Methods, devices, storage media and equipment for locating the inner boundary of the iris

Abstract

This invention discloses a method, apparatus, storage medium, and device for iris inner boundary localization, belonging to the field of iris recognition. The invention first performs initial localization of the iris inner boundary as a circle. Then, the annular region image containing the initially located iris inner boundary is unfolded into a rectangular region image. Next, gradient calculations are performed along the column directions of the rectangular region image, followed by binarization. The boundary points of the iris inner boundary are determined on the resulting binarized image. Finally, the boundary points are returned to the iris image to obtain the precisely located iris inner boundary. This invention is a universal method for iris inner boundary localization, capable of locating both irregular and regular circular iris inner boundaries, while ensuring accuracy and improving the accuracy of iris recognition.

Description

Technical Field

[0001] This invention relates to the field of iris recognition, and in particular to a method, apparatus, storage medium, and device for locating the inner boundary of the iris. Background Technology

[0002] The development of modern society has placed higher demands on the accuracy, security, and practicality of human identity verification. Identity verification is a ubiquitous issue in daily life, frequently requiring the verification of one's own identity and the identification of others. Traditional photo-based identity verification methods are far outdated, necessitating the search for more secure, reliable, and convenient new methods. Biometric identification technology offers the following advantages: it is less prone to being forgotten or lost, has strong anti-counterfeiting capabilities, is difficult to forge or steal, is portable, and can be used anytime, anywhere.

[0003] Biometric identification technology refers to the use of inherent physiological or behavioral characteristics of the human body for identity authentication, possessing characteristics such as non-replicability, uniqueness, universality, and stability. Iris recognition, with its random detailed and textured features and high stability, inherent isolation and protection capabilities, and the fact that it does not require contactless data collection, offers significant physiological advantages, making it a promising market prospect and valuable for scientific research.

[0004] The iris is a ring-shaped structure located between the pupil and the sclera, such as... Figure 1 The outer and inner circles of the iris. Figure 1 Due to the obstruction of the eyelids and eyelashes, some iris information is lost. The iris is approximately 12mm in diameter and 0.5mm thick. From a recognition perspective, the subtle, interwoven features resembling filaments or stripes in the iris are what make it unique. These features are typically the texture features of the iris and are used for iris recognition.

[0005] Iris localization refers to determining the inner and outer boundaries of the iris. The inner boundary of the iris is the boundary between the pupil and the iris, and is a boundary that resembles a circle. However, due to pupil dilation and constriction, tilting, and other diseases, the inner boundary of the iris is not a true circle. In practical applications, current technology locates the inner boundary of the iris as a circle. This method of locating the inner boundary as a circle leads to inaccurate localization, especially for… Figure 2 The images shown are of slightly slanted irises or irises with shallow texture. Iris recognition is a precise method, especially since the inner iris has rich texture and contains a large amount of information. A one-pixel deviation in locating the inner boundary of the iris can lead to a significant deviation in the recognition result, resulting in inaccurate iris recognition. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides a method, apparatus, storage medium, and device for iris inner boundary localization, which can locate both irregular and regular circular iris inner boundaries, ensuring positioning accuracy and improving the accuracy of iris recognition.

[0007] The technical solution provided by this invention is as follows:

[0008] In a first aspect, the present invention provides a method for locating the inner boundary of the iris, the method comprising:

[0009] On the iris image, the inner boundary of the iris is initially located according to a circle to obtain the initially located inner boundary of the iris;

[0010] The annular region image containing the initially located inner boundary of the iris is expanded into a rectangular region image using a coordinate transformation method.

[0011] Gradient calculation is performed on the rectangular region image along the column direction to obtain a gradient image;

[0012] The gradient image is binarized according to the binarization threshold to obtain a binarized image;

[0013] Obtain boundary points representing the inner boundary of the iris from the binarized image;

[0014] The boundary points are returned to the iris image using coordinate transformation to obtain the precisely located inner boundary of the iris.

[0015] Furthermore, the step of performing gradient calculations on the rectangular region image along the column direction to obtain a gradient image includes:

[0016] The rectangular region image is filtered according to the set filter kernel;

[0017] Perform row subtraction on the filtered image at intervals of a certain number of rows.

[0018] After subtracting the rows, keep the values ​​greater than 0 and set the values ​​less than or equal to 0 to 0 to obtain the gradient image.

