Iris recognition method and system based on GPGPU acceleration and storage medium

Abstract

The invention relates to the technical field of computer vision, and discloses an iris recognition method and system based on GPGPU acceleration and a storage medium, and the method comprises the steps: receiving an image obtained through the shooting of an eye part of a target object by an iris collection device based on a GPGPU, and screening an iris image with the definition meeting a preset requirement, extracting effective iris data in the iris image, removing pupil, eyelid and sclera data to obtain a preprocessed iris image, and performing normalization, image enhancement and feature extraction; and comparing the extracted feature information with iris features stored in a database to obtain an iris recognition result. According to the invention, the universal graphics processor is utilized to accelerate the implementation of an iris recognition system algorithm, adjustment and optimization are carried out aiming at the characteristics of the processor, global memory access is reduced, and the parallel computing capability of the universal graphics processor is fully exerted, so that efficient iris recognition can be realized.

Description

Technical Field

[0001] This application relates to the field of computer vision technology, for example, to an iris recognition method, system, and storage medium based on GPGPU acceleration. Background Art

[0002] The iris is a circular part located between the black pupil and the white sclera, containing many detailed features such as interlaced spots, filaments, coronas, stripes, crypts, etc. The iris has unique, stable, and complex texture features. Therefore, iris recognition is one of the most accurate and effective identity recognition methods and has been widely used in fields such as e-commerce and security at present.

[0003] With the rapid development of iris recognition theory, real-time iris recognition technology requires a large amount of processing resources, and it is impossible to achieve fast iris image processing using a digital signal processor (DSP). Because of its serial computer architecture, an ARM processor cannot efficiently process parallel operations.

[0004] Therefore, in the existing technical solutions, it is impossible to efficiently recognize the iris.

[0005] It should be noted that the information disclosed in the above background art section is only used to enhance the understanding of the background of this application, and therefore may include information that does not constitute the prior art known to those of ordinary skill in the art. Summary of the Invention

[0006] To have a basic understanding of some aspects of the disclosed embodiments, a simple summary is given below. This summary is not a general review, nor is it intended to identify key / important constituent elements or delineate the protection scope of these embodiments, but rather serves as a preface to the subsequent detailed description.

[0007] The iris recognition method based on GPGPU acceleration provided by the embodiments of the present disclosure includes:

[0008] Receiving, based on the GPGPU, an image obtained by an iris acquisition device taking a picture of the eyes of a target object, and screening iris images whose clarity meets a preset requirement;

[0009] For the iris images whose clarity meets the preset requirement, extracting valid iris data from the iris images, removing pupil, eyelid, and sclera data, to obtain a preprocessed iris image;

[0010] Performing normalization, image enhancement, and feature extraction on the preprocessed iris image;

[0011] Comparing the extracted feature information with the iris features stored in the database to obtain an iris recognition result.

[0012] In some embodiments, for a captured image with an M×N size, the iris images whose screening sharpness meets the preset requirements satisfy the formula:

[0013]

[0014] where (u, v) are the coordinates of the function f(x, y) transformed to the frequency domain by Fourier transform, u and v are frequency variables; x and y are image variables, and i is the imaginary part;

[0015] Based on the threshold corresponding to the preset requirements, compare with F(u, v), and screen out the images where F(u, v) is greater than or equal to the threshold corresponding to the preset requirements as the iris images whose sharpness meets the preset requirements.

[0016] In some embodiments, the extraction of effective iris data from the iris image, removing pupil, eyelid, and sclera data to obtain a preprocessed iris image includes:

[0017] Adopt the iris localization differential algorithm of circular difference, use an effective calculus operator to calculate the circle parameters to obtain the preprocessed iris image;

[0018] The iris localization differential algorithm of circular difference satisfies the formula:

[0019]

[0020] where I(x, y) represents the eye image containing the iris, dividing by 2πr is to perform normalization processing for different circular radii, and convolving with the Gaussian operator G σ (r), and the * operation is to perform smoothing filtering on the image. The parameter space (x c , y c , r) is the place to search for the maximum value of the gray change during the process of locating the inner and outer boundaries of the iris. This operator iteratively finds the optimal solution in the parameter space of the image, detects the circle in the image. During the iterative process, the operator integrates and normalizes the image on the circumference ds with the center (x, y) and radius r, then calculates the difference, and finally the (x c , y c , r) corresponding to the maximum value of the difference is the circle detected in the image.

