Fingerprint comparison method and apparatus, storage medium, and device

Abstract

The application discloses a fingerprint comparison method and device, a storage medium and equipment, and belongs to the fingerprint identification field. The method comprises the following steps: acquiring a first group of minutiae points of a first fingerprint image and a second group of minutiae points of a second fingerprint image, and obtaining a matching minutiae point pair; selecting one matching minutiae point pair as a reference minutiae point pair, and performing a rigid transformation on all minutiae points in the second group of minutiae points so that the two minutiae points of the reference minutiae point pair coincide; calculating a contribution value of the matching minutiae point pair according to the matching minutiae point pair and corresponding neighborhood minutiae point pairs; wherein the neighborhood minutiae point pairs are all matching minutiae point pairs in a neighborhood range of the matching minutiae point pair; and calculating a matching score according to the contribution values of all matching minutiae point pairs. The application calculates the contribution value according to the matching minutiae point pair and neighborhood minutiae point pairs in the neighborhood of the matching minutiae point pair, improves the accuracy of fingerprint comparison, and reduces the equal error rate and other evaluation indexes.

Description

Technical Field

[0001] This invention relates to the field of fingerprint recognition, and in particular to a fingerprint comparison method, apparatus, storage medium and device. Background Technology

[0002] Fingerprint recognition technology is one of the most mature and widely used technologies in the field of biometric identification. A fingerprint consists of ridges and valleys that intersect on the surface of the finger. Because fingerprint collection is convenient, hardware costs are low, performance is reliable, and it is easy to use, it is now widely used in many areas such as identity verification and access control, including mobile phone unlocking, access control devices, fingerprint collection when applying for ID cards, and fast-track immigration clearance systems.

[0003] Fingerprint recognition technology mainly includes two algorithmic processes: fingerprint feature extraction and fingerprint matching. Among the extracted fingerprint features, the most commonly used and fundamental feature is the minutiae, which includes two types: terminal points and bifurcation points. These two types correspond to the termination of a ridge and the division of a ridge into two ridges, respectively. For example... Figure 1 As shown, the circles represent detail points of the branching point type, and the squares represent detail points of the terminal point type. Each detail point can be represented by type t, coordinates (x, y), and angle. Composed of four-dimensional coordinates To express.

[0004] Currently, minutiae (small holes) are widely recognized as the most discriminative and reliable local features in fingerprints. The local structure of minutiae exhibits excellent adaptability to nonlinear deformation and noise. Minutia-based fingerprint matching algorithms are the most important method in fingerprint matching research. Their core problem is finding matching minutiae pairs and calculating a matching score based on certain rules.

[0005] The minutiae in the first fingerprint image are referred to as the first set of minutiae, and the minutiae in the second fingerprint image are referred to as the second set of minutiae. The main steps of existing minutiae-based fingerprint matching algorithms are as follows: calculating the local descriptor of each minutiae, calculating the local similarity between every two minutiae in the first and second sets of minutiae, finding matching minutiae pairs according to set rules, and finally obtaining a matching score based on the number of these matching minutiae pairs or the deterministic similarity value between them.

[0006] For example, Figure 2 An example of the first and second fingerprint images is given. Figure 2 The image on the left is the first fingerprint image, and the image on the right is the second fingerprint image. Both images are from the same fingerprint. Details of both are as follows... Figure 3 As shown, Figure 3The left side shows the first fingerprint image and its first set of minutiae, while the right side shows the second fingerprint image and its second set of minutiae. Each minutiae is then analyzed as follows: Figure 4 The number shown, Figure 4 The left side of the image shows the first group of minutiae (50 in total), and the right side shows the second group of minutiae (38 in total). The algorithm yielded 36 pairs of matching minutiae pairs, which are marked with lines, as shown below. Figure 5 As shown.

[0007] After finding matching minutiae pairs, the matching score can be calculated based on these pairs. Assuming the first set of minutiae has N1 minutiae, the second set has N2 minutiae, and the number of matching minutiae pairs is N (i.e., there are N pairs of matching minutiae), the formulas for calculating the matching score s in existing technologies generally include the following:

[0008]

[0009]

[0010]

[0011] Some studies have indicated that if there are 13 "completely" matching minutiae pairs, it means that the two fingerprint images being compared originate from the same fingerprint. However, "completely" is a subjective term used by experts and is difficult to express through concrete measurements. The inventors discovered through experiments that two fingerprints from different fingers can also achieve 13 or even 29 matching minutiae pairs, while two fingerprints from the same finger only have 8 matching minutiae pairs due to factors such as small overlap and poor image quality. Therefore, due to the different criteria for judging matching minutiae pairs, simply relying on the number of matching minutiae pairs or their similarity is insufficient to accurately evaluate the similarity between two fingerprint images. Summary of the Invention

[0012] To address the shortcomings of existing technologies, this invention provides a fingerprint comparison method, apparatus, storage medium, and device. This invention calculates the contribution value of matching minutiae pairs and neighboring minutiae pairs within the neighborhood of the matching minutiae pairs, thereby improving the accuracy of fingerprint comparison and reducing evaluation indicators such as the error rate.

[0013] The technical solution provided by this invention is as follows:

[0014] In a first aspect, the present invention provides a fingerprint comparison method, the method comprising:

[0015] Obtain the first set of minutiae from the first fingerprint image and the second set of minutiae from the second fingerprint image, and obtain matching minutiae pairs between the first set of minutiae and the second set of minutiae;

[0016] Select a matching minutiae pair as the reference minutiae pair, and perform a rigid transformation on all minutiae in the second set of minutiae such that the two minutiae of the reference minutiae pair coincide.

[0017] For each matching minutiae pair, the contribution value of the matching minutiae pair is calculated based on the matching minutiae pair and the corresponding neighboring minutiae pair;

[0018] Wherein, the neighborhood minutiae pair is all matching minutiae pairs within the neighborhood range defined by the matching minutiae pair;

[0019] The matching score between the first fingerprint image and the second fingerprint image is calculated based on the contribution values ​​of all matching minutiae pairs.