[0019] Furthermore, obtaining the boundary points representing the inner boundary of the iris on the binarized image includes:

[0020] Perform morphological operations on the binarized image;

[0021] For each column of the image after morphological operations, if the number of foreground points in the column is greater than a set number, the average value of all foreground points in the column is used as the boundary point of the column; if the number of foreground points in the column is less than or equal to the set number, local gradient calculation is performed on the column to obtain the boundary point of the column.

[0022] For each column of the image after morphological operations, if the number of foreground points in that column is 0, then the boundary points of that column are obtained by curve fitting using the boundary points of the nearest columns before and after that column.

[0023] Furthermore, after initially locating the inner boundary of the iris in a circular pattern on the iris image to obtain the initially located inner boundary of the iris, and before expanding the annular region image containing the initially located inner boundary of the iris image into a rectangular region image using a coordinate transformation method, the method further includes:

[0024] Obtain the noise region of the iris image and remove the noise region from the iris image.

[0025] Furthermore, the initial positioning of the inner boundary of the iris in the iris image according to a circle to obtain the initially positioned inner boundary of the iris includes:

[0026] On the iris image, the inner boundary of the iris is initially positioned as a circle to obtain the center and radius of the inner circle of the iris representing the initially positioned inner boundary of the iris.

[0027] The step of unfolding the annular region image containing the initially located inner boundary of the iris image into a rectangular region image using a coordinate transformation method includes:

[0028] Using the center of the inner circle of the iris as the center, the annular region image with a radius range of [r1, r2] on the iris image is expanded into a rectangular region image by coordinate transformation; where r1 < ri < r2, and ri is the radius of the inner circle of the iris.

[0029] Alternatively, the step of initially locating the inner boundary of the iris in a circular pattern on the iris image to obtain the initially located inner boundary of the iris includes:

[0030] On the iris image, the inner boundary and outer boundary of the iris are initially positioned according to a circle to obtain the center of the inner circle of the iris representing the initially positioned inner boundary, the radius of the inner circle of the iris, and the center of the outer circle representing the initially positioned outer boundary of the iris.

[0031] The step of unfolding the annular region image containing the initially located inner boundary of the iris image into a rectangular region image using a coordinate transformation method includes:

[0032] Using the average of the centers of the inner and outer circles of the iris as the center, the annular region image with a radius range of [r1, r2] on the iris image is expanded into a rectangular region image by coordinate transformation; where r1 < ri < r2, and ri is the radius of the inner circle of the iris.

[0033] Furthermore, r1 = ri - t, r2 = ri + t, where t is the set number of pixels;

[0034] Alternatively, r1 = a1ri, r2 = a2ri, where a1 and a2 are set coefficients, 0 <a1<1<a2。

[0035] Furthermore, the binarization threshold is determined by the following method:

[0036] Obtain the maximum value of each pixel in the gradient image;

[0037] The set percentage of the maximum value is taken as the binarization threshold;

[0038] Where 0 < the set percentage < 100%.

[0039] In a second aspect, the present invention provides an iris inner boundary positioning device, the device comprising:

[0040] The initial positioning module is used to initially locate the inner boundary of the iris in the iris image according to a circle, so as to obtain the initially located inner boundary of the iris;

[0041] The first coordinate transformation module is used to expand the annular region image containing the initially located inner boundary of the iris image into a rectangular region image according to the coordinate transformation method.

[0042] The gradient calculation module is used to perform gradient calculations on the rectangular region image in the column direction to obtain a gradient image.

[0043] The binarization module is used to binarize the gradient image according to the binarization threshold to obtain a binarized image;

[0044] A boundary point acquisition module is used to acquire boundary points representing the inner boundary of the iris on the binarized image;

[0045] The second coordinate transformation module is used to return the boundary points to the iris image according to the coordinate transformation method to obtain the precise positioning of the inner boundary of the iris.

[0046] Furthermore, the gradient calculation module includes:

[0047] The filtering unit is used to perform filtering operations on the rectangular region image according to the set filtering kernel;

[0048] The row subtraction unit is used to perform row subtraction operations on the filtered image at certain intervals;

[0049] The filtering unit is used to retain values ​​greater than 0 after row subtraction and set values ​​less than or equal to 0 to 0, thereby obtaining the gradient image.

[0050] Furthermore, the boundary point acquisition module includes:

[0051] A morphological operation unit is used to perform morphological operations on the binarized image;

[0052] The boundary point calculation unit is used to calculate the boundary points of each column of the image after morphological operations. If the number of foreground points in the column is greater than a set number, the average value of all foreground points in the column is used as the boundary point of the column. If the number of foreground points in the column is less than or equal to the set number, local gradient calculation is performed on the column to obtain the boundary points of the column.

