Identity verification method based on iris convolution information automatic analysis

Abstract

The invention discloses an identity verification method based on automatic analysis of iris convolution information. The method comprises the following steps: S1, respectively acquiring a template iris grayscale image and a test iris grayscale image; s2, converting the template iris grayscale image and the test iris grayscale image into normalized enhanced images; s3, performing convolution processing on the normalized enhanced image; s4, normalizing the operation data; s5, carrying out pooling operation; s6, carrying out expansion operation on the template iris pooling data; s7, analyzing the template iris expansion data to obtain a template iris recognition parameter range; and S8, matching the test iris pooling data of the tester with the template iris recognition parameter range to obtain an identity verification result. According to the method, the effect of people in the construction process of the recognition model is reduced, the manual participation process of threshold setting is reduced, the automation degree of the iris recognition process is improved, and the sensitivity of the recognition model to data is reduced so as to improve the adaptability of the recognition model.

Description

Technical Field

[0001] The present invention relates to the technical field of iris identity recognition, and in particular to an identity verification method based on automatic analysis of iris convolution information. Background Art

[0002] At present, with the development of social information industry, the demand for identity authentication has increased dramatically. In addition to the traditional password system, biometric technologies such as face, iris and fingerprint have also begun to be widely used, such as face fingerprint systems in railway stations and office places. Because of its physiological characteristics, the iris is not easy to be copied, and it is easier to be used in highly confidential and high-security places such as military bases.

[0003] The current traditional iris identity authentication method mainly uses thresholds to differentiate and obtain recognition results. This method faces many problems, such as: because of the small amount of training iris data, the unstable feature expression and difficulty in distinguishing caused by non-steady-state irises, the setting of thresholds needs to rely on manual experience judgment, the process is relatively complicated, and the threshold data is sensitive to the number and quality of samples, resulting in low model adaptability and difficulty in quickly responding to changes in sample data. Although the binary judgment of single threshold data on recognition results can be reduced by setting a matching method for continuous threshold intervals, and the adaptability of the threshold can be improved as much as possible, the uncertainty of non-steady-state irises and the errors in the manual experience judgment process will still reduce the accuracy of recognition. Summary of the invention

[0004] The present invention provides an identity authentication method based on automatic analysis of iris convolution information to overcome the technical problems in the prior art that the setting of thresholds needs to rely on manual experience and judgment, the process is relatively complicated, and the threshold data is sensitive to the number and quality of samples, resulting in low model adaptability and difficulty in quickly responding to changes in sample data.

[0005] In order to achieve the above object, the technical solution of the present invention is:

[0006] An identity authentication method based on automatic analysis of iris convolution information, the specific steps comprising:

[0007] S1: respectively obtaining a template iris grayscale image of the template person and a test iris grayscale image of the test person;

[0008] S2: The template iris grayscale image and the test iris grayscale image are converted into the template iris normalized enhanced image and the test iris normalized enhanced image respectively by the Daugman rubber band method and the equalization histogram method;

[0009] S3: performing convolution processing on the template iris normalized enhanced image and the test iris normalized enhanced image respectively through the gradient Laplacian convolution kernel to obtain template iris convolution data and test iris convolution data;

[0010] S4: performing normalization operations on the template iris convolution data and the test iris convolution data respectively to obtain template iris normalized processing data and test iris normalized processing data;

[0011] S5: performing a pooling operation on the template iris normalized processing data and the test iris normalized processing data to obtain template iris pooled data and test iris pooled data;

[0012] S6: performing an expansion operation on the template iris pooling data to obtain template iris expansion data;

[0013] S7: Analyze the template iris expansion data to obtain the template iris recognition parameter range, and use it as the basis for template iris recognition of the template person;

[0014] S8: Match the test iris pooling data of the tester with the template iris recognition parameter range to obtain the identity authentication result.

[0015] Furthermore, in S2, the template iris grayscale image and the test iris grayscale image are converted into a template iris normalized enhanced image and a test iris normalized enhanced image of 256×32 dimensions respectively by using the Daugman rubber band method and the equalization histogram method.

[0016] Furthermore, in S4, the general formula for normalizing the template iris convolution data and the test iris convolution data is:

[0017]

[0018] Where F(x,y) is the iris convolution data of the coordinate point (x,y); F p is the average value of the iris convolution data; N(x,y) is the iris normalized processing data of the coordinate point (x,y); x represents the horizontal coordinate and y represents the vertical coordinate.