[0020] Furthermore, for each matching minutiae pair, calculating the contribution value of the matching minutiae pair based on the matching minutiae pair and the corresponding neighboring minutiae pair includes:

[0021] The first contribution value of the matching minutiae pair is calculated based on the distance and angle difference between the two minutiae points of the matching minutiae pair.

[0022] For each neighborhood minutiae pair within the defined neighborhood range of the matching minutiae pair, a rigid transformation is performed on the second set of minutiae points that makes the two minutiae points of the matching minutiae pair coincide.

[0023] For each neighborhood minutiae pair, calculate the second contribution value of each neighborhood minutiae pair based on the distance and angle difference between the two minutiae in the neighborhood minutiae pair, as well as the corresponding interior angle value;

[0024] Wherein, the interior angle value is the angle value of the interior angle formed by the two minutiae of the neighborhood minutiae pair and the matching minutiae pair;

[0025] The contribution value of the matching minutiae pair is calculated based on the first contribution value of the matching minutiae pair and the second contribution value of all neighboring minutiae pairs.

[0026] Furthermore, the first contribution value of the matching minutiae pair is calculated using the following formula;

[0027] sd=Z(d i ,μ1,τ1)*Z(Δθ i (μ2, τ2)

[0028] Where sd is the first contribution value of the matched minutiae pair, Z(,,) is the sigmoid function, and d i and Δθ iLet μ1, τ1, μ2, and τ2 be the distance and angle difference between the two minutiae of the i-th matching minutiae pairs, respectively, where i = 1, 2, ..., N, N is the number of matching minutiae pairs, and μ1, τ1, μ2, and τ2 are the set parameters.

[0029] The second contribution value of each neighborhood minutiae pair is calculated using the following formula;

[0030] Among them, sn j Sn represents the second contribution value of the j-th neighborhood minutiae pair, where j = 1, 2, ..., M, and M is the number of neighborhood minutiae pairs within the defined neighborhood range of the matched minutiae pair. j =Z(d) j ,μ1,τ1)*Z(Δθ j ,μ2,τ2)*Z(α j ,μ3,τ3),d j and Δθ j Let α be the distance and angle difference between the two minutiae of the j-th neighborhood minutiae pair. j Let μ3 and τ3 be the angle values ​​of the interior angle formed by the two minutiae of the j-th neighborhood minutiae pair and the matching minutiae pair, where μ3 and τ3 are set parameters.

[0031] The contribution value of the matched minutiae pair is calculated using the following formula;

[0032]

[0033] Among them, pairscore i Let np and nq be the contribution values ​​of the i-th matching minutiae pair, respectively, and let np and nq be the number of the first group of minutiae and the second group of minutiae within the set neighborhood of the matching minutiae pair.

[0034]

[0035] Furthermore, the matching score between the first fingerprint image and the second fingerprint image is calculated using the following formula:

[0036]

[0037] Where s is the matching score between the first fingerprint image and the second fingerprint image, and N1 and N2 are the number of minutiae in the first group and the number of minutiae in the second group, respectively.

[0038] Furthermore, selecting a matching minutiae pair as a reference minutiae pair includes:

[0039] Calculate the centroid of the first set of minutiae for all matching minutiae pairs;

[0040] Find the first set of minutiae closest to the centroid from all matching minutiae pairs, and use the matching minutiae pair to which the found first set of minutiae belongs as the reference minutiae pair.

[0041] Furthermore, the rigid transformation performed on all detail points in the second group, ensuring that the two detail points of the reference detail point pair coincide, includes:

[0042] Calculate the angular difference Δθ and coordinate differences Δx and Δy between the two details of the reference detail pair;

[0043] Δθ=θ2-θ1

[0044]

[0045] Where (x2, y2, θ2) are the coordinates and orientation angles of the first set of detail points of the reference detail point pair, and (x1, y1, θ1) are the coordinates and orientation angles of the second set of detail points of the reference detail point pair;

[0046] Based on the angle difference Δθ and coordinate difference Δx and Δy between the two details of the reference detail point pair, calculate the coordinates and orientation angles of the second set of detail points after rigid transformation;

[0047]

[0048] θ′=θ-Δθ

[0049] Where (x, y, θ) are the coordinates and orientation angles of the second set of detail points before rigid transformation, and (x′, y′, θ′) are the coordinates and orientation angles of the second set of detail points after rigid transformation.

[0050] Furthermore, the method also includes:

[0051] The fingerprint comparison result is obtained based on the matching score between the first fingerprint image and the second fingerprint image.

[0052] In a second aspect, the present invention provides a fingerprint comparison device, the device comprising:

[0053] The data preparation module is used to acquire the first set of minutiae of the first fingerprint image and the second set of minutiae of the second fingerprint image, and to obtain matching minutiae pairs of the first set of minutiae and the second set of minutiae.

[0054] The rigid transformation module is used to select a matching pair of detail points as a reference pair of detail points, and to perform a rigid transformation on all detail points in the second group of detail points so that the two detail points of the reference pair of detail points coincide.

[0055] The contribution value calculation module is used to calculate the contribution value of each matching minutiae pair based on the matching minutiae pair and the corresponding neighboring minutiae pair.

[0056] Wherein, the neighborhood minutiae pair is all matching minutiae pairs within the neighborhood range defined by the matching minutiae pair;

[0057] The matching score calculation module is used to calculate the matching score between the first fingerprint image and the second fingerprint image based on the contribution values ​​of all matching minutiae pairs.

[0058] Furthermore, the contribution value calculation module includes:

[0059] The first calculation unit is used to calculate the first contribution value of the matching minutiae pair based on the distance and angle difference between the two minutiae of the matching minutiae pair;

[0060] A rigid transformation unit is used to perform a rigid transformation on the second set of minutiae of each neighborhood minutiae pair within a defined neighborhood range of the matching minutiae pair, such that the two minutiae of the matching minutiae pair coincide.