[0053] The boundary completion unit is used to perform curve fitting on each column of the image after morphological operations. If the number of foreground points in the column is 0, the boundary points of the column are obtained by using the boundary points of the nearest columns before and after the column.

[0054] Furthermore, the device also includes:

[0055] The noise acquisition module is used to acquire the noise region of the iris image and remove the noise region from the iris image.

[0056] Furthermore, the initial positioning module includes:

[0057] The first positioning unit is used to initially position the inner boundary of the iris in the iris image according to a circle, and obtain the center and radius of the inner circle of the iris representing the initially positioned inner boundary of the iris.

[0058] The first coordinate transformation module includes:

[0059] The first transformation unit is used to expand an annular region image with a radius range of [r1, r2] on the iris image into a rectangular region image by means of coordinate transformation, with the center of the inner circle of the iris as the center; where r1 < ri < r2, and ri is the radius of the inner circle of the iris.

[0060] or;

[0061] The initial positioning module includes:

[0062] The second positioning unit is used to initially position the inner boundary and outer boundary of the iris in the iris image according to a circle, so as to obtain the center of the inner circle of the iris representing the initially positioned inner boundary, the radius of the inner circle of the iris, and the center of the outer circle representing the initially positioned outer boundary of the iris.

[0063] The first coordinate transformation module includes:

[0064] The second transformation unit is used to expand a ring-shaped region image with a radius range of [r1, r2] on the iris image into a rectangular region image by means of coordinate transformation, with the average value of the center of the inner circle and the center of the outer circle of the iris as the center; where r1 < ri < r2, and ri is the radius of the inner circle of the iris.

[0065] Furthermore, r1 = ri - t, r2 = ri + t, where t is the set number of pixels;

[0066] Alternatively, r1 = a1ri, r2 = a2ri, where a1 and a2 are set coefficients, 0 <a1<1<a2。

[0067] Furthermore, the binarization threshold is determined through the following process:

[0068] Obtain the maximum value of each pixel in the gradient image;

[0069] The set percentage of the maximum value is taken as the binarization threshold;

[0070] Where 0 < the set percentage < 100%.

[0071] Thirdly, the present invention provides a computer-readable storage medium for iris intraboundary localization, including a memory for storing processor-executable instructions, which, when executed by the processor, implement the steps of the iris intraboundary localization method described in the first aspect.

[0072] Fourthly, the present invention provides a device for iris inner boundary localization, comprising at least one processor and a memory storing computer-executable instructions, wherein the processor executes the instructions to implement the steps of the iris inner boundary localization method described in the first aspect.

[0073] The present invention has the following beneficial effects:

[0074] This invention first performs initial positioning of the inner iris boundary as a circle. Then, the annular region image containing the initially positioned inner iris boundary is unfolded into a rectangular region image. Next, gradient calculation is performed along the column direction of the rectangular region image, followed by binarization. The boundary points of the inner iris boundary are determined on the resulting binarized image. Finally, the boundary points are returned to the iris image to obtain the precisely positioned inner iris boundary. This invention is a universal method for locating the inner iris boundary, capable of locating both irregular and regular circular inner iris boundaries, while ensuring accuracy and improving the accuracy of iris recognition. Attached Figure Description

[0075] Figure 1 This is a schematic diagram of a normal iris image;

[0076] Figure 2 A schematic diagram of the iris in strabismus;

[0077] Figure 3 A flowchart illustrating one embodiment of the iris inner boundary localization method of the present invention;

[0078] Figure 4 This is a schematic diagram for initially locating the inner boundary of the iris;

[0079] Figure 5 This is a schematic diagram of iris noise.

[0080] Figure 6 This is a schematic diagram of a rectangular region image;

[0081] Figure 7 This is a schematic diagram of a binarized image;

[0082] Figure 8 A schematic diagram before the iris boundary points are filled in;

[0083] Figure 9 A schematic diagram showing the completed iris boundary points;

[0084] Figure 10 A schematic diagram for precisely locating the inner boundary of the iris;

[0085] Figure 11 A flowchart illustrating another embodiment of the iris inner boundary localization method of the present invention;

[0086] Figure 12 This is a schematic diagram of one embodiment of the iris inner boundary positioning device of the present invention;

[0087] Figure 13 This is a schematic diagram of another embodiment of the iris inner boundary positioning device of the present invention. Detailed Implementation

[0088] 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.

[0089] This invention provides a method for locating the inner boundary of the iris, such as... Figure 3As shown, the method includes:

[0090] S100: On the iris image, the inner boundary of the iris is initially located according to a circle to obtain the initially located inner boundary of the iris.