[0019] Furthermore, in S5, the general formula for performing a pooling operation on the template iris normalized processing data and the test iris normalized processing data is:

[0020]

[0021] Where N(x1×16+i1,y1×16+t1) is the iris normalized data with coordinate point (x1×16+i1,y1×16+t1); T(x1,y1) is the iris pooled data with coordinate point (x1,y1).

[0022] Furthermore, in S6, the process of expanding the template iris pooling data is as follows:

[0023] S61: Calculate the analysis parameters of all template iris grayscale images at each coordinate point. The general calculation formula of the analysis parameters is:

[0024]

[0025] In the formula, H t (x2, y2) represents the analysis parameters of the t-th template iris grayscale image at the coordinate point (x2, y2); T t (x2, y2) represents the iris pooling data of the t-th template iris grayscale image at the coordinate point (x2, y2); T t (1,1) represents the iris pooling data of the t-th template iris grayscale image at the coordinate point (1,1); T i (x2, y2) represents the iris pooling data of the i-th template iris grayscale image at the coordinate point (x2, y2); T i (1,1) represents the iris pooling data of the i-th template iris grayscale image at coordinate point (1,1);

[0026] S62: comparing the analysis parameters of all template iris grayscale images at each coordinate point, obtaining the minimum value of the analysis parameter corresponding to each coordinate point, and then obtaining the template iris grayscale image corresponding to the minimum value of the analysis parameter;

[0027] The template ratio parameter of the coordinate point is calculated based on the iris pooling data of the template iris grayscale image corresponding to the minimum value of the analysis parameter. The general calculation formula is:

[0028]

[0029] Where, T s (x2, y2) represents the iris pooling data of the sth template iris grayscale image at the coordinate point (x2, y2); T s (1,1) represents the iris pooling data of the sth template iris grayscale image at the coordinate point (1,1).

[0030] S63: Expand the template iris pooling data based on the template ratio parameter of the coordinate point. The general formula is:

[0031] T new (x2,y2)=k1×k2(x2,y2)×T(1,1)

[0032] Where, T(1,1) represents the template iris pooling data with coordinate point (1,1); T new(x2, y2) represents the template iris expansion data of the coordinate point (x2, y2); k1 is the template multiple parameter, which is a randomly set positive decimal integer.

[0033] Furthermore, in S7, the process of analyzing the template iris expansion data to obtain the template iris recognition parameter range is as follows:

[0034] S71: Calculate the template distance analysis value of all template iris expansion data at each coordinate point. The general calculation formula is:

[0035]

[0036] In the formula, G t2 (x3, y3) represents the template distance analysis value of the t2th template iris grayscale image at the coordinate point (x3, y3); T t2-new (x3, y3) represents the template iris expansion data of the t2th template iris grayscale image at the coordinate point (x3, y3); i2-new (x3, y3) represents the template iris expansion data of the i2th template iris grayscale image at the coordinate point (x3, y3);

[0037] S72: Compare the template distance analysis values ​​of all template iris grayscale images at each coordinate point to obtain the minimum and maximum template distance analysis values ​​corresponding to each coordinate point;

[0038] S73: sequentially determine the template iris expansion data corresponding to the minimum template distance analysis value and the template iris expansion data corresponding to the maximum template distance analysis value corresponding to each coordinate point, assign the larger value of the template iris expansion data corresponding to the minimum template distance analysis value and the template iris expansion data corresponding to the maximum template distance analysis value to the template iris recognition basis parameter Line1 (x3, y3), and assign the smaller value to the template iris recognition basis parameter Line2 (x3, y3);

[0039] S74: According to the values ​​of Line2(x3,y3) and Line1(x3,y3), the parameter range of the corresponding coordinate point is set to [Line2(x3,y3), Line1(x3,y3)];

[0040] S75: Determine the parameter range of all coordinate points and use it as the template iris recognition basis parameter for the template tester.

[0041] Furthermore, in S8, the process of matching the test iris pooling data of the tester with the template iris recognition parameter range to obtain the identity authentication result is:

[0042] S81: Calculate the similarity data DZ between the test person's test iris pooling data and the template iris recognition parameter range. The general calculation formula is:

[0043]

[0044] Where Z(x4,y4) is the matching parameter of the coordinate point (x4,y4); Line1(x4,y4) and Line2(x4,y4) represent the template iris recognition basis parameters of the coordinate point (x4,y4); k1 represents the template multiple parameter of the template tester; T test (x4, y4) represents the iris pooling data of the tester at the coordinate point (x4, y4);

[0045] S82: Set a similarity threshold T. If the similarity data DZ is greater than the similarity threshold T, it means that the irises of the tester and the template tester belong to the same person. If the similarity data DZ is less than or equal to the similarity threshold T, it means that the irises of the tester and the template tester do not belong to the same person.