[0021] In some embodiments, the normalization of the preprocessed iris image includes:

[0022] Taking the inner and outer circle center coordinates and radius as parameters, map the gray image corresponding to the preprocessed iris image in the rectangular coordinate system to a double dimensionless polar coordinate system.

[0023] In some embodiments, the image enhancement of the preprocessed iris image includes:

[0024] Using histogram equalization technology, the gray-level intervals with a concentration greater than the threshold in the gray-level histogram corresponding to the normalized iris image are transformed into a uniform distribution within the entire gray-level range;

[0025] The discrete expression of the gray-level histogram satisfies the formula:

[0026]

[0027] where r k represents the gray-level value of the k-th (k = 0, 1, 2... l - 1) level in the image, and n k is the number of pixels with the gray level r k in the image, and n is the total number of pixels in the image.

[0028] In some embodiments, the histogram equalization satisfies the formula:

[0029]

[0030] where r j represents the gray-level value of the j-th level in the image.

[0031] In some embodiments, feature extraction is performed on the preprocessed iris image, including:

[0032] Extracting the feature information of the iris from the preprocessed iris image, using the parallel processing ability of GPGPU to calculate the features of the iris image and generate specific encoded information;

[0033] Among them, the iris feature information extraction uses a 2D-Gabor filter, in the form of:

[0034]

[0035] μ 0 is the filter width, the length is v 0 , the position coordinates are (x 0 , y 0 ), α and β are scale factors that limit the filter size, i is the imaginary unit, the frequency tuning value is represented by (μ 0 , v 0 ), the direction is The frequency space is represented as

[0036] In some embodiments, the generation of specific encoded information includes:

[0037] Performing feature encoding on the signs of the filtering results in polar coordinates (ρ, φ), and the expression of the feature encoding is:

[0038]

[0039] Among them, I(ρ, φ) is the polar coordinate form of the image; h {Re,Im} is the feature encoding.

[0040] The iris recognition system based on GPGPU acceleration provided by the embodiments of the present disclosure includes:

[0041] A quality assessment module, configured to receive, based on the GPGPU, an image obtained by an iris acquisition device photographing the eyes of a target object, and screen an iris image whose clarity meets a preset requirement;

[0042] A preprocessing module, configured to, for an iris image whose clarity meets a preset requirement, extract valid iris data in the iris image, and remove pupil, eyelid, and sclera data to obtain a preprocessed iris image;

[0043] A feature extraction module, configured to perform normalization, image enhancement, and feature extraction on the preprocessed iris image;

[0044] A feature comparison module, configured to compare the extracted feature information with the iris features stored in a database to obtain an iris recognition result.

[0045] In some embodiments, a storage medium stores program instructions, and when the program instructions are running, they execute the iris recognition method based on GPGPU acceleration in the above embodiments.

[0046] The iris recognition method, system, and storage medium provided by the embodiments of the present disclosure can achieve the following technical effects:

[0047] Utilize the general-purpose graphics processing unit GPGPU to accelerate the implementation of the iris recognition system algorithm, optimize according to the characteristics of the GPGPU, reduce global memory access, and give full play to the parallel computing ability of the GPGPU, so as to efficiently recognize the iris.

[0048] The above general description and the following description are only exemplary and explanatory, and are not used to limit this application. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] One or more embodiments are exemplarily illustrated by corresponding drawings. These exemplary illustrations and the drawings do not constitute limitations on the embodiments. Elements with the same reference numerals in the drawings are shown as similar elements. The drawings do not constitute a proportional limitation, and among them:

[0050] Figure 1 is a flowchart of a method for iris recognition based on GPGPU acceleration provided by an embodiment of the present disclosure;

[0051] Figure 2 It is a schematic flowchart of another iris recognition method based on GPGPU acceleration provided by an embodiment of the present disclosure;

[0052] Figure 3 It is a schematic structural diagram of an iris recognition system based on GPGPU acceleration provided by an embodiment of the present disclosure;

[0053] Figure 4 It is a schematic structural diagram of another iris recognition system based on GPGPU acceleration provided by an embodiment of the present disclosure. Detailed implementation manners

[0054] In order to be able to understand the features and technical content of the embodiments of the present disclosure in more detail, the implementation of the embodiments of the present disclosure will be described in detail below with reference to the accompanying drawings. The accompanying drawings are only for reference and illustration purposes and are not used to limit the embodiments of the present disclosure. In the following technical descriptions, for the sake of explanation, sufficient understanding of the disclosed embodiments is provided through multiple details. However, one or more embodiments can still be implemented without these details. In other cases, well-known structures and devices can be shown in a simplified manner.