[0091] This step first acquires and preprocesses the iris image, then takes the inner boundary of the iris as a circle, and performs initial positioning using algorithms such as radial symmetry transformation and calculus detection operator to obtain the approximate iris boundary, which is the initially positioned inner boundary of the iris.

[0092] This invention can perform initial positioning only on the inner boundary of the iris to obtain the center (yi, xi) and radius ri of the inner circle of the iris representing the initially positioned inner boundary. Alternatively, it can perform initial positioning on both the inner and outer boundaries of the iris to obtain the center (yi, xi) and radius ri of the inner circle of the iris representing the initially positioned inner boundary, and the center (yo, xo) and radius ro of the outer circle representing the initially positioned outer boundary.

[0093] An example of initial localization of the inner boundary of the iris is as follows: Figure 4 As shown, the white arc line represents the initial positioning of the inner boundary of the iris, which is the edge of the circle, and the white cross mark represents the center of the initially positioned inner circle of the iris.

[0094] S300: Expand the annular region image containing the initially located inner boundary of the iris in the iris image into a rectangular region image using a coordinate transformation method.

[0095] The initial positioning in S100 has roughly determined the position of the inner boundary of the iris. Based on the initially positioned inner boundary, a certain neighborhood range is extended in both the inward and outward directions to determine the annular region image containing the initially positioned inner boundary. This annular region image can then be unfolded into a rectangular region image. This invention does not limit the specific implementation of the unfolding; two examples are given below for illustration:

[0096] Example 1: When S100 only performs initial positioning of the inner boundary of the iris, this step can take the center of the inner circle of the iris (yixi) as the center and expand the annular region image with a radius range of [r1,r2] (r1<ri<r2) on the iris image into a rectangular region image by means of coordinate transformation.

[0097] The input image is a ring-shaped region centered at (yi, xi) with a radius varying within [r1, r2]. This ring-shaped region within the range [0, 2π] is expanded using coordinate transformation. The expanded rectangular region has r2 - r1 + 1 rows and 360 columns. The expanded rectangular region looks like this. Figure 6 As shown.

[0098] Example 2: When S100 performs initial positioning on the inner boundary and outer boundary of the iris, in this step, the center of the circular area image with a radius range of [r1, r2] on the iris image can be expanded into a rectangular area image according to the coordinate transformation method with the average value of the center of the inner circle of the iris (yi, xi) and the center of the outer circle of the iris (yo, xo) as the center.

[0099] In Example 1 and Example 2, the values of r1 and r2 are not restricted, as long as the inner boundary of the iris can be included in the circular area image. As a preference, the values of r1 and r2 should be such that the circular area image does not exceed the outer boundary of the effective area of the iris, so as not to affect the recognition of the inner boundary during the gradient operation.

[0100] For example, r1 = r - t, r2 = r + t, where t is the set number of pixels. The value of t can be set according to experience or according to the radius of the outer boundary of the iris, so as to avoid the circular area image exceeding the outer boundary of the effective area of the iris. For example, when t = 30, the number of rows of the rectangular area image is 61 and the number of columns is 360.

[0101] Another example, r1 = a1r, r2 = a2r, where a1 and a2 are set coefficients, 0 < a1 < 1 < a2. The values of a1 and a2 can be set according to experience or according to the radius of the outer boundary of the iris, so as to avoid the circular area image exceeding the outer boundary of the effective area of the iris.

[0102] S400: Perform a gradient operation on the rectangular area image in the column direction to obtain a gradient image.

[0103] This step is used to perform a gradient operation on the rectangular area image in the column direction. Corresponding to the circular area image (i.e., the original iris image), it is a gradient operation in the direction of the radius of the inner circle of the iris. Since the gradient change at the inner boundary of the iris on the circular area image is the most intense, the pixel values at the inner boundary of the iris on the gradient image are relatively large.

[0104] S500: Binarize the gradient image according to the binarization threshold to obtain a binarized image.

[0105] Since the pixel values at the inner boundary of the iris on the gradient image are relatively large, by setting an appropriate binarization threshold, the pixels at the inner boundary of the iris can be distinguished from the pixels at other locations. During binarization, the pixels in the gradient image that are greater than or equal to the binarization threshold are set to 255, that is, foreground points, which represent the points at the inner boundary of the iris, and the pixels less than the binarization threshold are set to 0, that is, background points. An example of the binarized image is as Figure 7 shown.