[0046] Beneficial effects: The present invention reduces the manual participation process of threshold setting by performing a series of automatic processing on the template person's template iris grayscale image and the test person's test iris grayscale image, thereby reducing the role of humans in the recognition model construction process and improving the automation level of the iris recognition process; at the same time, through a series of operations, a data range that can only be reached by the data of each iris category is set and used as a basis for calculating the similarity, thereby completing the recognition process, improving the versatility of the iris recognition model in different scenarios, and reducing the sensitivity of the recognition model to data; the present invention also improves the adaptability of the recognition model and ensures the accuracy of non-steady-state iris recognition identity authentication in different acquisition states. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0048] Figure 1 It is a flow chart of an identity authentication method based on automatic analysis of iris convolution information in the present invention;

[0049] Figure 2 This is a comparison result diagram in an embodiment of the present invention. DETAILED DESCRIPTION

[0050] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0051] This embodiment provides an identity authentication method based on automatic analysis of iris convolution information. Figure 1 As shown, the specific steps include:

[0052] S1: respectively obtaining a template iris grayscale image of the template person and a test iris grayscale image of the test person;

[0053] Specifically, in this embodiment, 2000 template iris grayscale images of the template person and 1 test iris grayscale image of the test person are collected by using any common iris collector on the market;

[0054] S2: The template iris grayscale image and the test iris grayscale image are converted into a 256×32 dimensional template iris normalized enhanced image and a 256×32 dimensional test iris normalized enhanced image respectively by using the Daugman rubber band method and the equalization histogram method;

[0055] In the field of iris recognition, when the feature expression and recognition process is completed by a normalized enhancement process, a 256×32 dimensional iris grayscale image is usually captured, which can preserve relatively complete information.

[0056] S3: The template iris normalized enhanced image and the test iris normalized enhanced image are convolved with the gradient Laplacian convolution kernel to obtain 256×32 dimensional template iris convolution data and 256×32 dimensional test iris convolution data. The normalized enhanced image can be converted into a binary array through the convolution operation.

[0057] Specifically, the gradient Laplacian convolution kernel is expressed as follows:

[0058] 1 1 1 1 -8 1 1 1 1

[0059] S4: performing normalization operations on the template iris convolution data of 256×32 dimensions and the test iris convolution data of 256×32 dimensions respectively, to obtain template iris normalized processing data of 256×32 dimensions and test iris normalized processing data of 256×32 dimensions;

[0060] S5: performing a pooling operation on the template iris normalized processing data of 256×32 dimensions and the test iris normalized processing data of 256×32 dimensions to obtain template iris pooled data of 16×2 dimensions and test iris pooled data of 16×2 dimensions;

[0061] S6: performing an expansion operation on the template iris pooling data of 16×2 dimensions to obtain template iris expansion data of 16×2 dimensions;

[0062] S7: Analyze the template iris expansion data to obtain a 16×2 dimensional template iris recognition parameter range, and use it as a basis for template iris recognition of the template person;

[0063] S8: Match the test iris pooling data of the tester with the template iris recognition parameter range to obtain the identity authentication result.

[0064] In a specific embodiment, in S4, the general formula for normalizing the template iris convolution data and the test iris convolution data is:

[0065]

[0066] Where, F(x, y) is the iris convolution data of the coordinate point (x, y) based on the plane rectangular coordinate system; F p is the average value of the iris convolution data; N(x, y) is the iris normalized processing data of the coordinate point (x, y); x represents the horizontal coordinate and y represents the vertical coordinate.

[0067] In a specific embodiment, in S5, the general formula for performing a pooling operation on the template iris normalized processing data and the test iris normalized processing data is:

[0068]

[0069] Where N(x1×16+i1,y1×16+t1) is the iris normalized data with coordinate point (x1×16+i1,y1×16+t1); T(x1,y1) is the iris pooled data with coordinate point (x1,y1).