[0055] The terms "first", "second", etc. in the embodiments of the present disclosure are used to distinguish similar objects and do not necessarily need to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged under appropriate circumstances so as to implement the embodiments of the present disclosure described here. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusion.

[0056] Unless otherwise specified, the term "plurality" means two or more.

[0057] In the embodiments of the present disclosure, the character " / " indicates that the objects before and after are in an "or" relationship. For example, A / B means: A or B.

[0058] The term "and / or" is a description of the association relationship of objects and indicates that three relationships can exist. For example, A and / or B means: A or B, or, the three relationships of A and B.

[0059] The term "corresponding" can refer to an association relationship or a binding relationship. A corresponding to B means that there is an association relationship or a binding relationship between A and B.

[0060] The iris recognition method, system, and storage medium provided by the embodiments of the present disclosure will be described below with reference to the accompanying drawings.

[0061] Figure 1 It is a schematic flowchart of an iris recognition method based on GPGPU acceleration provided by an embodiment of the present disclosure, as Figure 1As shown, the iris recognition method based on GPGPU acceleration may include:

[0062] S01, receiving, based on the GPGPU, an image obtained by an iris acquisition device photographing the eyes of a target object, and screening for an iris image whose clarity meets a preset requirement;

[0063] S02, for an iris image whose clarity meets a preset requirement, extracting valid iris data in the iris image, removing pupil, eyelid, and sclera data, to obtain a preprocessed iris image;

[0064] S03, performing normalization, image enhancement, and feature extraction on the preprocessed iris image;

[0065] S04, comparing the extracted feature information with the iris features stored in the database to obtain an iris recognition result.

[0066] In the present disclosure, a general-purpose graphics processing unit GPGPU is used to accelerate the implementation of the iris recognition system algorithm. Tuning is performed according to the characteristics of the GPGPU, global memory access is reduced, and the parallel computing power of the GPGPU is fully utilized, so as to efficiently recognize the iris.

[0067] In some embodiments, for a captured image of size M×N, the screening for an iris image whose clarity meets a preset requirement satisfies the formula:

[0068]

[0069] where (u,v) are the coordinates of the function f(x,y) transformed to the frequency domain by Fourier transform, u and v are frequency variables; x and y are image variables, and i is the imaginary part;

[0070] Based on the threshold corresponding to the preset requirement, compare with F(u,v), and screen out the images where F(u,v) is greater than or equal to the threshold corresponding to the preset requirement as the iris images whose clarity meets the preset requirement.

[0071] In some embodiments, the extracting valid iris data in the iris image, removing pupil, eyelid, and sclera data, to obtain a preprocessed iris image includes:

[0072] Adopting a circular differential iris positioning differential algorithm, using an effective calculus operator to calculate the circle parameters to obtain a preprocessed iris image;

[0073] The circular differential iris positioning differential algorithm satisfies the formula:

[0074]

[0075] Among them, I(x, y) represents the eye image containing the iris. Dividing by 2πr is to perform normalization processing for different circumferential radii, and it is convolved * with the Gaussian operator G σ (r). The convolution * operation is to perform smoothing filtering on the image. The parameter space (x c , y c , r) is the place to search for the maximum value of the gray-scale change during the process of locating the inner and outer boundaries of the iris. This operator iteratively finds the optimal solution in the parameter space of the image to detect the circle in the image. During the iterative process, the operator integrates and normalizes the image on the circumferential ds with the center (x, y) and radius r, then calculates the difference, and finally the (x c , y c , r) corresponding to the maximum value of the difference is the circle detected in the image.

[0076] In some embodiments, the normalization of the preprocessed iris image includes:

[0077] Using the center coordinates and radius of the inner and outer circles as parameters, mapping the gray-scale image corresponding to the preprocessed iris image in the rectangular coordinate system to a double non-dimensional polar coordinate system.

[0078] In some embodiments, the image enhancement of the preprocessed iris image includes:

[0079] Adopting histogram equalization technology to transform the gray-scale interval with a concentration greater than the threshold in the gray-scale histogram corresponding to the normalized iris image into a uniform distribution within the entire gray-scale range;

[0080] The discrete expression of the gray-scale histogram satisfies the formula:

[0081]

[0082] Among them, r k represents the gray-scale value of the k-th (k = 0, 1, 2... l - 1) level in the image, n k is the number of pixels with the gray-scale of r k in the image, and n is the total number of pixels in the image.

