprint image according to claim 1, wherein: in the step S4, the sum of absolute values of pixel differences between all two adjacent pixel points in the eight neighborhoods of the finger vein intersection is 6 × 255, and the specific step of the step S4 includes:
s4.1, dividing the finger vein contour map into a plurality of block areas, and extracting a direction field of each block area;
s4.2, traversing and searching fingerprint finger vein intersections in the finger vein contour map, and calculating the directions of the fingerprint finger vein intersections based on the vein direction fields of the fingerprint finger vein intersections and the direction fields of the block areas where the fingerprint finger vein intersections are located;
and S4.3, repairing the vein texture based on the direction of the fingerprint vein intersection.
6. An identification method for converting a vein image into a fingerprint image according to claim 5, wherein: the specific steps in step S4.1 include:
s4.1.1, estimating the direction of each pixel point i in the finger vein contour map, comprising the following steps: placing the pixel point in
Figure DEST_PATH_IMAGE044
In the neighborhood window, the trend of the pixel point is divided into 8 directions, and the included angle between two adjacent directions is
Figure DEST_PATH_IMAGE046
Separately calculating said
Figure 893566DEST_PATH_IMAGE044
Dividing the gray level average values in 8 directions in the neighborhood window into four groups according to the direction vertical to each other, calculating the difference value of the two gray level average values in each group, wherein the two directions with the largest difference absolute value are the prediction directions of the pixel point, and respectively calculatingThe difference value between the gray average value in the two prediction directions and the gray value of the pixel point is taken as the final direction field of the pixel point in the direction with smaller absolute value of the difference value
Figure DEST_PATH_IMAGE048
；
S4.1.2, dividing the finger vein contour map into
Figure DEST_PATH_IMAGE050
The block area of the size is calculated according to the step S4.1, and the direction fields of all pixel points in the block area are calculated according to the direction fields
Figure DEST_PATH_IMAGE051
The direction with the highest ratio is taken as the direction of the block area
Figure DEST_PATH_IMAGE053
。
7. An identification method for converting a vein image into a fingerprint image according to claim 6, wherein: in step S4.2, the calculation formula of the vein direction field of the fingerprint finger vein intersection is:
Figure DEST_PATH_IMAGE055
in the formula, the first step is that,
Figure DEST_PATH_IMAGE057
representing the magnitude of the field in the direction of the intersection of the finger veins,
Figure DEST_PATH_IMAGE059
the coordinates of the point are represented by,
Figure DEST_PATH_IMAGE061
representing the derivative of the fingerprint finger vein intersection in the x-direction,
Figure DEST_PATH_IMAGE063
indicating finger vein intersection in fingerprint
Figure DEST_PATH_IMAGE065
The derivative of the direction.
8. An identification method for converting a vein image into a fingerprint image according to claim 7, wherein: in step S4.2, the calculation formula of the direction of the fingerprint finger vein intersection is:
Figure DEST_PATH_IMAGE067
in the formula, the first step is that,
Figure DEST_PATH_IMAGE069
the direction of the intersection of the finger veins is the fingerprint,
Figure DEST_PATH_IMAGE071
and
Figure DEST_PATH_IMAGE073
respectively the direction of the block area
Figure DEST_PATH_IMAGE075
Vein direction field of intersection point of finger vein and fingerprint
Figure DEST_PATH_IMAGE077
A weighting coefficient of
Figure DEST_PATH_IMAGE079
。
9. An identification method for converting a vein image into a fingerprint image according to claim 8, wherein: the specific steps in step S4.3 are:
s4.3.1. Setting a matching thresholdValue of
Figure DEST_PATH_IMAGE081
；
S4.3.2, calculating any fingerprint finger vein intersection point on one side of edge of finger vein contour line
Figure DEST_PATH_IMAGE083
And any fingerprint finger vein intersection point on the other side of the edge of the finger vein contour line
Figure DEST_PATH_IMAGE085
Difference in direction of
Figure DEST_PATH_IMAGE087
When is coming into contact with
Figure DEST_PATH_IMAGE089
Judging the matching of the two points;
s4.3.3, if the number of the fingerprint finger vein intersections which can not be successfully matched in any intersection point on one side of the edge of the finger vein contour line is more than 30 percent of the total number of the fingerprint finger vein intersections on the corresponding side, judging that the matching threshold value set in the step S5.1 is matched
Figure 190162DEST_PATH_IMAGE081
Invalid; otherwise, if the number of the fingerprint finger vein intersections which are not successfully matched in any side intersection of the edge of the finger vein contour line is less than or equal to 30% of the total number of the fingerprint finger vein intersections on the corresponding side, the matching threshold set in the step S5.1 is determined
Figure 747045DEST_PATH_IMAGE081
Valid, record the matching threshold
Figure 607554DEST_PATH_IMAGE081
The lower fingerprint refers to the group number of vein intersections successfully matched;
s4.3.4. Increasing the matching threshold gradually
Figure 275296DEST_PATH_IMAGE081
Until the threshold is matched, steps S5.2 and S5.3 are repeated
Figure 288382DEST_PATH_IMAGE081
Increasing the matching threshold to 30 degrees, counting the effective matching threshold and the corresponding number of the matched groups, and taking the matching threshold with the largest number of the matched groups as a reference matching threshold;
and S4.3.5, connecting the fingerprint finger vein intersections matched with each other under the reference matching threshold value to simulate the finger vein texture.
CN202211140193.1A 2022-09-20 2022-09-20 Identification method for converting vein image into fingerprint image Active CN115223211B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202211140193.1A CN115223211B (en) 2022-09-20 2022-09-20 Identification method for converting vein image into fingerprint image

