Palm vein image processing method and system and storage medium

Abstract

The invention relates to the technical field of image processing, and discloses a palm vein image processing method and system and a storage medium, and the method comprises the steps: obtaining a to-be-recognized palm vein image; according to the to-be-recognized palm vein image, a direction factor is calculated through a hand shape judgment algorithm; performing binarization processing on the palm vein image to be identified to obtain a binarized palm vein image; separating the binary palm vein image through a convolution algorithm to obtain a palm vein trunk image and a palm vein branch image; according to the palm vein trunk image, the palm vein branch image and the direction factor, obtaining a palm vein recombination code through a recombination coding algorithm; and storing the palm vein recombinant code and user information corresponding to a pre-acquired palm vein image into a database. According to the method, the palm vein image processing and storage efficiency can be improved by considering the palm input direction and the fusion codes of the palm vein trunk and the palm vein branches.

Description

Technical Field

[0001] The present invention relates to the technical field of image processing, and in particular to a palm vein image processing method, system and storage medium. Background Art

[0002] At present, the most widely used technology in the field of data security is cryptographic technology based on identity data. This technology uses identity data as a key or identifier for secure operations, but identity data is easily stolen and leaked, resulting in data tampering and identity imitation. Biometric technology is the third generation of identity recognition technology after keyboards and fingerprints. It is the most effective means of identifying individuals. Biometric recognition has the characteristics of being non-copyable and non-tamperable. Biometric technology includes fingerprints, palm prints, faces, irises, retinas, sounds, etc. From the methods of biometric collection and computer vision analysis, it is mainly divided into traditional recognition technology and AI pattern recognition technology. Among them, vein recognition technology is a biometric technology that uses human vein distribution images for identity recognition. Vein recognition technology mainly uses human palm veins. Palm vein information includes palm shape, palm hill, palm print, palm flow map, and palm vein map. During the recognition process, the extracted palm vein information is compared with the database template, and identity recognition is achieved through matching of image features, texture features, geometric features, dynamic features, spectrum features, etc. Palm vein recognition technology can be applied to finance, security, social security, medical and other application fields.

[0003] In the image acquisition process of an existing palm vein acquisition device, a dedicated infrared camera is used to emit near-infrared light, which penetrates the skin and is absorbed by hemoglobin in the blood. Since the blood in the vein strongly absorbs near-infrared light, the palm vein image can be captured. However, since the shape of the palm vein is similar to that of the fingerprint, it is localized and easily generates pseudo-vein images, resulting in recognition errors. At the same time, the palm vein information contains the geometric information and fingerprint information of the palm, resulting in the inability to separate the palm vein and the fingerprint. The palm vein data has a large amount of information redundancy. During the data storage process, the trunk information and branch image information need to be stored, which is not easy to extract. If it is stored directly, the amount of data will be very large, resulting in low efficiency in palm vein image processing and storage. Summary of the invention

[0004] The present invention provides a palm vein image processing method, system and storage medium to solve the problem of low palm vein image processing and storage efficiency.

[0005] In a first aspect, in order to solve the above technical problem, the present invention provides a method for processing a palm vein image, comprising: Acquire a palm vein image to be identified; According to the palm vein image to be identified, a direction factor is calculated by a hand shape determination algorithm; Binarizing the palm vein image to be identified to obtain a binary palm vein image; Separating the binary palm vein image into a palm vein trunk image and a palm vein branch image by a convolution algorithm; Obtaining a palm vein reorganization code by a reorganization coding algorithm according to the palm vein trunk image, the palm vein branch image and the direction factor; The palm vein recombinant code and user information corresponding to the pre-acquired palm vein image are stored in a database.

[0006] In an optional implementation, the calculating the direction factor according to the palm vein image to be identified by a hand shape determination algorithm includes: Extracting the finger touch point coordinates and the palm center coordinates according to the palm vein image; Calculate a first finger vector, a second finger vector and a third finger vector according to the finger touch point coordinates and the palm center coordinates; According to the first finger vector, the second finger vector and the third finger vector, respectively calculate the cosine values ​​of the angles between the three vectors; Calculate the direction factor according to the cosine value of the angle and a preset cosine value threshold; When all the cosine values ​​of the included angles are smaller than a preset cosine value threshold, the direction factor is 0.