[0083] In some embodiments, the histogram equalization satisfies the formula:

[0084]

[0085] Among them, r j represents the gray-scale value of the j-th level in the image.

[0086] In some embodiments, the feature extraction of the preprocessed iris image includes:

[0087] Extract the feature information of the iris from the preprocessed iris image, and utilize the parallel processing ability of GPGPU to calculate the features of the iris image and generate specific coding information;

[0088] Among them, the extraction of iris feature information adopts a 2D-Gabor filter, and its form is:

[0089]

[0090] μ 0 is the filter width, and the length is v 0 , and the position coordinates are (x 0 , y 0 ). α and β are scale factors that limit the filter size, i is the imaginary unit, and the frequency tuning is represented by (μ 0 , v 0 ), and the direction is The frequency space is represented as

[0091] In some embodiments, the generation of specific coding information includes:

[0092] Perform feature encoding on the sign of the filtering result in polar coordinates (ρ, φ), and the expression of the feature encoding is:

[0093]

[0094] Among them, I(ρ, φ) is the polar coordinate form of the image; h {Re,Im} is the feature encoding.

[0095] In a specific example, Figure 1 The modules that need to be deployed to GPGPU in the method in

[0096] Quality assessment module: Select the images that meet the judgment criteria as the input of the identification system, that is, perform quality assessment on the collected images. The assessment criteria mainly include judging whether the images are defocused, whether there is light blur, whether the eyelashes are blocked, and whether the pupils are excessively deformed, etc. GPGPU can be used to accelerate the quality evaluation of image sharpness and quickly screen out images with poor quality. If these images enter the recognition system, it will increase the false matching rate of the subsequent system and consume the running time of the system.

[0097] Preprocessing module: Since the captured eye images contain a lot of redundant information and do not meet the requirements in terms of clarity, etc., preliminary image processing is required before iris feature extraction. GPGPU is used to accelerate preprocessing operations such as iris localization segmentation, normalization, and image enhancement.

[0098] Feature extraction module: Extract the feature information of the iris, such as spots, filaments, stripes, etc., from the preprocessed image. Utilize the parallel processing ability of GPGPU to quickly calculate the features of the iris image and generate specific encoded information.

[0099] Feature comparison module: Compare the extracted feature information with the iris features stored in the database to determine the identity. GPGPU speeds up the matching speed by parallelly processing multiple feature vectors.

[0100] Through the above acceleration module, GPGPU can significantly improve the performance of the iris recognition system, especially in scenarios that require processing a large amount of data or real-time recognition.

[0101] Figure 2 is a schematic flowchart of another iris recognition method provided by an embodiment of the present disclosure. In combination with Figure 2 , Figure 1 the method in

[0102] In a specific example, Figure 1 the method in

[0103] Step 1: Iris image acquisition. Use a specific digital camera device to capture the entire eye of a person and transmit the captured image to the GPGPU memory.

[0104] Step 2: After the information is collected, the host sends a quality assessment signal to the GPGPU to screen out iris images with higher clarity.

[0105] Assume the size of an image is M×N. The discrete Fourier transform of the function f(x, y) is given by the following formula:

[0106]

[0107] If the image is out of focus or motion blurred, the high-frequency component value of the image will be small. Therefore, the method of setting a threshold for the high-frequency band energy value can be used to detect whether the image has sufficient clarity. For images with relatively blurred clarity, a signal is returned to request reshooting.

[0108] Step 3: For images with good quality, it is necessary to extract the effective iris data in the image, remove non-iris data such as the pupil, eyelids, and sclera, and improve the accuracy of the data. The host sends a preprocessing signal to the GPGPU. The GPGPU first accelerates the iris localization and segmentation operations.

[0109] Adopt the iris localization differential algorithm of circular difference. This algorithm uses an effective calculus operator to calculate the circle parameters:

[0110]

[0111] where I(x, y) represents the eye image containing the iris. Dividing by 2πr normalizes different circular radii, and convolving with the Gaussian operator G σ (r). The convolution * operation is to perform smoothing filtering on the image. The role of this operator is to iteratively find the optimal solution in the parameter space (x c , y c , r) of the image, so as to detect the circle in the image. During the iteration process, the operator integrates and normalizes the image on the circumference ds with the center (x, y) and radius r, then calculates the difference, and finally the (x c , y c , r) corresponding to the maximum value of the difference is the circle detected in the image. The circular difference algorithm can perform iterative search from coarse to fine by setting different iteration steps. Using this operator to detect the circle in the image has high robustness and accuracy, and its accuracy can reach a single pixel.