[0061] The second calculation unit is used to calculate the second contribution value of each neighborhood minutiae pair based on the distance and angle difference between the two minutiae in the neighborhood minutiae pair and the corresponding interior angle value.

[0062] Wherein, the interior angle value is the angle value of the interior angle formed by the two minutiae of the neighborhood minutiae pair and the matching minutiae pair;

[0063] The third calculation unit is used to calculate the contribution value of the matching minutiae pair based on the first contribution value of the matching minutiae pair and the second contribution values ​​of all neighboring minutiae pairs.

[0064] Furthermore, the first contribution value of the matching minutiae pair is calculated using the following formula;

[0065] sd=Z(d i ,μ1,τ1)*Z(Δθ i (μ2, τ2)

[0066] Where sd is the first contribution value of the matched minutiae pair, Z(,,) is the sigmoid function, and d i and Δθ i Let μ1, τ1, μ2, and τ2 be the distance and angle difference between the two minutiae of the i-th matching minutiae pairs, respectively, where i = 1, 2, ..., N, N is the number of matching minutiae pairs, and μ1, τ1, μ2, and τ2 are the set parameters.

[0067] The second contribution value of each neighborhood minutiae pair is calculated using the following formula;

[0068] Among them, sn j Sn represents the second contribution value of the j-th neighborhood minutiae pair, where j = 1, 2, ..., M, and M is the number of neighborhood minutiae pairs within the defined neighborhood range of the matched minutiae pair. j =Z(d) j ,μ1,τ1)*Z(Δθ j ,μ2,τ2)*Z(α j ,μ3,τ3),d j and Δθ j Let α be the distance and angle difference between the two minutiae of the j-th neighborhood minutiae pair. j Let μ3 and τ3 be the angle values ​​of the interior angle formed by the two minutiae of the j-th neighborhood minutiae pair and the matching minutiae pair, where μ3 and τ3 are set parameters.

[0069] The contribution value of the matched minutiae pair is calculated using the following formula;

[0070]

[0071] Among them, pairscore i Let np and nq be the contribution values ​​of the i-th matching minutiae pair, respectively, and let np and nq be the number of the first group of minutiae and the second group of minutiae within the set neighborhood of the matching minutiae pair.

[0072]

[0073] Furthermore, the matching score between the first fingerprint image and the second fingerprint image is calculated using the following formula:

[0074]

[0075] Where s is the matching score between the first fingerprint image and the second fingerprint image, and N1 and N2 are the number of minutiae in the first group and the number of minutiae in the second group, respectively.

[0076] Furthermore, the rigid transformation module includes:

[0077] The centroid calculation unit is used to calculate the centroid of the first set of minutiae for all matching minutiae pairs.

[0078] The reference detail point pair selection unit is used to find the first group detail point that is closest to the centroid from the first group detail points of all matching detail point pairs, and to take the matching detail point pair to which the found first group detail point belongs as the reference detail point pair;

[0079] The fourth calculation unit is used to calculate the angle difference Δθ and coordinate difference Δx, Δy between the two details of the reference detail point pair;

[0080] Δθ=θ2-θ1

[0081]

[0082] Where (x2, y2, θ2) are the coordinates and orientation angles of the first set of detail points of the reference detail point pair, and (x1, y1, θ1) are the coordinates and orientation angles of the second set of detail points of the reference detail point pair;

[0083] The transformation unit is used to calculate the coordinates and orientation angles of the second set of detail points after rigid transformation based on the angle difference Δθ and coordinate difference Δx, Δy between the two details of the reference detail point pair.

[0084]

[0085] θ′=θ-Δθ

[0086] Where (x, y, θ) are the coordinates and orientation angles of the second set of detail points before rigid transformation, and (x′, y′, θ′) are the coordinates and orientation angles of the second set of detail points after rigid transformation.

[0087] Furthermore, the device also includes:

[0088] The comparison module is used to obtain the fingerprint comparison result based on the matching score between the first fingerprint image and the second fingerprint image.

[0089] Thirdly, the present invention provides a computer-readable storage medium for fingerprint comparison, including a memory for storing processor-executable instructions, which, when executed by the processor, implement the steps of the fingerprint comparison method described in the first aspect.

[0090] Fourthly, the present invention provides an apparatus for fingerprint comparison, comprising at least one processor and a memory storing computer-executable instructions, wherein the processor executes the instructions to implement the steps of the fingerprint comparison method described in the first aspect.

[0091] The present invention has the following beneficial effects:

[0092] When calculating the contribution value of a matching minutiae pair, this invention not only calculates based on the two minutiae of the matching minutiae pair, but also on the neighboring minutiae pairs within the neighborhood of the matching minutiae pair. This fully utilizes the useful information contained in the matching minutiae pair itself and within a certain range around it, effectively evaluating the similarity between two fingerprint images and improving the accuracy of fingerprint comparison. Attached Figure Description

[0093] Figure 1 Example diagram of terminal points and bifurcation points on a fingerprint;

[0094] Figure 2 Here is an example image of the first and second fingerprint images;

[0095] Figure 3 An example image of the first fingerprint image and its first set of minutiae, and a second fingerprint image and its second set of minutiae;

[0096] Figure 4 An example image showing the first and second groups of detail points and their numbering;

[0097] Figure 5 An example diagram for matching lines connecting detail point pairs;

[0098] Figure 6 This is a flowchart of the fingerprint comparison method of the present invention;

[0099] Figure 7 For α j A schematic diagram;

[0100] Figure 8 This is a schematic diagram of the fingerprint comparison device of the present invention. Detailed Implementation

[0101] To make the technical problems, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below in conjunction with the accompanying drawings and specific embodiments. Obviously, the described embodiments are only a part of the embodiments of this invention, and not all of them. The components of the embodiments of this invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without inventive effort are within the scope of protection of this invention.