[0070] In a specific embodiment, in S6, the process of expanding the template iris pooling data is as follows:

[0071] S61: Calculate the analysis parameters of all template iris grayscale images at each coordinate point. The general calculation formula of the analysis parameters is:

[0072]

[0073] In the formula, H t(x2, y2) represents the analysis parameters of the t-th template iris grayscale image at the coordinate point (x2, y2); T t (x2, y2) represents the iris pooling data of the t-th template iris grayscale image at the coordinate point (x2, y2); T t (1,1) represents the iris pooling data of the t-th template iris grayscale image at the coordinate point (1,1); T i (x2, y2) represents the iris pooling data of the i-th template iris grayscale image at the coordinate point (x2, y2); T i (1,1) represents the iris pooling data of the i-th template iris grayscale image at coordinate point (1,1);

[0074] Specifically, each coordinate point refers to each coordinate point in the template iris pooling data of 16×2 dimensions.

[0075] S62: comparing the analysis parameters of all 2000 template iris grayscale images at each coordinate point, obtaining the minimum value of the analysis parameter corresponding to each coordinate point, and then obtaining the template iris grayscale image corresponding to the minimum value of the analysis parameter;

[0076] For example, the minimum value of the analysis parameter of the 2000 template iris grayscale images corresponding to a coordinate point (x2, y2) in the 16×2 dimension is min(x2, y2), and the minimum value min(x2, y2) corresponds to the s-th template iris grayscale image;

[0077] Then, the template ratio parameter of the coordinate point is calculated based on the iris pooling data of the template iris grayscale image corresponding to the minimum value of the analysis parameter. The general calculation formula is:

[0078]

[0079] Where, T s (x2, y2) represents the iris pooling data of the sth template iris grayscale image at the coordinate point (x2, y2); T s (1,1) represents the iris pooling data of the sth template iris grayscale image at the coordinate point (1,1).

[0080] S63: Perform an expansion operation on the template iris pooling data based on the template ratio parameter of the coordinate point (x2, y2). The general formula is:

[0081] T new (x2,y2)=k1×k2(x2,y2)×T(1,1)

[0082] Where, T(1,1) represents the template iris pooling data with coordinate point (1,1); T new(x2, y2) represents the template iris expansion data of the coordinate point (x2, y2); k1 is the template multiple parameter, which is a randomly set positive decimal integer.

[0083] In a specific embodiment, in S7, the process of analyzing the template iris expansion data to obtain the template iris recognition parameter range is as follows:

[0084] S71: Calculate the template distance analysis value of all 2000 template iris expansion data at each coordinate point. The general calculation formula is:

[0085]

[0086] In the formula, G t2 (x3, y3) represents the template distance analysis value of the t2th template iris grayscale image at the coordinate point (x3, y3); Represents the template iris expansion data of the t2th template iris grayscale image at the coordinate point (x3, y3); Represents the template iris expansion data of the i2th template iris grayscale image at the coordinate point (x3, y3);

[0087] S72: Compare the template distance analysis values ​​of all 2000 template iris grayscale images at each coordinate point to obtain the minimum and maximum template distance analysis values ​​corresponding to each coordinate point;

[0088] S73: sequentially determine the template iris expansion data corresponding to the minimum template distance analysis value and the template iris expansion data corresponding to the maximum template distance analysis value corresponding to each coordinate point, assign the larger value of the template iris expansion data corresponding to the minimum template distance analysis value and the template iris expansion data corresponding to the maximum template distance analysis value to the template iris recognition basis parameter Line1 (x3, y3), and assign the smaller value to the template iris recognition basis parameter Line2 (x3, y3);

[0089] For example, assuming that the minimum value of the template distance analysis value of a coordinate point (x3, y3) in the 16×2 dimension is min1(x3, y3) and the maximum value is max1(x3, y3); then the iris expansion data corresponding to min1(x3, y3) is T t2-new-1 (x3, y3), max1(x3, y3) corresponds to the iris expansion data T t2-new-2 (x3,y3), T t2-new-1 (x3,y3) and T t2-new-2 The larger value in (x3, y3) is assigned to the template iris recognition based parameter Line1 (x3, y3), and the smaller value is assigned to the template iris recognition based parameter Line2 (x3, y3);

[0090] S74: According to the values ​​of Line2(x3,y3) and Line1(x3,y3), the parameter range of the coordinate point (x3,y3) is set to [Line2(x3,y3), Line1(x3,y3)];

[0091] S75: Determine the parameter range of all coordinate points and use it as the template iris recognition basis parameter for the template tester.