[0007] In an optional implementation, binarizing the to-be-identified palm vein image to obtain a binarized palm vein image includes: The palm vein image to be identified is converted into a grayscale image by the following formula: in, is the pixel gray value of the grayscale image, R, G and B Respectively represent the red, green and blue channel intensities of the pixels in the palm vein image to be identified; The grayscale image is converted into a binary palm vein image by the following formula: in, is the pixel value of the binary palm vein image, is the preset grayscale threshold.

[0008] In an optional implementation, the binary palm vein image is separated into a palm vein trunk image and a palm vein branch image by a convolution algorithm, comprising: The convolution result feature map is calculated by the following formula: in, is the convolution result feature map, is the pixel coordinate on the convolution result feature map, is the pixel coordinate of the binary palm vein image, For the preset convolution kernel The weight of the position, is the intermediate process quantity of convolution; According to the convolution result feature map and a preset trunk length threshold, a palm vein trunk image and a palm vein branch image are separated and obtained.

[0009] In an optional implementation, the size of the sliding window of the preset convolution kernel is 5×5.

[0010] In an optional implementation, obtaining the palm vein reorganization code by a reorganization coding algorithm according to the palm vein trunk image, the palm vein branch image and the direction factor includes: Performing an XOR operation on the palm vein trunk image and the palm vein branch image to obtain a palm vein trunk template image; Performing rotation correction on the palm vein trunk template image according to the direction factor to obtain a reconstructed palm vein trunk template image; Binarizing the reconstructed palm vein trunk template image to obtain a palm vein reconstructed code; Wherein, the palm vein recombination code is The matrix of is the size of the reconstructed palmar vein trunk template image.

[0011] In an optional implementation, the performing rotation correction on the palm vein trunk template image according to the direction factor to obtain a reconstructed palm vein trunk template image includes: The reconstructed palmar vein trunk template image is calculated by the following formula: in, is the pixel coordinate of the palmar vein trunk template image, is the pixel coordinate of the translation during the process, is the pixel coordinate of the center point of the palm vein trunk template image, is the pixel coordinate of the reconstructed palmar vein trunk template image, is the direction factor.

[0012] In a second aspect, the present invention provides a palm vein image processing system, comprising: An image acquisition module, used to acquire a palm vein image to be identified; A direction factor module, used to calculate the direction factor according to the palm vein image to be identified by using a hand shape judgment algorithm; A binarization module, used for performing binarization processing on the palm vein image to be identified to obtain a binarized palm vein image; A trunk separation module, used for separating the binary palm vein image into a palm vein trunk image and a palm vein branch image by a convolution algorithm; A coding module, used for obtaining palm vein reorganization coding by a reorganization coding algorithm according to the palm vein trunk image, the palm vein branch image and the direction factor; The storage module is used to store the palm vein reorganization code and the user information corresponding to the pre-acquired palm vein image in a database.

[0013] In a third aspect, the present invention also provides an electronic device, comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein when the processor executes the computer program, the method for processing a palm vein image described in any one of the above is implemented.

[0014] In a fourth aspect, the present invention also provides a computer-readable storage medium, which includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute any one of the above-mentioned methods for processing palm vein images.