[0106] The binarization threshold of this invention can be set to a fixed value, or it can be adaptively set according to the gradient image to adapt to different gradient images. One method for determining the binarization threshold is as follows:

[0107] S1: Obtain the maximum value of each pixel in the gradient image.

[0108] S2: Take the set percentage of the maximum value as the binarization threshold.

[0109] Where 0 < the set percentage < 100%, for example, a value of 20% can be taken.

[0110] S600: Obtain the boundary point representing the inner boundary of the iris on the binarized image.

[0111] As mentioned above, the foreground points in the binarized image represent points at the inner boundary of the iris. However, these foreground points are not accurate and contain noise (such as noise from eyelashes). Statistical and / or morphological operations are needed to remove the noise and obtain the boundary points that truly represent the inner boundary of the iris. The boundary points of the inner boundary of the iris are represented as a horizontal curve in the binarized image, such as... Figure 9 As shown.

[0112] S700: The boundary points are returned to the iris image using coordinate transformation to obtain a precisely located inner boundary of the iris.

[0113] The boundary points obtained above, on the binarized image of the rectangular region, need to undergo a coordinate transformation opposite to S300 to return the boundary points to the original iris image, thus obtaining the precisely located inner iris boundary. The precisely located inner iris boundary appears as a closed curve resembling a circle in the iris image, such as... Figure 10 As shown.

[0114] This invention first performs initial positioning of the inner iris boundary as a circle. Then, the annular region image containing the initially positioned inner iris boundary is unfolded into a rectangular region image. Next, gradient calculation is performed along the column direction of the rectangular region image, followed by binarization. The boundary points of the inner iris boundary are determined on the resulting binarized image. Finally, the boundary points are returned to the iris image to obtain the precisely positioned inner iris boundary. This invention is a universal method for locating the inner iris boundary, capable of locating both irregular and regular circular inner iris boundaries, while ensuring accuracy and improving the accuracy of iris recognition.

[0115] As an improvement to an embodiment of the present invention, the aforementioned S400 includes:

[0116] S410: Perform a filtering operation on the rectangular region image according to the set filtering kernel.

[0117] S420: Performs a row subtraction operation on the filtered image at intervals of a certain number of rows.

[0118] S430: After subtracting rows, keep the values ​​greater than 0 and set the values ​​less than or equal to 0 to 0 to obtain the gradient image.

[0119] For example, in Figure 6 The rectangular region image shown is subjected to 3*3 median filtering using a 3*3 filter kernel. Then, a row subtraction operation is performed at intervals of a certain number of rows (e.g., 3 rows). That is, the nth row is subtracted from the n+4th row. Values ​​greater than 0 after subtraction are retained, and values ​​less than or equal to 0 are set to 0, thus obtaining the gradient image.

[0120] like Figure 11 As shown, embodiments of the present invention may further include:

[0121] S200: Obtain the noise region of the iris image and remove the noise region from the iris image.

[0122] This step primarily utilizes gradient transformation, grayscale pixel values, and least squares fitting methods to identify noisy regions, remove these regions, and eliminate light spots and the influence of the upper and lower eyelids. For example... Figure 5 As shown, the white areas represent the upper and lower eyelids and the area of ​​the light spot, which are treated as noise.

[0123] The present invention is not limited to a method for obtaining boundary points representing the inner boundary of the iris on the binarized image, and in one example includes:

[0124] S610: Perform morphological operations on the binarized image.

[0125] Morphological operations include closing and opening operations. Closing and opening operations can eliminate the influence of noise such as eyelashes, and remove some boundary points with small gradient changes, thus determining a portion of the boundary points.

[0126] S620: For each column of the image after morphological operation, if the number of foreground points in the column is greater than a set number, the average value of all foreground points in the column is used as the boundary point of the column. If the number of foreground points in the column is less than or equal to the set number, local gradient operation is performed on the column to obtain the boundary point of the column.

[0127] This step first counts the number of foreground points in each column. If the number is greater than a set number (e.g., 2), the average value of all foreground points is taken as the boundary point of the column. If the number is less than or equal to the set number (e.g., 2), local gradient calculation is performed to eliminate noise and obtain the optimal point as the boundary point of the column. In this way, the coordinates of the boundary points of each column can be obtained.

[0128] S630: For each column of the image after morphological operations, if the number of foreground points in that column is 0, then perform curve fitting using the boundary points of the nearest columns before and after that column to obtain the boundary points of that column.

[0129] Iris images are easily affected by noise such as light spots, eyelashes, and eyelids. Figure 5 , 6 The white area is shown. Due to the noise, a portion of the points on the curve formed by the boundary points in the binarized image are missing, as shown... Figure 8 As shown, we need to fill in these missing points to complete the entire boundary.