[0092] In a specific embodiment, in S8, the process of matching the test iris pooling data of the tester with the template iris recognition parameter range to obtain the identity authentication result is:

[0093] S81: Calculate the similarity data DZ between the test person's test iris pooling data and the template iris recognition parameter range. The general calculation formula is:

[0094]

[0095]

[0096] Where Z(x4,y4) is the matching parameter of the coordinate point (x4,y4); Line1(x4,y4) and Line2(x4,y4) represent the template iris recognition basis parameters at the coordinate point (x4,y4); k1 represents the template multiple parameter of the template tester; T test (x4, y4) represents the iris pooling data of the tester at the coordinate point (x4, y4);

[0097] S82: Set a similarity threshold T. If the similarity data DZ is greater than the similarity threshold T, it means that the irises of the tester and the template tester belong to the same person. If the similarity data DZ is less than or equal to the similarity threshold T, it means that the irises of the tester and the template tester do not belong to the same person.

[0098] To verify the effect of this embodiment, 100 people were used as experimental categories, and 2000 training samples from the JLU iris database and 2000 training samples from the CASIA iris database were used for training. 20 test irises were selected for each experimental category for experiment, and each test image was subjected to one same category and one different category recognition experiment (a total of 4000 times). The iris recognition comparison methods were respectively adopted: Case 0: the iris recognition method proposed in this embodiment; Case 1: the convolutional neural network (R-CNN) of the Inception v2 model; Case 2: the recognition algorithm based on deep learning; Case 3: the OCaNet iris recognition structure; Case 4: the iris feature recognition method based on the k-means clustering analysis method; Case 5: the iris recognition based on the end-to-end unified framework of MADNet and DSANet.

[0099] The ROC curve of the comparative experimental results is as follows Figure 2 As shown, including (a) ROC curve obtained using JLU iris database; (b) ROC curve obtained using CASIA iris database, Figure 2 In the figure, the horizontal axis represents the false acceptance rate FAR, and the vertical axis represents the false rejection rate FRR. The relationship between the two is described by adjusting the decision basis. The inflection point of the two reflects the EER. The basis for judging the effect of the ROC curve is: the closer the value of the EER in the ROC curve is to the origin, the better the ROC effect is considered. Figure 2 It can be seen that the ROC curve of the method proposed in this embodiment is closer to the origin, indicating that the effect of this solution is better.

[0100] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. An identity authentication method based on automatic analysis of iris convolution information, characterized in that: The specific steps include: S1: respectively obtaining a template iris grayscale image of the template person and a test iris grayscale image of the test person; S2: The template iris grayscale image and the test iris grayscale image are converted into the template iris normalized enhanced image and the test iris normalized enhanced image respectively by using the Daugman rubber band method and the equalization histogram method; S3: performing convolution processing on the template iris normalized enhanced image and the test iris normalized enhanced image respectively through the gradient Laplacian convolution kernel to obtain template iris convolution data and test iris convolution data; S4: performing normalization operations on the template iris convolution data and the test iris convolution data respectively to obtain template iris normalized processing data and test iris normalized processing data; S5: performing a pooling operation on the template iris normalized processing data and the test iris normalized processing data to obtain template iris pooled data and test iris pooled data; S6: performing an expansion operation on the template iris pooling data to obtain template iris expansion data; S7: Analyze the template iris expansion data to obtain the template iris recognition parameter range, and use it as the basis for template iris recognition of the template person; S8: Match the test iris pooling data of the tester with the template iris recognition parameter range to obtain the identity authentication result.

2. The identity authentication method based on automatic analysis of iris convolution information according to claim 1, characterized in that: In S2, the template iris grayscale image and the test iris grayscale image are converted into a template iris normalized enhanced image and a test iris normalized enhanced image of 256×32 dimensions respectively by using the Daugman rubber band method and the equalization histogram method.

3. The identity authentication method based on automatic analysis of iris convolution information according to claim 2, characterized in that: In S4, the general formula for normalizing the template iris convolution data and the test iris convolution data is: Where F(x,y) is the iris convolution data of the coordinate point (x,y); F p is the average value of the iris convolution data; N(x,y) is the iris normalized processing data of the coordinate point (x,y); x represents the horizontal coordinate and y represents the vertical coordinate.

4. The identity authentication method based on automatic analysis of iris convolution information according to claim 3, characterized in that: In S5, the general formula for performing pooling operations on the template iris normalized processing data and the test iris normalized processing data is: Where N(x1×16+i1,y1×16+t1) is the iris normalized data with coordinate point (x1×16+i1,y1×16+t1); T(x1,y1) is the iris pooled data with coordinate point (x1,y1).