[0112] Step 4: Iris normalization. Since the distances from the human eye to the lens in two different shootings cannot be exactly the same, there are certain size differences in the iris images. After iris localization, iris normalization can unify different iris images into a fixed size and corresponding position, thereby eliminating the influence of translation, scaling, and rotation on iris recognition. Using the inner and outer circle center coordinates and radii as parameters, the iris grayscale image in the rectangular coordinate system is mapped to a double dimensionless polar coordinate system image. Normalizing the iris image can expand the annular iris image into a rectangular image with a fixed resolution. The iris texture information in the rectangular image is further encoded for matching.

[0113] Step 5: Image enhancement. The normalized iris image is complete but still has noise, including unreasonable lighting, light and shadow and other interference information. Therefore, it is necessary to perform an enhancement operation on the iris information to make its information more prominent. The histogram equalization technique can be used to enhance the normalized iris image, changing the gray histogram of the original image from a relatively concentrated gray interval to a uniform distribution within the entire gray range, thus enhancing the overall contrast of the image.

[0114] The discrete expression of the grayscale histogram is as follows:

[0115]

[0116] where r k represents the grayscale value of the k-th (k = 0, 1, 2... l - 1) level in the image, and n k is the number of pixels with the grayscale value of r k in the image, and n is the total number of pixels in the image. The following formula is used to perform equalization operation on the image:

[0117]

[0118] After performing histogram equalization on the iris normalized image, the range of the image grayscale values is increased, and the

[0119] contrast of the image is enhanced

[0120] Step 6: Feature extraction. Extract the feature information of the iris from the preprocessed image, and utilize the parallel processing ability of GPGPU to quickly calculate the features of the iris image and generate specific encoded information. The iris feature extraction adopts a 2D-Gabor filter, and its form is:

[0121]

[0122] μ 0 is the filter width, the length is v 0 , the position coordinates are (x 0 , y 0 ), α and β are scale factors that limit the filter size, i is the imaginary unit, the frequency tuning value is represented by (μ 0 , v 0 ), and the direction is The frequency space is expressed as

[0123] Encode the sign of the filtering result in polar coordinates, and the expression of the feature encoding is:

[0124]

[0125] In the formula, (ρ, φ) is the polar coordinate form of the image; h {Re,Im} is the feature encoding. Since the two-dimensional Gabor filter has a real part and an imaginary part, the filtering result has the sign information of the real part and the imaginary part. By encoding the real part and the imaginary part of the filtering result with 0 and 1 respectively, several different sequences can be obtained.

[0126] Step 7: Feature comparison. Compare the extracted feature information with the iris features stored in the database. The GPGPU speeds up the matching process by processing multiple feature vectors in parallel. Let A and B represent the feature sequences to be recognized and in the database respectively. Then the similarity between A and B is:

[0127]

[0128] N is the total length of the feature encoding, A i and B i are used to represent the feature encoding of the i-th bit in the feature sequence to be recognized. is the exclusive OR operation, and HD represents the Hamming distance. The smaller the HD value, the more similar the two irises are.

[0129] In the present disclosure, the general-purpose image processor GPGPU has a large number of computing cores, which specifically support multiple data parallel operations to improve the processing speed. Since the iris recognition algorithm involves processing a large amount of data, GPGPU is selected for acceleration processing, which not only improves performance but also avoids waste of a large amount of infrastructure resources and saves costs. In addition, due to the large memory requirements for iris recognition, the reasonable utilization of various memories of GPGPU can significantly improve the overall performance of the program. Using the general-purpose graphics processor GPGPU to accelerate the implementation of the iris recognition system algorithm, optimizing for the characteristics of GPGPU, reducing global memory access, and giving full play to the parallel computing ability of GPGPU, so as to efficiently recognize irises.

[0130] Figure 3 is a schematic structural diagram of an iris recognition system based on GPGPU acceleration provided by an embodiment of the present disclosure. As Figure 3 shown, the iris recognition system based on GPGPU acceleration may include:

[0131] A quality assessment module 301, configured to receive, based on the GPGPU, an image obtained by an iris acquisition device photographing the eyes of a target object, and screen an iris image whose clarity meets a preset requirement;

[0132] A preprocessing module 302, configured to, for an iris image whose clarity meets a preset requirement, extract valid iris data in the iris image, and remove pupil, eyelid, and sclera data to obtain a preprocessed iris image;

[0133] A feature extraction module 303, configured to perform normalization, image enhancement, and feature extraction on the preprocessed iris image;

[0134] A feature comparison module 304, configured to compare the extracted feature information with the iris features stored in the database to obtain an iris recognition result.