[0130] If the number of foreground points in a column is 0, it means that the column is missing boundary points and needs to be filled in using curve fitting. Curve fitting methods include Gaussian process regression iteration, least squares method, etc.

[0131] Taking Gaussian process regression iteration as an example, the steps are as follows:

[0132] Using the nearest (e.g., 5) boundary points surrounding the missing point as sampling points for the Gaussian process, and using the missing point as the prediction point, Gaussian process regression iteratively fills in all the missing points. The result after filling in the missing points is as follows: Figure 9 As shown, the filled-in boundary points form a continuous, smooth curve on the rectangular region image, while in the original iris image they appear as a closed, smooth curve resembling a circle. By filling in the missing points, the accuracy of the inner boundary of the iris is improved.

[0133] This invention also provides an iris inner boundary positioning device, such as... Figure 12 As shown, the device includes:

[0134] Initial positioning module 1 is used to initially position the inner boundary of the iris in a circle on the iris image to obtain the initially positioned inner boundary of the iris.

[0135] The first coordinate transformation module 3 is used to expand the annular region image containing the initially located inner boundary of the iris into a rectangular region image using a coordinate transformation method.

[0136] The gradient calculation module 4 is used to perform gradient calculations on the rectangular region image in the column direction to obtain a gradient image.

[0137] Binarization module 5 is used to binarize the gradient image according to a binarization threshold to obtain a binarized image.

[0138] Boundary point acquisition module 6 is used to acquire boundary points representing the inner boundary of the iris on the binarized image.

[0139] The second coordinate transformation module 7 is used to return the boundary points to the iris image according to the coordinate transformation method to obtain the precisely located inner boundary of the iris.

[0140] This invention first performs initial positioning of the inner iris boundary as a circle. Then, the annular region image containing the initially positioned inner iris boundary is unfolded into a rectangular region image. Next, gradient calculation is performed along the column direction of the rectangular region image, followed by binarization. The boundary points of the inner iris boundary are determined on the resulting binarized image. Finally, the boundary points are returned to the iris image to obtain the precisely positioned inner iris boundary. This invention is a universal method for locating the inner iris boundary, capable of locating both irregular and regular circular inner iris boundaries, while ensuring accuracy and improving the accuracy of iris recognition.

[0141] As an improvement to this embodiment of the invention, the gradient calculation module includes:

[0142] The filtering unit is used to perform filtering operations on the rectangular region image according to the set filtering kernel.

[0143] The row subtraction unit is used to perform row subtraction operations on the filtered image at intervals of a certain number of rows.

[0144] The filtering unit is used to retain values ​​greater than 0 after row subtraction and set values ​​less than or equal to 0 to 0, thereby obtaining the gradient image.

[0145] This invention is not limited to the specific implementation of the boundary point acquisition module. In one example, it includes:

[0146] A morphological operation unit is used to perform morphological operations on the binarized image.

[0147] The boundary point calculation unit is used to calculate the boundary points of each column of the image after morphological operations. If the number of foreground points in the column is greater than a set number, the average value of all foreground points in the column is used as the boundary point of the column. If the number of foreground points in the column is less than or equal to the set number, local gradient calculation is performed on the column to obtain the boundary points of the column.

[0148] The boundary completion unit is used to perform curve fitting on each column of the image after morphological operations. If the number of foreground points in the column is 0, the boundary points of the column are obtained by using the boundary points of the nearest columns before and after the column.

[0149] like Figure 13 As shown, the apparatus of the present invention may further include:

[0150] The noise acquisition module 2 is used to acquire the noise region of the iris image and remove the noise region from the iris image.

[0151] In some of these examples, the initial positioning module includes:

[0152] The first positioning unit is used to initially position the inner boundary of the iris in the iris image according to a circle, and obtain the center and radius of the inner circle of the iris representing the initially positioned inner boundary of the iris.

[0153] The first coordinate transformation module includes:

[0154] The first transformation unit is used to expand an annular region image with a radius range of [r1, r2] on the iris image into a rectangular region image by means of coordinate transformation, with the center of the inner circle of the iris as the center; where r1 < ri < r2, and ri is the radius of the inner circle of the iris.

[0155] Alternatively; in some other examples, the initial positioning module includes:

[0156] The second positioning unit is used to initially position the inner boundary and outer boundary of the iris in the iris image according to a circle, so as to obtain the center of the inner circle of the iris representing the initially positioned inner boundary, the radius of the inner circle of the iris, and the center of the outer circle representing the initially positioned outer boundary of the iris.