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202211140193.1A CN115223211B (en) 2022-09-20 2022-09-20 Identification method for converting vein image into fingerprint image

Publications (2)

Publication Number Publication Date
CN115223211A CN115223211A (en) 2022-10-21
CN115223211B true CN115223211B (en) 2022-12-02

Family

ID=83617006

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202211140193.1A Active CN115223211B (en) 2022-09-20 2022-09-20 Identification method for converting vein image into fingerprint image

Country Status (1)

Country Link
CN (1) CN115223211B (en)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116778172B (en) * 2023-08-18 2023-11-07 江苏圣点世纪科技有限公司 Finger back vein image enhancement method
CN116993960B (en) * 2023-09-28 2024-02-09 江苏圣点世纪科技有限公司 Method for removing fingerprints in finger vein image

Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101093539A (en) * 2007-07-27 2007-12-26 哈尔滨工程大学 Matching identification method by extracting characters of vein from finger

Family Cites Families (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20040033495A1 (en) * 2001-08-03 2004-02-19 Eos Biotechnology, Inc. Methods of diagnosis of angiogenesis, compositions and methods of screening for angiogenesis modulators
CN101840511B (en) * 2010-06-04 2013-08-28 哈尔滨工程大学 Method for extracting, matching and recognizing characteristics of finger veins
US10296734B2 (en) * 2015-01-27 2019-05-21 Idx Technologies Inc. One touch two factor biometric system and method for identification of a user utilizing a portion of the person's fingerprint and a vein map of the sub-surface of the finger
CN108319887A (en) * 2017-01-18 2018-07-24 中国移动通信有限公司研究院 A kind of identity authentication method and system
CN108520211A (en) * 2018-03-26 2018-09-11 天津大学 The extracting method of finger venous image feature based on finger folding line
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
CN109598247B (en) * 2018-12-07 2022-09-06 黑龙江大学 Two-dimensional code identity authentication method based on vein image detail point and grain characteristics
CN111639560B (en) * 2020-05-15 2023-05-30 圣点世纪科技股份有限公司 Finger vein feature extraction method and device based on dynamic fusion of vein skeleton line and topographic concave-convex characteristics

Patent Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101093539A (en) * 2007-07-27 2007-12-26 哈尔滨工程大学 Matching identification method by extracting characters of vein from finger

Also Published As

Publication number Publication date
CN115223211A (en) 2022-10-21

Similar Documents

Publication Publication Date Title
CN115223211B (en) Identification method for converting vein image into fingerprint image
CN108009520B (en) Finger vein identification method and system based on convolution variational self-encoder network
US9633269B2 (en) Image-based liveness detection for ultrasonic fingerprints
Rattani et al. Feature level fusion of face and fingerprint biometrics
WO2017059591A1 (en) Finger vein identification method and device
CN102043961B (en) Vein feature extraction method and method for carrying out identity authentication by utilizing double finger veins and finger-shape features
CN105956578A (en) Face verification method based on identity document information
CN107729820B (en) Finger vein identification method based on multi-scale HOG
CN104239769A (en) Identity recognition method and system based on finger vein characteristics
CN111507206B (en) Finger vein identification method based on multi-scale local feature fusion
Sharma et al. Two-stage quality adaptive fingerprint image enhancement using Fuzzy C-means clustering based fingerprint quality analysis
CN106096569A (en) A kind of finger vein identification method
Kim et al. New enhancement algorithm for fingerprint images
CN110348289A (en) A kind of finger vein identification method based on binary map
Cho et al. Core-based fingerprint image classification
CN111223063A (en) Finger vein image NLM denoising method based on texture features and binuclear function
CN110163123B (en) Fingerprint finger vein fusion identification method based on single near-infrared finger image
CN105373781A (en) Binary image processing method for identity authentication
Lin et al. Feature level fusion of fingerprint and finger vein biometrics
CN116823820B (en) Aluminum-carbon integral stopper rod contour bending detection method
CN116778537B (en) Nail detection and processing method in dorsal digital vein image recognition process
Rajkumar et al. A secondary fingerprint enhancement and minutiae extraction
CN115311696B (en) Finger area detection method based on vein texture features
Huang et al. Noise removal and impainting model for iris image
Wang et al. A novel finger vein pattern extraction method using oriented filtering technology

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
GR01 Patent grant
GR01 Patent grant