[0135] In some embodiments, for a captured image with an M×N size, the above-mentioned iris images whose sharpness meets the preset requirements satisfy the formula:

[0136]

[0137] where (u, v) are the coordinates of the function f(x, y) transformed to the frequency domain by Fourier transform, u and v are variables of frequency; x and y are variables of the image, and i is the imaginary part;

[0138] Based on the threshold corresponding to the preset requirements, compare with F(u, v), and select the images where F(u, v) is greater than or equal to the threshold corresponding to the preset requirements as the iris images whose sharpness meets the preset requirements.

[0139] In some embodiments, the above-mentioned extraction of effective iris data from the iris image, removing pupil, eyelid, and sclera data to obtain a preprocessed iris image includes:

[0140] Adopt the iris localization differential algorithm of circular difference, use an effective calculus operator to calculate the circle parameters to obtain the preprocessed iris image;

[0141] The iris localization differential algorithm of circular difference satisfies the formula:

[0142]

[0143] where I(x, y) represents the eye image containing the iris, dividing by 2πr is to perform normalization processing for different circular radii, and convolving with the Gaussian operator G σ (r) is to perform smoothing filtering on the image. The parameter space is the place where the maximum value of the gray change is searched during the process of locating the inner and outer boundaries of the iris. This operator iteratively finds the optimal solution in the parameter space (x c , y c , r) of the image, detects the circle in the image. During the iterative process, the operator integrates and normalizes the image on the circumference ds with the center (x, y) and radius r, then calculates the difference, and finally the (x c , y c , r) corresponding to the maximum value of the difference is the circle detected in the image.

[0144] In some embodiments, the above-mentioned normalization of the preprocessed iris image includes:

[0145] Using the inner and outer circle center coordinates and radius as parameters, map the grayscale image corresponding to the preprocessed iris image in the rectangular coordinate system to a double dimensionless polar coordinate system.

[0146] In some embodiments, the above-mentioned image enhancement of the preprocessed iris image includes:

[0147] Using histogram equalization technology, the gray-level interval with a concentration greater than the threshold in the gray-level histogram corresponding to the normalized iris image is transformed into a uniform distribution within the entire gray-level range;

[0148] The discrete expression of the gray-level histogram satisfies the formula:

[0149]

[0150] where r k represents the gray-level value of the k-th (k = 0, 1, 2... l - 1) level in the image, and n k is the number of pixels with the gray level of r k in the image, and n is the total number of pixels in the image.

[0151] In some embodiments, the above histogram equalization satisfies the formula:

[0152]

[0153] where r j represents the gray-level value of the j-th level in the image.

[0154] In some embodiments, the above feature extraction from the preprocessed iris image includes:

[0155] Extracting the feature information of the iris from the preprocessed iris image, using the parallel processing ability of GPGPU to calculate the features of the iris image and generate specific coding information;

[0156] Among them, the iris feature information extraction uses a 2D-Gabor filter, in the form of:

[0157]

[0158] μ 0 is the filter width, the length is v 0 , the position coordinates are (x 0 , y 0 ), α and β are scale factors that limit the filter size, i is the imaginary unit, the frequency tuning value is represented by (μ 0 , v 0 ), and the direction is The frequency space is represented as

[0159] In some embodiments, the above generation of specific coding information includes:

[0160] Performing feature coding on the sign of the filtering result in polar coordinates (ρ, φ), and the expression of the feature coding is:

[0161]

[0162] where I(ρ, φ) is the polar coordinate form of the image; h {Re,Im} is the feature encoding.

[0163] It should be noted that, for the foregoing method embodiments, for the sake of simple description, they are all expressed as a series of action combinations. However, those skilled in the art should know that the present disclosure is not limited by the described action sequence, because according to the present disclosure, certain steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all optional embodiments, and the actions and modules involved are not necessarily essential to the present disclosure.

[0164] Combined with Figure 4 As shown, the embodiment of the present disclosure provides another iris recognition system 400 based on GPGPU acceleration, including a processor 404 and a memory 401. Optionally, the system may further include a communication interface 402 and a bus 403. Among them, the processor 404, the communication interface 402, and the memory 401 can complete mutual communication through the bus 403. The communication interface 402 can be used for information transmission. The processor 404 can call the logical instructions in the memory 401 to execute the iris recognition method based on GPGPU acceleration in the above embodiments.