[0157] The first coordinate transformation module includes:

[0158] The second transformation unit is used to expand a ring-shaped region image with a radius range of [r1, r2] on the iris image into a rectangular region image by means of coordinate transformation, with the average value of the center of the inner circle and the center of the outer circle of the iris as the center; where r1 < ri < r2, and ri is the radius of the inner circle of the iris.

[0159] Where r1 = ri - t, r2 = ri + t, and t is the set number of pixels; or, r1 = a1ri, r2 = a2ri, and a1 and a2 are set coefficients, 0 <a1<1<a2。

[0160] The binarization threshold of this invention can be set to a fixed value, or it can be adaptively set according to the gradient image to adapt to different gradient images. One method for determining the binarization threshold is as follows:

[0161] Obtain the maximum value of each pixel in the gradient image; take a set percentage of the maximum value as the binarization threshold; where 0 < the set percentage < 100%.

[0162] The apparatus provided in this embodiment of the invention operates on the same principle and produces the same technical effects as the aforementioned method embodiments. For the sake of brevity, any parts not mentioned in this apparatus embodiment can be referred to the corresponding content in the aforementioned method embodiments. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the apparatus and units described above can all be referred to the corresponding processes in the aforementioned method embodiments, and will not be repeated here.

[0163] The methods described in the above embodiments of the present 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 the method embodiments of this specification. Therefore, embodiments of the present invention also provide a computer-readable storage medium for iris intraboundary localization, including a memory for storing processor-executable instructions. When the instructions are executed by the processor, they implement the steps of the iris intraboundary localization method of the foregoing embodiments.

[0164] This invention first performs initial positioning of the inner iris boundary as a circle. Then, the annular region image containing the initially positioned inner iris boundary is unfolded into a rectangular region image. Next, gradient calculation is performed along the column direction of the rectangular region image, followed by binarization. The boundary points of the inner iris boundary are determined on the resulting binarized image. Finally, the boundary points are returned to the iris image to obtain the precisely positioned inner iris boundary. This invention is a universal method for locating the inner iris boundary, capable of locating both irregular and regular circular inner iris boundaries, while ensuring accuracy and improving the accuracy of iris recognition.

[0165] 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.

[0166] The storage medium described above may also include other implementation methods according to the description of the method embodiments. The implementation principle and technical effects of this embodiment are the same as those of the foregoing method embodiments. For details, please refer to the description of the relevant method embodiments, which will not be repeated here.

[0167] This invention also provides a device for iris inner boundary localization. The device can be a standalone computer, or it can include an actual operating device that uses one or more of the methods or embodiments described in this specification. The iris inner boundary localization 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 the iris inner boundary localization method described in any one or more of the above embodiments.

[0168] This invention first performs initial positioning of the inner iris boundary as a circle. Then, the annular region image containing the initially positioned inner iris boundary is unfolded into a rectangular region image. Next, gradient calculation is performed along the column direction of the rectangular region image, followed by binarization. The boundary points of the inner iris boundary are determined on the resulting binarized image. Finally, the boundary points are returned to the iris image to obtain the precisely positioned inner iris boundary. This invention is a universal method for locating the inner iris boundary, capable of locating both irregular and regular circular inner iris boundaries, while ensuring accuracy and improving the accuracy of iris recognition.

[0169] The device described above may also include other implementation methods according to the method embodiments. The implementation principle and technical effects of this embodiment are the same as those of the foregoing method embodiments. For details, please refer to the description of the relevant method embodiments, which will not be repeated here.

[0170] 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 method for locating the inner boundary of the iris, characterized in that, The method includes: On the iris image, the inner boundary of the iris is initially located according to a circle to obtain the initially located inner boundary of the iris; The annular region image containing the initially located inner boundary of the iris is expanded into a rectangular region image using a coordinate transformation method. Gradient calculation is performed on the rectangular region image along the column direction to obtain a gradient image; The gradient image is binarized according to the binarization threshold to obtain a binarized image; Obtain boundary points representing the inner boundary of the iris from the binarized image; The boundary points are returned to the iris image using coordinate transformation to obtain a precisely located inner boundary of the iris. The step of obtaining the boundary points representing the inner boundary of the iris on the binarized image includes: Perform morphological operations on the binarized image; For each column of the image after morphological operations, if the number of foreground points in the column is greater than a set number, the average value of all foreground points in the column is used as the boundary point of the column; if the number of foreground points in the column is less than or equal to the set number, local gradient calculation is performed on the column to obtain the boundary point of the column. For each column of the image after morphological operations, if the number of foreground points in that column is 0, then the boundary points of that column are obtained by curve fitting using the boundary points of the nearest columns before and after that column.