[0015] Compared with the prior art, the present invention has the following beneficial effects: The present invention discloses a palm vein image processing method, system and storage medium, the method comprising: obtaining a palm vein image to be identified; calculating a direction factor according to the palm vein image to be identified by a hand shape judgment algorithm; binarizing the palm vein image to be identified to obtain a binarized palm vein image; separating the binarized palm vein image by a convolution algorithm to obtain a palm vein trunk image and a palm vein branch image; obtaining a palm vein recombinant code according to the palm vein trunk image, the palm vein branch image and the direction factor by a recombinant coding algorithm; storing the palm vein recombinant code and the user information corresponding to the pre-acquired palm vein image in a database. The present invention proposes a palm vein image processing method combining image segmentation, convolution algorithm, XOR operation and other technologies, which can efficiently extract palm vein trunk and branch vein information and generate unique palm vein ID information. At the same time, the direction factor is judged by the posture of the finger on the touch screen to realize the dynamic recombinant of palm vein information, thereby improving the accuracy and flexibility of authentication. Furthermore, the trunk feature is encoded into a matrix by a coding algorithm and stored in the database, thereby reducing the palm vein information capacity. The trunk features are extracted from the palm vein trunk, and the branch vein information is used as verification information. The branch vein shape and the palm vein trunk shape are XORed to reduce information redundancy, reduce computational complexity, and improve palm vein image processing and storage efficiency. The recognition efficiency of feature extraction is guaranteed by using the convolution algorithm for feature extraction. The recombinant coding is combined with the XOR operation to reduce the palm vein trunk feature information and branch feature information, and the recognition accuracy of the branch vein feature extracted by the convolution algorithm is improved.

[0016] Specifically, the calculated direction factor can be dynamically adjusted according to the actual touch state of the finger to adapt to the hand shape and touch method of different users. At the same time, by calculating the finger vector and the cosine value of its angle, the hand shape and posture judgment process is simplified, making the algorithm more efficient and reducing the computational complexity. By comparing with the preset cosine value threshold, the horizontal, vertical and tilted states of the finger can be effectively distinguished, thereby providing accurate posture information for subsequent image processing and recognition. Binarization simplifies the pixel values ​​of the image into black and white states, reducing the amount of image data and making the palm vein features more prominent in the image. The computational complexity of the binary image is lower than that of the color or grayscale image, which helps to improve the execution speed of the algorithm and improve data processing efficiency.

[0017] In summary, this method can improve the efficiency of palm vein image processing and storage while ensuring coding accuracy by considering the direction of the palm when entering and the fusion coding of the palm vein trunk and branches. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 It is a schematic flow chart of a palm vein image processing method provided by the first embodiment of the present invention; Figure 2 1 is a schematic diagram of the structure of a palm vein image processing system provided by the second embodiment of the present invention. DETAILED DESCRIPTION

[0019] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0020] At present, the most widely used technology in the field of data security is cryptographic technology based on identity data. This technology uses identity data as a key or identifier for secure operations, but identity data is easily stolen and leaked, resulting in data tampering and identity imitation. Biometric technology is the third generation of identity recognition technology after keyboards and fingerprints. It is the most effective means of identifying individuals. Biometric recognition has the characteristics of being non-copyable and non-tamperable. Biometric technology includes fingerprints, palm prints, faces, irises, retinas, sounds, etc. From the methods of biometric collection and computer vision analysis, it is mainly divided into traditional recognition technology and AI pattern recognition technology. Among them, vein recognition technology is a biometric technology that uses human vein distribution images for identity recognition. Vein recognition technology mainly uses human palm veins. Palm vein information includes palm shape, palm hill, palm print, palm flow map, and palm vein map. During the recognition process, the extracted palm vein information is compared with the database template, and identity recognition is achieved through matching of image features, texture features, geometric features, dynamic features, spectrum features, etc. Palm vein recognition technology can be applied to finance, security, social security, medical and other application fields.

[0021] In the image acquisition process of an existing palm vein acquisition device, a dedicated infrared camera is used to emit near-infrared light, which penetrates the skin and is absorbed by hemoglobin in the blood. Since the blood in the vein strongly absorbs near-infrared light, the palm vein image can be captured. However, since the shape of the palm vein is similar to that of the fingerprint, it is localized and easily generates pseudo-vein images, resulting in recognition errors. At the same time, the palm vein information contains the geometric information and fingerprint information of the palm, resulting in the inability to separate the palm vein and the fingerprint. The palm vein data has a large amount of information redundancy. During the data storage process, the trunk information and branch image information need to be stored, which is not easy to extract. If it is stored directly, the amount of data will be very large, resulting in low efficiency in palm vein image processing and storage.