[0165] In addition, when the logical instructions in the above-mentioned memory 401 are implemented in the form of software functional units and sold or used as an independent product, they can be stored in a computer-readable storage medium.

[0166] The memory 401, as a computer-readable storage medium, can be used to store software programs and computer-executable programs, such as the program instructions / modules corresponding to the methods in the embodiments of the present disclosure. The processor 404 executes functional applications and data processing by running the program instructions / modules stored in the memory 401, that is, implements the iris recognition method based on GPGPU acceleration in the above embodiments.

[0167] The memory 401 may include a program storage area and a data storage area. Among them, the program storage area can store an operating system and application programs required for at least one function; the data storage area can store data created according to the use of the terminal device, etc. In addition, the memory 401 may include a high-speed random access memory and may also include a non-volatile memory.

[0168] An embodiment of the present disclosure provides a computer-readable storage medium storing computer-executable instructions, which are configured to execute an iris recognition method accelerated by a GPGPU.

[0169] The above computer-readable storage medium may be a transient computer-readable storage medium or a non-transient computer-readable storage medium.

[0170] The technical solution of the embodiment of the present disclosure may be embodied in the form of a software product. The computer software product is stored in a storage medium and includes one or more instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method of the embodiment of the present disclosure. The foregoing storage medium may be a non-transient storage medium, including: various media such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disc that can store program codes, or may also be a transient storage medium.

[0171] The above description and the drawings fully illustrate the embodiments of the present disclosure so that those skilled in the art can practice them. Other embodiments may include structural, logical, electrical, process, and other changes. Embodiments merely represent possible variations. Unless explicitly required, individual components and functions are optional, and the order of operations may vary. Parts and features of some embodiments may be included in or substituted for parts and features of other embodiments. As used in the description of the embodiments, unless the context clearly indicates otherwise, the singular forms "a", "an", and "the" are intended to also include the plural forms. Similarly, as used in this application, the term "and / or" refers to any and all possible combinations including one or more of the associated listed items. Additionally, when used in this application, the term "comprise" and its variants "comprises" and / or "comprising" etc. mean the presence of the stated features, wholes, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components, and / or groupings thereof. Without further limitation, an element defined by the statement "comprising an..." does not exclude the presence of additional identical elements in the process, method, or device including the element. In this document, each embodiment may focus on the differences from other embodiments, and the same or similar parts among the various embodiments may be referred to each other. For the methods, products, etc. disclosed in the embodiments, if they correspond to the method part disclosed in the embodiments, the relevant parts may refer to the description of the method part.

[0172] Those skilled in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner can depend on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods for each specific application to implement the described functions, but such implementation should not be considered to exceed the scope of the embodiments of the present disclosure. Those skilled in the art can clearly understand that for the convenience and conciseness of description, the specific working processes of the systems, devices, and units described above can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0173] In the embodiments disclosed herein, the disclosed methods, products (including but not limited to devices, equipment, etc.) can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of units can be merely a logical function division, and there can be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces. The indirect couplings or communication connections of devices or units can be in electrical, mechanical, or other forms. The units described as separate components can be or can not be physically separated, and the components displayed as units can be or can not be physical units, that is, they can be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to implement this embodiment. In addition, in the embodiments of the present disclosure, the various functional units can be integrated in a processing unit, or each unit can exist physically alone, or two or more units can be integrated in one unit.

[0174] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to embodiments of the present disclosure. In this regard, each block in the flowchart or block diagram may represent a module, a segment of code, or a part thereof, which contains one or more executable instructions for implementing the specified logical function. In some alternative implementations, the functions noted in the blocks may occur in a different order than noted in the accompanying drawings. For example, two consecutive blocks may in fact be executed substantially in parallel, or they may sometimes be executed in the reverse order, depending on the functions involved. In the descriptions corresponding to the flowcharts and block diagrams in the accompanying drawings, the operations or steps corresponding to different blocks may also occur in a different order than disclosed in the description, and sometimes there is no specific order between different operations or steps. For example, two consecutive operations or steps may in fact be executed substantially in parallel, or they may sometimes be executed in the reverse order, depending on the functions involved. Each block in the block diagram and / or flowchart, and combinations of blocks in the block diagram and / or flowchart, may be implemented by a dedicated hardware-based system that performs the specified functions or actions, or may be implemented by a combination of dedicated hardware and computer instructions.