[0102] Example 1:

[0103] This invention provides a fingerprint comparison method, such as... Figure 6 As shown, the method includes:

[0104] S100: Obtain the first set of minutiae from the first fingerprint image and the second set of minutiae from the second fingerprint image, and obtain matching minutiae pairs of the first set of minutiae and the second set of minutiae.

[0105] The first and second sets of minutiae in this step are obtained by extracting the first and second fingerprint images used for comparison using a fingerprint feature extraction algorithm. This invention does not limit the specific implementation of the fingerprint feature extraction algorithm. Furthermore, this invention does not limit the method of obtaining the matching minutiae pairs of the first and second sets of minutiae; various methods in the prior art can be used.

[0106] For example, the number of minutiae in the first group is N1, the number of minutiae in the second group is N2, and the number of matching minutiae pairs between the first and second groups is N (N <= N1 and N <= N2). These N matching minutiae pairs are represented by (p i q i Let p be an integer, i = 1, 2, ..., N. i Let i = 1, 2, ..., N be the N detail points belonging to the first group of detail points, and q i , i = 1, 2, ..., N are N detail points belonging to the second group of detail points.

[0107] S200: Select a matching detail point pair as the reference detail point pair, and perform a rigid transformation on all detail points in the second set of detail points so that the two detail points of the reference detail point pair coincide.

[0108] The selected reference detail point pair can be denoted as (p a q a ), p a For details belonging to the first group of details, q a These are details belonging to the second group of details. Then q a Perform a rigid transformation so that q a With p a The points overlap, and the other detail points in the second group also undergo the same rigid transformation. The coordinates of the second group of detail points are then converted to the coordinates of the first group of detail points for easier subsequent operations.

[0109] S300: For each matching minutiae pair, calculate the contribution value of the matching minutiae pair based on the matching minutiae pair and the corresponding neighboring minutiae pair.

[0110] Wherein, the neighborhood minutiae pair is all matching minutiae pairs within the neighborhood range defined by the matching minutiae pair.

[0111] This invention improves the method of calculating the contribution value of matching minutiae pairs. When calculating the contribution value of matching minutiae pairs, it calculates not only based on the two minutiae of the matching minutiae pair, but also based on the neighboring minutiae pairs in the neighborhood of the matching minutiae pair. This fully utilizes the useful information contained in the matching minutiae pairs and can improve the accuracy of fingerprint comparison.

[0112] S400: Calculate the matching score between the first fingerprint image and the second fingerprint image based on the contribution values ​​of all matching minutiae pairs.

[0113] The matching score is a statistical value of the contribution of all minutiae pairs. After calculating the matching score, the fingerprint comparison result can be obtained based on the matching score.

[0114] This invention first obtains a first set of minutiae from a first fingerprint image and a second set of minutiae from a second fingerprint image, and obtains matching minutiae pairs between the first and second sets of minutiae. Then, a matching minutiae pair is selected as a reference minutiae pair, and a rigid transformation is performed on all minutiae in the second set of minutiae so that the two minutiae of the reference minutiae pair coincide. For each matching minutiae pair, the contribution value of the matching minutiae pair is calculated based on the two minutiae of the matching minutiae pair and the minutiae of the neighboring minutiae pairs within a defined neighborhood of the matching minutiae pair. Finally, a matching score is calculated based on the contribution values ​​of all minutiae pairs.

[0115] When calculating the contribution value of a matching minutiae pair, this invention not only calculates based on the two minutiae of the matching minutiae pair, but also on the neighboring minutiae pairs within the neighborhood of the matching minutiae pair. This fully utilizes the useful information contained in the matching minutiae pair itself and within a certain range around it, effectively evaluating the similarity between two fingerprint images and improving the accuracy of fingerprint comparison.

[0116] As an improvement to the embodiment of the present invention, the aforementioned S300 includes:

[0117] S310: Calculate the first contribution value of the matching minutiae pair based on the distance and angle difference between the two minutiae points of the matching minutiae pair.

[0118] In one example, the first contribution value of the matching minutiae pair can be calculated using the following formula;

[0119] sd=Z(d i ,μ1,τ1)*Z(Δθ i (μ2, τ2)

[0120] Where sd is the first contribution value of the matched minutiae pair, Z(,,) is the sigmoid function, and d i and Δθ i Each of the i matching minutiae pairs (p) i q i The distance and angle difference between two minutiae, i = 1, 2, ..., N, where N is the number of matching minutiae pairs, and μ1, τ1, μ2, τ2 are set parameters.

[0121] The formula for the sigmoid function is as follows:

[0122]

[0123] Where v is a variable, μ and τ are two parameters, and the sigmoid function takes values ​​in the range [0, 1]. The distance d... i and angle difference Δθ i The μ and τ parameters are passed as variables to the sigmoid function, and a value between 0 and 1 can be calculated based on the set μ and τ parameters to represent a certain similarity. For example, the closer the distance, the larger the value; the smaller the angle difference, the larger the value.

[0124] S320: For the matched minutiae pair (p i q i For each neighborhood minutiae pair within the neighborhood range, a rigid transformation is performed on the second set of minutiae that makes the two minutiae of the matching minutiae pair coincide.

[0125] Assume p i The number of the first set of detail points within a 90-radius neighborhood is np, q i The number of the second set of detail points within a 90-radius neighborhood is nq, p i and q i There are M matching neighborhood minutiae within a radius of 90, where M <= np and M <= nq, because not every minutiae within the neighborhood has a matching pair.

[0126] Let M neighborhood detail point pairs within the neighborhood be denoted as (pNeighbor). j qNeighbor j ), j = 1, 2, ..., M, use a method similar to that in S200 to make a rigid change so that p i and q i Overlapping, and simultaneously, the qNeighbor pairs of M neighborhood detail points within the neighborhood range. j Perform the same rigid transformation.

[0127] S330: For each neighborhood minutiae pair, calculate the second contribution value of each neighborhood minutiae pair based on the distance and angle difference between the two minutiae in the neighborhood minutiae pair, as well as the corresponding interior angle value.