[0022] To solve the above problems, refer to Figure 1 A first embodiment of the present invention provides a method for processing a palm vein image, comprising the following steps: S11, obtaining a palm vein image to be identified; S12, calculating a direction factor based on the palm vein image to be identified by using a hand shape determination algorithm; S13, performing binarization processing on the palm vein image to be identified to obtain a binarized palm vein image; S14, separating the binary palm vein image into a palm vein trunk image and a palm vein branch image by a convolution algorithm; S15, obtaining a palm vein reorganization code by a reorganization coding algorithm according to the palm vein trunk image, the palm vein branch image and the direction factor; S16, storing the palm vein recombinant code and the user information corresponding to the pre-acquired palm vein image in a database.

[0023] In step S11, a palm vein image to be identified is obtained.

[0024] In one embodiment, the palm vein acquisition device uses a near-infrared camera, starts the palm vein acquisition software, and adjusts the camera to ensure that the user's palm is completely within the acquisition area. Use near-infrared light to illuminate the palm to highlight the vein pattern, because the vein absorbs infrared light, making the vein appear as a darker area in the image. The collected original image is preprocessed, including noise reduction, contrast enhancement, etc., to improve the visibility of the vein. The preprocessed image is saved as the palm vein image to be identified.

[0025] It is worth noting that before the acquisition process begins, the user is instructed to clean the palm to reduce the impact of stains and sweat on the image quality, and to clean the palm vein acquisition device regularly to prevent dust and dirt from affecting the image quality. Furthermore, ensure that the light in the acquisition environment is appropriate and avoid direct strong light, as this will interfere with the acquisition effect of near-infrared light. The user should place the palm on the device with appropriate pressure. Too light or too heavy pressure will affect the clarity of the vein image.

[0026] In step S12, a direction factor is calculated based on the palm vein image to be identified by using a hand shape determination algorithm.

[0027] In one embodiment, the finger touch point coordinates and the palm center coordinates are extracted based on the palm vein image; the finger vector No. 1, the finger vector No. 2 and the finger vector No. 3 are calculated based on the finger touch point coordinates and the palm center coordinates; the cosine values ​​of the angles between the three vectors are calculated based on the finger vector No. 1, the finger vector No. 2 and the finger vector No. 3, respectively; the direction factor is calculated based on the cosine values ​​of the angles and a preset cosine value threshold; wherein, when the cosine values ​​of the angles are all less than the preset cosine value threshold, the direction factor is 0.

[0028] In one embodiment, the hand shape judgment rule is as follows: First, the contact points of the fingers are mapped into a two-dimensional coordinate system. Then, the posture of the fingers is analyzed by connecting the touch points. If the finger is stretched flat on the touch screen, it is considered to be in a horizontal state. If the finger is not fully extended outside the touch screen, it is considered to be in a vertical state. When the finger contact point is located at the edge of the touch screen, the finger is considered to be tilted, and the direction factor is calculated accordingly. Specifically, the direction factor is calculated by: selecting three points on the line from the palm to the fingertip, assuming that the distance from the palm to the fingertip is L, and the vectors from the palm to the three points are vectors a, b and c respectively. Then, the cosine values ​​of the angles between vector b and vectors a and c are calculated respectively. These cosine values ​​are compared with the preset cosine value thresholds in the database: if the cosine values ​​of all angles are less than the preset cosine value threshold, the finger is considered to be placed horizontally. If the cosine values ​​of two angles are greater than the preset cosine value threshold, and the cosine value of one angle is less than the preset cosine value threshold, the finger is considered to be placed vertically. If the cosine value of one angle is greater than the preset cosine value threshold, and both are less than the preset cosine value threshold, the finger is considered to be tilted to the right. In the case where the cosine values ​​of all angles are greater than the preset cosine value threshold, the finger is considered not only placed vertically but also tilted upward. In these cases, the tilt angle is used as an angle factor for subsequent rotation correction.