2. The method for locating the inner boundary of the iris according to claim 1, characterized in that, The step of performing gradient calculations on the rectangular region image along the column direction to obtain a gradient image includes: The rectangular region image is filtered according to the set filter kernel; Perform row subtraction on the filtered image at intervals of a certain number of rows. After subtracting the rows, keep the values ​​greater than 0 and set the values ​​less than or equal to 0 to 0 to obtain the gradient image.

3. The method for locating the inner boundary of the iris according to claim 1, characterized in that, After initially locating the inner boundary of the iris in a circular pattern on the iris image, and before expanding the annular region image containing the initially located inner boundary of the iris into a rectangular region image using coordinate transformation, the method further includes: Obtain the noise region of the iris image and remove the noise region from the iris image.

4. The method for locating the inner boundary of the iris according to any one of claims 1-3, characterized in that, The initial positioning of the inner boundary of the iris on the iris image according to a circle, to obtain the initially positioned inner boundary of the iris, includes: On the iris image, the inner boundary of the iris is initially positioned as a circle to obtain the center and radius of the inner circle of the iris representing the initially positioned inner boundary of the iris. The step of unfolding the annular region image containing the initially located inner boundary of the iris image into a rectangular region image using a coordinate transformation method includes: Using the center of the inner circle of the iris as the center, the annular region image with a radius range of [r1, r2] on the iris image is expanded into a rectangular region image by coordinate transformation; where r1 < ri < r2, and ri is the radius of the inner circle of the iris.

5. The method for locating the inner boundary of the iris according to any one of claims 1-3, characterized in that, The initial positioning of the inner boundary of the iris on the iris image according to a circle, to obtain the initially positioned inner boundary of the iris, includes: On the iris image, the inner boundary and outer boundary of the iris are initially positioned according to a circle to obtain the center of the inner circle of the iris representing the initially positioned inner boundary, the radius of the inner circle of the iris, and the center of the outer circle representing the initially positioned outer boundary of the iris. The step of unfolding the annular region image containing the initially located inner boundary of the iris image into a rectangular region image using a coordinate transformation method includes: Using the average of the centers of the inner and outer circles of the iris as the center, an annular region image with a radius range of [r1, r2] on the iris image is expanded into a rectangular region image by coordinate transformation; where r1 < ri < r2, and ri is the radius of the inner circle of the iris.

6. The method for locating the inner boundary of the iris according to any one of claims 1-3, characterized in that, The binarization threshold is determined by the following method: Obtain the maximum value of each pixel in the gradient image; The set percentage of the maximum value is taken as the binarization threshold; Where 0 < the set percentage < 100%.

7. A device for locating the inner boundary of the iris, characterized in that, The device includes: The initial positioning module is used to initially locate the inner boundary of the iris in the iris image according to a circle, so as to obtain the initially located inner boundary of the iris; The first coordinate transformation module is used to expand the annular region image containing the initially located inner boundary of the iris image into a rectangular region image according to the coordinate transformation method. The gradient calculation module is used to perform gradient calculations on the rectangular region image in the column direction to obtain a gradient image. The binarization module is used to binarize the gradient image according to a binarization threshold to obtain a binarized image; A boundary point acquisition module is used to acquire boundary points representing the inner boundary of the iris on the binarized image; The second coordinate transformation module is used to return the boundary points to the iris image according to the coordinate transformation method to obtain the precise positioning of the inner boundary of the iris; The boundary point acquisition module includes: A morphological operation unit is used to perform morphological operations on the binarized image; The boundary point calculation unit is used to calculate the boundary points of each column of the image after morphological operations. If the number of foreground points in the column is greater than a set number, the average value of all foreground points in the column is used as the boundary point of the column. If the number of foreground points in the column is less than or equal to the set number, local gradient calculation is performed on the column to obtain the boundary points of the column. The boundary completion unit is used to perform curve fitting on each column of the image after morphological operations. If the number of foreground points in the column is 0, the boundary points of the column are obtained by using the boundary points of the nearest columns before and after the column.

8. A computer-readable storage medium for iris inner boundary localization, characterized in that, It includes a memory for storing processor-executable instructions, which, when executed by the processor, implement the steps of the intrairis boundary localization method according to any one of claims 1-6.

9. A device for locating the inner boundary of the iris, 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 intrairis boundary localization method according to any one of claims 1-6.

Images