5. The identity authentication method based on automatic analysis of iris convolution information according to claim 4, characterized in that: In S6, the process of expanding the template iris pooling data is as follows: S61: Calculate the analysis parameters of all template iris grayscale images at each coordinate point. The general calculation formula of the analysis parameters is: In the formula, H t (x2, y2) represents the analysis parameters of the t-th template iris grayscale image at the coordinate point (x2, y2); T t (x2, y2) represents the iris pooling data of the t-th template iris grayscale image at the coordinate point (x2, y2); T t (1,1) represents the iris pooling data of the t-th template iris grayscale image at the coordinate point (1,1); T i (x2, y2) represents the iris pooling data of the i-th template iris grayscale image at the coordinate point (x2, y2); T i (1,1) represents the iris pooling data of the i-th template iris grayscale image at coordinate point (1,1); S62: Compare the analysis parameters of all template iris grayscale images at each coordinate point to obtain the minimum value of the analysis parameter corresponding to each coordinate point, and then obtain the template iris grayscale image corresponding to the minimum value of the analysis parameter; The template ratio parameter of the coordinate point is calculated based on the iris pooling data of the template iris grayscale image corresponding to the minimum value of the analysis parameter. The general calculation formula is: Where, T s (x2, y2) represents the iris pooling data of the sth template iris grayscale image at the coordinate point (x2, y2); T s (1,1) represents the iris pooling data of the sth template iris grayscale image at the coordinate point (1,1); S63: Expand the template iris pooling data based on the template ratio parameter of the coordinate point. The general formula is: T new (x2,y2)=k1×k2(x2,y2)×T(1,1) Where, T(1,1) represents the template iris pooling data with coordinate point (1,1); T new (x2, y2) represents the template iris expansion data of the coordinate point (x2, y2); k1 is the template multiple parameter, which is a randomly set positive decimal integer.

6. The identity authentication method based on automatic analysis of iris convolution information according to claim 5, characterized in that: In S7, the process of analyzing the template iris expansion data to obtain the template iris recognition parameter range is as follows: S71: Calculate the template distance analysis value of all template iris expansion data at each coordinate point. The general calculation formula is: In the formula, G t2 (x3, y3) represents the template distance analysis value of the t2th template iris grayscale image at the coordinate point (x3, y3); Represents the template iris expansion data of the t2th template iris grayscale image at the coordinate point (x3, y3); Represents the template iris expansion data of the i2th template iris grayscale image at the coordinate point (x3, y3); S72: Compare the template distance analysis values ​​of all template iris grayscale images at each coordinate point to obtain the minimum and maximum template distance analysis values ​​corresponding to each coordinate point; S73: sequentially determine the template iris expansion data corresponding to the minimum template distance analysis value and the template iris expansion data corresponding to the maximum template distance analysis value corresponding to each coordinate point, assign the larger value of the template iris expansion data corresponding to the minimum template distance analysis value and the template iris expansion data corresponding to the maximum template distance analysis value to the template iris recognition basis parameter Line1 (x3, y3), and assign the smaller value to the template iris recognition basis parameter Line2 (x3, y3); S74: According to the values ​​of Line2(x3,y3) and Line1(x3,y3), the parameter range of the corresponding coordinate point is set to [Line2(x3,y3), Line1(x3,y3)]; S75: Determine the parameter range of all coordinate points and use it as the template iris recognition basis parameter for the template tester.

7. The identity authentication method based on automatic analysis of iris convolution information according to claim 6, characterized in that: In S8, the process of matching the test iris pooling data of the tester with the template iris recognition parameter range to obtain the identity authentication result is as follows: S81: Calculate the similarity data DZ between the test person's test iris pooling data and the template iris recognition parameter range. The general calculation formula is: x4=1,...,16y4=1,2 Where Z(x4,y4) is the matching parameter of the coordinate point (x4,y4); Line1(x4,y4) and Line2(x4,y4) represent the template iris recognition basis parameters of the coordinate point (x4,y4); k1 represents the template multiple parameter of the template tester; T test (x4, y4) represents the iris pooling data of the tester at the coordinate point (x4, y4); S82: Set a similarity threshold T. If the similarity data DZ is greater than the similarity threshold T, it means that the irises of the tester and the template tester belong to the same person. If the similarity data DZ is less than or equal to the similarity threshold T, it means that the irises of the tester and the template tester do not belong to the same person.

Images