[0128] Wherein, the interior angle value is the angle value of the interior angle formed by the two minutiae of the neighborhood minutiae pair and the matching minutiae pair.

[0129] In one example, the second contribution value of each neighborhood minutiae pair can be calculated using the following formula;

[0130] Among them, sn jFor the j-th neighborhood detail pair (pNeighbor) j qNeighbor j The second contribution value, j = 1, 2, ..., M, where M is the number of neighborhood minutiae pairs within the defined neighborhood range of the matched minutiae pair, and sn j =Z(d) j ,μ1,τ1)*Z(Δθ j ,μ2,τ2)*Z(α j ,μ3,τ3),d j and Δθ j These are the j-th neighborhood detail pairs (pNeighbor) j qNeighbor j The distance and angle difference between two detail points, α j For the j-th neighborhood detail pair (pNeighbor) j qNeighbor j The two minutiae of the point and the matching minutiae pair (p) i q i The angle value of the interior angle formed by α j like Figure 7 As shown, μ3 and τ3 are the set parameters.

[0131] S340: Calculate the contribution value of the matching minutiae pair based on the first contribution value of the matching minutiae pair and the second contribution values ​​of all neighboring minutiae pairs.

[0132] In one example, the contribution value of the matching minutiae pair can be calculated using the following formula;

[0133]

[0134] Among them, pairscore i Let np and nq be the contribution values ​​of the i-th matching minutiae pair, respectively, and let np and nq be the number of the first group of minutiae and the second group of minutiae within the set neighborhood of the matching minutiae pair.

[0135]

[0136] This invention, when calculating the contribution value of a matching minutiae pair, considers not only the distance and angle difference between the two minutiae within the matching minutiae pair itself, but also the distance and angle difference between the two minutiae of neighboring minutiae pairs within a certain neighborhood of the matching minutiae pair, as well as the interior angle value formed with the matching minutiae pair. This fully utilizes the distance and angle information of the matching minutiae pair itself and surrounding matching minutiae pairs within a certain range, thus improving the accuracy of fingerprint comparison.

[0137] Then, the matching score between the first fingerprint image and the second fingerprint image is calculated using the following formula:

[0138]

[0139] Where s is the matching score between the first fingerprint image and the second fingerprint image, and N1 and N2 are the number of minutiae in the first group and the number of minutiae in the second group, respectively.

[0140] This invention can select reference detail point pairs through various methods. In one example, the method for selecting reference detail point pairs includes:

[0141] S210: Calculate the centroid of the first set of minutiae for all matching minutiae pairs.

[0142] Matching details are (p i q i ), i = 1, 2, ..., N, where p i Let i = 1, 2, ..., N be the N minutiae belonging to the first group of minutiae, which is the first group of minutiae among all matching minutiae pairs. Then the coordinates (xc, yc) of its centroid C are:

[0143]

[0144] S220: Find the first set of minutiae closest to the centroid from all matching minutiae pairs, and use the matching minutiae pair to which the found first set of minutiae belongs as the reference minutiae pair.

[0145] At the above N detail points p i In this context, assume that the detail point closest to the centroid C at a distance (e.g., Euclidean distance) is p. a And its corresponding matching minutiae pair is (p a q a ), will (p a q a () as a reference detail point pair.

[0146] Once the reference detail point pairs are obtained, a rigid transformation can be performed. This invention does not limit the specific implementation of the rigid transformation; one example includes:

[0147] S230: Calculate the angular difference Δθ and coordinate difference Δx, Δy between the two detail points of the reference detail point pair;

[0148] Δθ=θ2-θ1

[0149]

[0150] Where (x2, y2, θ2) are the coordinates and orientation angles of the first set of detail points in the reference detail point pair, and (x1, y1, θ1) are the coordinates and orientation angles of the second set of detail points in the reference detail point pair.

[0151] Assuming a reference detail pair (p) a q a ) in q a The coordinates and orientation angles are (x2, y2, θ2), p a Given the coordinates and orientation angle (x1, y1, θ1), calculate q. a and p a The angle difference Δθ and the coordinate difference Δx, Δy.

[0152] S240: Calculate the coordinates and orientation angles of the second set of detail points after rigid transformation based on the angle difference Δθ and coordinate difference Δx and Δy between the two detail points of the reference detail point pair.

[0153]

[0154] θ′=θ-Δθ

[0155] Where (x, y, θ) are the coordinates and orientation angles of the second set of detail points before rigid transformation, and (x′, y′, θ′) are the coordinates and orientation angles of the second set of detail points after rigid transformation.

[0156] Make q a and p a The coincident rigid transformations consist of a rotation and translation operation T, and the angles of the second set of detail points are all reduced by Δθ.

[0157] To more clearly illustrate the effects of the present invention, the following tests were conducted: A fingerprint test set was constructed and tested entirely according to the existing fingerprint matching algorithm based on minutiae. The error rate was 5.5%, while the error rate of the fingerprint matching method of the present invention was 3.5%. Compared with the existing technology, the error rate was reduced by about 36.3%, and the results showed a significant improvement.

[0158] In summary, this invention optimizes the method for calculating fingerprint matching scores. When calculating the contribution value of matching minutiae pairs, it considers not only information such as the distance and angle difference between the two minutiae within the matching minutiae pair itself, but also information such as the distance and angle difference between the two minutiae of neighboring minutiae pairs within a certain neighborhood of the matching minutiae pair, as well as the interior angle value formed with the matching minutiae pair. This fully utilizes the distance and angle information contained in the matching minutiae pair itself and surrounding matching minutiae pairs within a certain range, which is helpful for fingerprint matching, thus improving the accuracy of fingerprint matching. The test results on the test set show a significant improvement compared to existing technologies, increasing the accuracy of fingerprint matching and reducing evaluation indicators such as the equal error rate.