Claims

1. An iris recognition method based on GPGPU acceleration, characterized in that: The method comprises: Based on GPGPU, the image of the target object's eyes captured by the iris acquisition device is received, and the iris image with a clarity that meets the preset requirements is selected; For an iris image whose clarity meets the preset requirements, extract effective iris data from the iris image, remove pupil, eyelid and sclera data, and obtain a preprocessed iris image; Normalize, enhance and extract features of the preprocessed iris image; The extracted feature information is compared with the iris features stored in the database to obtain the iris recognition result.

2. The method according to claim 1, characterized in that For the captured image of size M×N, the iris image whose definition meets the preset requirement is selected to satisfy the formula: Among them, (u,v) is the coordinate of the function f(x,y) transformed from Fourier to the frequency domain, u and v are frequency variables; x and y are image variables, and i is the imaginary part; Based on the threshold value corresponding to the preset requirement, it is compared with F(u,v), and the image whose F(u,v) is greater than or equal to the threshold value corresponding to the preset requirement is selected as the iris image whose clarity meets the preset requirement.

3. The method according to claim 1, characterized in that The extracting effective iris data from the iris image, removing pupil, eyelid and sclera data, and obtaining a preprocessed iris image includes: The iris localization difference algorithm based on circle difference is adopted, and an effective calculus operator is used to calculate the circle parameters to obtain the preprocessed iris image; The iris positioning difference algorithm of the circumferential difference satisfies the formula: Among them, I(x, y) represents the eye image including the iris, and the division by 2πr is to normalize the different circle radii, which is different from the Gaussian operator G with a standard deviation of σ. σ (r), the convolution operation is to smooth the image, the parameter space (x c ,y c ,r) is the place where the maximum grayscale change is searched in the process of locating the inner and outer boundaries of the iris. The operator iterates in the parameter space in the image to find the optimal solution and detect the circle in the image. During the iteration process, the operator integrates and normalizes the image on the circle ds with the center (x, y) and radius r, and then calculates the difference. Finally, the maximum difference corresponds to (x c ,y c ,r) is the circle detected in the image.

4. The method according to claim 1, characterized in that: The step of normalizing the preprocessed iris image comprises: The grayscale image corresponding to the preprocessed iris image in the rectangular coordinate system is mapped into a dual dimensionless polar coordinate system using the center coordinates and radius of the inner and outer circles as parameters.

5. The method according to claim 1, characterized in that Perform image enhancement on the preprocessed iris image, including: By using histogram equalization technology, the grayscale intervals with a concentration greater than a threshold in the grayscale histogram corresponding to the normalized iris image are transformed into a uniform distribution within the entire grayscale range; The discrete expression of the grayscale histogram satisfies the formula: Among them, r k Indicates the kth (k=0,1,2…l-1) grayscale value in the image, n k For the image where r appears k The number of pixels of this grayscale, n is the total number of pixels in the image.

6. The method according to claim 5, characterized in that The histogram equalization satisfies the formula: Among them, r j Represents the j-th grayscale value in the image.

7. The method according to claim 6, characterized in that Perform feature extraction on the preprocessed iris image, including: Extract iris feature information from the preprocessed iris image, use the parallel processing capability of GPGPU to calculate the features of the iris image, and generate specific coding information; Among them, iris feature information extraction uses 2D-Gabor filter in the form of: μ0 is the filter width, the length is v0, the position coordinate is (x0, y0), α and β are the scale factors that limit the filter size, i is the imaginary unit, the frequency adjustment value is expressed as (μ0, v0), and the direction is The frequency space is represented as 8. The method according to claim 7, characterized in that The generating of specific coding information includes: The sign of the filtering result under the polar coordinates (ρ, φ) is feature encoded, and the expression of feature encoding is: Among them, I(ρ,φ) is the polar coordinate form of the image; h {Re,Im} Encode the features.

9. An iris recognition system based on GPGPU acceleration, characterized in that: The system comprises: A quality assessment module is used to receive images of the eyes of the target object captured by the iris acquisition device based on GPGPU, and select iris images whose clarity meets preset requirements; A preprocessing module is used to extract effective iris data from an iris image whose clarity meets preset requirements, remove pupil, eyelid, and sclera data, and obtain a preprocessed iris image; A feature extraction module is used to normalize, enhance and extract features of the preprocessed iris image; The feature comparison module is used to compare the extracted feature information with the iris features stored in the database to obtain the iris recognition result.

10. A storage medium storing program instructions, characterized in that: When the program instructions are executed, the iris recognition method based on GPGPU acceleration is executed as described in any one of claims 1 to 7.

Images