[0159] Example 2:

[0160] This invention provides a fingerprint comparison device, such as... Figure 8 As shown, the device includes:

[0161] Data preparation module 1 is used to acquire the first set of minutiae of the first fingerprint image and the second set of minutiae of the second fingerprint image, and to obtain matching minutiae pairs of the first set of minutiae and the second set of minutiae.

[0162] The rigid transformation module 2 is used to select a matching detail point pair as a reference detail point pair, and to perform a rigid transformation on all detail points in the second set of detail points so that the two detail points of the reference detail point pair coincide.

[0163] The contribution value calculation module 3 is used to calculate the contribution value of each matching minutiae pair based on the matching minutiae pair and the corresponding neighboring minutiae pair.

[0164] Wherein, the neighborhood minutiae pair is all matching minutiae pairs within the neighborhood range defined by the matching minutiae pair.

[0165] The matching score calculation module 4 is used to calculate the matching score between the first fingerprint image and the second fingerprint image based on the contribution values ​​of all matching minutiae pairs.

[0166] When calculating the contribution value of a matching minutiae pair, this invention not only calculates based on the two minutiae of the matching minutiae pair, but also on the neighboring minutiae pairs within the neighborhood of the matching minutiae pair. This fully utilizes the useful information contained in the matching minutiae pair itself and within a certain range around it, effectively evaluating the similarity between two fingerprint images and improving the accuracy of fingerprint comparison.

[0167] As an improvement to this embodiment of the invention, the contribution value calculation module includes:

[0168] The first calculation unit is used to calculate the first contribution value of the matching minutiae pair based on the distance and angle difference between the two minutiae points of the matching minutiae pair.

[0169] A rigid transformation unit is used to perform a rigid transformation on the second set of minutiae of each neighborhood minutiae pair within a defined neighborhood range of the matching minutiae pair, such that the two minutiae of the matching minutiae pair coincide.

[0170] The second calculation unit is used to calculate the second contribution value of each neighborhood minutiae pair based on the distance and angle difference between the two minutiae in the neighborhood minutiae pair and the corresponding interior angle value.

[0171] Wherein, the interior angle value is the angle value of the interior angle formed by the two minutiae of the neighborhood minutiae pair and the matching minutiae pair.

[0172] The third calculation unit is used to calculate the contribution value of the matching minutiae pair based on the first contribution value of the matching minutiae pair and the second contribution values ​​of all neighboring minutiae pairs.

[0173] Furthermore, the first contribution value of the matching minutiae pair can be calculated using the following formula;

[0174] sd=Z(d i ,μ1,τ1)*Z(Δθ i (μ2, τ2)

[0175] Where sd is the first contribution value of the matched minutiae pair, Z(,,) is the sigmoid function, and d i and Δθ i Let μ1, τ1, μ2, and τ2 be the distance and angle difference between the two minutiae of the i-th matching minutiae pairs, respectively, where i = 1, 2, ..., N, N is the number of matching minutiae pairs, and μ1, τ1, μ2, and τ2 are the set parameters.

[0176] The second contribution value of each neighborhood minutiae pair is calculated using the following formula;

[0177] Among them, sn j Sn represents the second contribution value of the j-th neighborhood minutiae pair, where j = 1, 2, ..., M, and M is the number of neighborhood minutiae pairs within the defined neighborhood range of the matched minutiae pair. j =Z(d) j ,μ1,τ1)*Z(Δθ j ,μ2,τ2)*Z(α j ,μ3,τ3),d j and Δθ j Let α be the distance and angle difference between the two minutiae of the j-th neighborhood minutiae pair. j Let μ3 and τ3 be the angle values ​​of the interior angle formed by the two minutiae of the j-th neighborhood minutiae pair and the matching minutiae pair, where μ3 and τ3 are set parameters.

[0178] The contribution value of the matched minutiae pair is calculated using the following formula;

[0179]

[0180] Among them, pairscore i Let np and nq be the contribution values ​​of the i-th matching minutiae pair, respectively, and let np and nq be the number of the first group of minutiae and the second group of minutiae within the set neighborhood of the matching minutiae pair.

[0181]

[0182] The present invention can calculate the matching score between the first fingerprint image and the second fingerprint image using the following formula:

[0183]

[0184] Where s is the matching score between the first fingerprint image and the second fingerprint image, and N1 and N2 are the number of minutiae in the first group and the number of minutiae in the second group, respectively.

[0185] The aforementioned rigid transformation module includes:

[0186] The centroid calculation unit is used to calculate the centroid of the first set of minutiae for all matching minutiae pairs.

[0187] The reference detail point pair selection unit is used to find the first group detail point that is closest to the centroid from the first group detail points of all matching detail point pairs, and to take the matching detail point pair to which the found first group detail point belongs as the reference detail point pair.

[0188] The fourth calculation unit is used to calculate the angle difference Δθ and coordinate difference Δx, Δy between the two details of the reference detail point pair.

[0189] Δθ=θ2-θ1

[0190]

[0191] Where (x2, y2, θ2) are the coordinates and orientation angles of the first set of detail points of the reference detail point pair, and (x1, y1, θ1) are the coordinates and orientation angles of the second set of detail points of the reference detail point pair.

[0192] The transformation unit is used to calculate the coordinates and orientation angles of the second set of detail points after rigid transformation based on the angle difference Δθ and coordinate difference Δx, Δy between the two details of the reference detail point pair.

[0193]

[0194] θ′=θ-Δθ

[0195] Where (x, y, θ) are the coordinates and orientation angles of the second set of detail points before rigid transformation, and (x′, y′, θ′) are the coordinates and orientation angles of the second set of detail points after rigid transformation.

[0196] The apparatus of the present invention may further include:

[0197] The comparison module is used to obtain the fingerprint comparison result based on the matching score between the first fingerprint image and the second fingerprint image.

[0198] The device provided in this embodiment of the invention has the same implementation principle and technical effects as the aforementioned method embodiment 1. For the sake of brevity, any parts not mentioned in this device embodiment can be referred to the corresponding content in the aforementioned method embodiment 1. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the aforementioned device and unit can all be referred to the corresponding processes in the aforementioned method embodiment 1, and will not be repeated here.

[0199] Example 3:

[0200] The method described in Embodiment 1 of this invention can implement business logic through a computer program and record it on a storage medium. This storage medium can be read and executed by a computer, achieving the effects of the scheme described in Embodiment 1 of this specification. Therefore, this invention also provides a computer-readable storage medium for fingerprint comparison, including a memory for storing processor-executable instructions. When executed by a processor, the instructions implement the steps of the fingerprint comparison method of Embodiment 1.

[0201] When calculating the contribution value of a matching minutiae pair, this invention not only calculates based on the two minutiae of the matching minutiae pair, but also on the neighboring minutiae pairs within the neighborhood of the matching minutiae pair. This fully utilizes the useful information contained in the matching minutiae pair itself and within a certain range around it, effectively evaluating the similarity between two fingerprint images and improving the accuracy of fingerprint comparison.

[0202] The storage medium may include a physical device for storing information, typically digitizing the information and then storing it using electrical, magnetic, or optical methods. The storage medium may include: devices that store information using electrical energy, such as various types of memory, like RAM and ROM; devices that store information using magnetic energy, such as hard disks, floppy disks, magnetic tapes, magnetic core memory, bubble memory, and USB flash drives; and devices that store information using optical methods, such as CDs or DVDs. Of course, there are other readable storage media, such as quantum memories and graphene memories.

[0203] The storage medium described above may also include other implementation methods according to the description of method embodiment 1. The implementation principle and technical effects of this embodiment are the same as those of the aforementioned method embodiment 1. For details, please refer to the description of the relevant method embodiment 1, which will not be repeated here.

[0204] Example 4:

[0205] The present invention also provides a device for fingerprint comparison. The device may be a standalone computer, or it may include an actual operating device that uses one or more of the methods or embodiments described in this specification. The fingerprint comparison device may include at least one processor and a memory storing computer-executable instructions. When the processor executes the instructions, it implements the steps of any one or more of the fingerprint comparison methods described in Embodiment 1.

[0206] When calculating the contribution value of a matching minutiae pair, this invention not only calculates based on the two minutiae of the matching minutiae pair, but also on the neighboring minutiae pairs within the neighborhood of the matching minutiae pair. This fully utilizes the useful information contained in the matching minutiae pair itself and within a certain range around it, effectively evaluating the similarity between two fingerprint images and improving the accuracy of fingerprint comparison.

[0207] The device described above may include other implementation methods according to the description of method embodiment 1. The implementation principle and technical effects of this embodiment are the same as those of the aforementioned method embodiment 1. For details, please refer to the description of the relevant method embodiment 1, which will not be repeated here.

[0208] Finally, it should be noted that the above-described embodiments are merely specific implementations of the present invention, used to illustrate the technical solutions of the present invention, and not to limit it. The scope of protection of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments within the scope of the technology disclosed in the present invention, or make equivalent substitutions for some of the technical features; and these modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention. All should be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A fingerprint comparison method, characterized in that, The method includes: Obtain the first set of minutiae from the first fingerprint image and the second set of minutiae from the second fingerprint image, and obtain matching minutiae pairs between the first set of minutiae and the second set of minutiae; Select a matching minutiae pair as the reference minutiae pair, and perform a rigid transformation on all minutiae in the second set of minutiae such that the two minutiae of the reference minutiae pair coincide. For each matching minutiae pair, the contribution value of the matching minutiae pair is calculated based on the matching minutiae pair and the corresponding neighboring minutiae pair; Wherein, the neighborhood minutiae pair is all matching minutiae pairs within the neighborhood range defined by the matching minutiae pair; The matching score between the first fingerprint image and the second fingerprint image is calculated based on the contribution values ​​of all matching minutiae pairs. For each matching minutiae pair, the contribution value of the matching minutiae pair is calculated based on the matching minutiae pair and the corresponding neighboring minutiae pair, including: The first contribution value of the matching minutiae pair is calculated based on the distance and angle difference between the two minutiae points of the matching minutiae pair. For each neighborhood minutiae pair within the defined neighborhood range of the matching minutiae pair, a rigid transformation is performed on the second set of minutiae points that makes the two minutiae points of the matching minutiae pair coincide. For each neighborhood minutiae pair, calculate the second contribution value of each neighborhood minutiae pair based on the distance and angle difference between the two minutiae in the neighborhood minutiae pair, as well as the corresponding interior angle value; Wherein, the interior angle value is the angle value of the interior angle formed by the two minutiae of the neighborhood minutiae pair and the matching minutiae pair; The contribution value of the matching minutiae pair is calculated based on the first contribution value of the matching minutiae pair and the second contribution values ​​of all neighboring minutiae pairs; The first contribution value of the matching minutiae pair is calculated using the following formula; sd=Z(d i ,μ1,τ1)*Z(Δθ i ,m2,t2) Where sd is the first contribution value of the matched minutiae pair, Z(,,) is the sigmoid function, and d i and Δθ i Let μ1, τ1, μ2, τ2 be the distance and angle difference between the two minutiae of the i-th matching minutiae pairs, respectively, i = 1, 2, ..., N, where N is the number of matching minutiae pairs, and μ1, τ1, μ2, τ2 are the set parameters; The second contribution value of each neighborhood minutiae pair is calculated using the following formula; Among them, sn j Sn represents the second contribution value of the j-th neighborhood minutiae pair, where j = 1, 2, ..., M, and M is the number of neighborhood minutiae pairs within the defined neighborhood range of the matched minutiae pair. j =Z(d) j ,μ1,τ1)*Z(Δθ j ,μ2,τ2)*Z(α j ,μ3,τ3),d j and Δθ j Let α be the distance and angle difference between the two minutiae of the j-th neighborhood minutiae pair. j Let μ3,τ3 be the angle value of the interior angle formed by the two minutiae of the j-th neighborhood minutiae pair and the matching minutiae pair, where μ3,τ3 are set parameters; The contribution value of the matched minutiae pair is calculated using the following formula; Among them, pairscore i Let np and nq be the contribution values ​​of the i-th matching minutiae pair, where np and nq are the number of the first group of minutiae and the second group of minutiae within the set neighborhood of the matching minutiae pair, respectively.

2. The fingerprint comparison method according to claim 1, characterized in that, The matching score between the first fingerprint image and the second fingerprint image is calculated using the following formula: Where s is the matching score between the first fingerprint image and the second fingerprint image, and N1 and N2 are the number of minutiae in the first group and the number of minutiae in the second group, respectively.

3. The fingerprint comparison method according to claim 1, characterized in that, The step of selecting a matching minutiae pair as the reference minutiae pair includes: Calculate the centroid of the first set of minutiae for all matching minutiae pairs; Find the first set of minutiae closest to the centroid from all matching minutiae pairs, and use the matching minutiae pair to which the found first set of minutiae belongs as the reference minutiae pair.

4. The fingerprint comparison method according to claim 1, characterized in that, The rigid transformation performed on all detail points in the second group, ensuring that the two detail points of the reference detail point pair coincide, includes: Calculate the angle difference Δθ and coordinate difference Δx, Δy between the two details of the reference detail pair; Δθ=θ2-θ1 Where (x2,y2,θ2) are the coordinates and orientation angles of the first set of detail points of the reference detail point pair, and (x1,y1,θ1) are the coordinates and orientation angles of the second set of detail points of the reference detail point pair; Calculate the coordinates and orientation angles of the second set of detail points after rigid transformation based on the angle difference Δθ and coordinate difference Δx, Δy between the two detail points of the reference detail point pair; θ′=θ-Δθ Where (x,y,θ) are the coordinates and orientation angles of the second set of detail points before rigid transformation, and (x′,y′,θ′) are the coordinates and orientation angles of the second set of detail points after rigid transformation.

5. The fingerprint comparison method according to any one of claims 1-4, characterized in that, The method further includes: The fingerprint comparison result is obtained based on the matching score between the first fingerprint image and the second fingerprint image.

6. A fingerprint comparison device, characterized in that, The device includes: The data preparation module is used to acquire the first set of minutiae of the first fingerprint image and the second set of minutiae of the second fingerprint image, and to obtain matching minutiae pairs of the first set of minutiae and the second set of minutiae. The rigid transformation module is used to select a matching pair of detail points as a reference pair of detail points, and to perform a rigid transformation on all detail points in the second group of detail points so that the two detail points of the reference pair of detail points coincide. The contribution value calculation module is used to calculate the contribution value of each matching minutiae pair based on the matching minutiae pair and the corresponding neighboring minutiae pair. Wherein, the neighborhood minutiae pair is all matching minutiae pairs within the neighborhood range defined by the matching minutiae pair; The matching score calculation module is used to calculate the matching score between the first fingerprint image and the second fingerprint image based on the contribution values ​​of all matching minutiae pairs. The contribution value calculation module includes: The first calculation unit is used to calculate the first contribution value of the matching minutiae pair based on the distance and angle difference between the two minutiae of the matching minutiae pair; A rigid transformation unit is used to perform a rigid transformation on the second set of minutiae of each neighborhood minutiae pair within a defined neighborhood range of the matching minutiae pair, such that the two minutiae of the matching minutiae pair coincide. The second calculation unit is used to calculate the second contribution value of each neighborhood minutiae pair based on the distance and angle difference between the two minutiae in the neighborhood minutiae pair and the corresponding interior angle value. Wherein, the interior angle value is the angle value of the interior angle formed by the two minutiae of the neighborhood minutiae pair and the matching minutiae pair; The third calculation unit is used to calculate the contribution value of the matching minutiae pair based on the first contribution value of the matching minutiae pair and the second contribution values ​​of all neighboring minutiae pairs. The first contribution value of the matching minutiae pair is calculated using the following formula; sd=Z(d i ,μ1,τ1)*Z(Δθ i ,m2,t2) Where sd is the first contribution value of the matched minutiae pair, Z(,,) is the sigmoid function, and d i and Δθ i Let μ1, τ1, μ2, τ2 be the distance and angle difference between the two minutiae of the i-th matching minutiae pairs, respectively, i = 1, 2, ..., N, where N is the number of matching minutiae pairs, and μ1, τ1, μ2, τ2 are the set parameters; The second contribution value of each neighborhood minutiae pair is calculated using the following formula; Among them, sn j Sn represents the second contribution value of the j-th neighborhood minutiae pair, where j = 1, 2, ..., M, and M is the number of neighborhood minutiae pairs within the defined neighborhood range of the matched minutiae pair. j =Z(d) j ,μ1,τ1)*Z(Δθ j ,μ2,τ2)*Z(α j ,μ3,τ3),d j and Δθ j Let α be the distance and angle difference between the two minutiae of the j-th neighborhood minutiae pair. j Let μ3,τ3 be the angle value of the interior angle formed by the two minutiae of the j-th neighborhood minutiae pair and the matching minutiae pair, where μ3,τ3 are set parameters; The contribution value of the matched minutiae pair is calculated using the following formula; Among them, pairscore i Let np and nq be the contribution values ​​of the i-th matching minutiae pair, where np and nq are the number of the first group of minutiae and the second group of minutiae within the set neighborhood of the matching minutiae pair, respectively.

7. A computer-readable storage medium for fingerprint comparison, characterized in that, It includes a memory for storing processor-executable instructions, which, when executed by the processor, implement the steps of the fingerprint comparison method according to any one of claims 1-5.

8. A device for fingerprint comparison, characterized in that, It includes at least one processor and a memory storing computer-executable instructions, wherein the processor executes the instructions to implement the steps of the fingerprint comparison method according to any one of claims 1-5.

Images