[0029] In another embodiment, the collected palm vein image is preprocessed, including grayscale, binarization, denoising and other operations; an edge detection algorithm is used to extract edge information in the image, such as Canny edge detection; and the contour of the finger is found through a contour detection algorithm, such as the findContours function provided by OpenCV. A feature point detection algorithm is used to detect the key points of the finger, including the fingertip and joint positions, such as Harris corner point detection or SIFT / SURF feature point detection; the inclination angle of the finger is calculated based on the detected contours and key points. Including: for each contour, its minimum circumscribed rectangle can be calculated, and the angle of the rectangle can be obtained to estimate the direction of the finger. If SIFT or other feature point detection algorithms are used, the inclination angle of the finger can be calculated based on the position of the key points. For example, assuming that the key points of the fingertips and joints have been detected, the direction of the finger can be determined by calculating the vectors between these points.

[0030] In another embodiment, the posture prediction model can be used to predict the distribution map of the main trunks of the palm veins at any other angles through the distribution map of the main trunks of the palm veins at one angle. The posture prediction model can be trained using the public distribution map of the main trunks of the palm veins as a training set.

[0031] It is worth noting that by calculating the angle factor, the system can adjust the captured palm vein image so that it is aligned with the template image in the database in terms of direction. This helps to reduce recognition errors caused by different finger placement angles. The angle factor corrects the angle of the palm vein image in the palm vein recognition system to ensure consistency with the database template, thereby improving recognition accuracy and the system's anti-interference ability. Considering the angle factor can reduce storage requirements and avoid frequent storage of multi-angle images due to differences in finger placement angles, thereby optimizing data storage efficiency. If the angle factor is ignored, for the palm artery of the same user, multi-angle images will be frequently stored at different angles, reducing data storage efficiency.

[0032] In step S13, the palm vein image to be identified is binarized to obtain a binary palm vein image; In one implementation, the palm vein image to be identified is converted into a grayscale image using the following formula: in, is the pixel gray value of the grayscale image, R, G and B Respectively represent the red, green and blue channel intensities of the pixels in the palm vein image to be identified; The grayscale image is converted into a binary palm vein image by the following formula: in, is the pixel value of the binary palm vein image, is the preset grayscale threshold.

[0033] It is worth noting that the palm vein image to be identified is an image in RGB format. R,G,B The intensity values ​​of the red, green and blue color channels of each pixel of the original color palm vein image. The grayscale formula is to calculate the grayscale value by weight according to the sensitivity of the human eye to different colors. The human eye is most sensitive to green (weight 0.587), followed by red (weight 0.299), and least sensitive to blue (weight 0.114). In this way, the color image can be converted to a grayscale image while retaining as much visual information as possible. In order to further improve the quality of the binary image, the following post-processing steps can be performed, including: Morphological operations: such as opening operations (corrosion followed by expansion) to remove small white noise, and closing operations (expansion followed by corrosion) to fill in small black spots in the vein. Removal of isolated white spots: Remove too small white spots through area thresholding, which are usually noise. Smoothing: Use morphological filters to smooth the edges of the veins and reduce the roughness of the image.

[0034] It is worth noting that for images with uneven illumination or complex backgrounds, local adaptive thresholding methods such as Local Binary Pattern (LBP) or adaptive thresholding based on neighborhood average can be used. For images with uniform illumination, the Otsu algorithm can automatically find the optimal threshold to maximize the inter-class variance and segment the entire image using a fixed threshold. After threshold segmentation, in order to eliminate small noise points or fill holes, morphological operations such as opening or closing operations can be used, and connected region marking can be used to identify and separate different objects.

[0035] It is worth noting that for the selection of the threshold, a fixed threshold can be manually selected. This method is suitable for situations where the scene is relatively fixed and the lighting conditions do not change much. Or by analyzing the grayscale histogram of the image, a suitable threshold can be found. For example, a valley or peak in the histogram can be selected as the threshold. In some cases, the user may want to manually select the threshold. This can be achieved by displaying the image and letting the user adjust the threshold through a slider or other means.

[0036] In step S14, the binary palm vein image is separated into a palm vein trunk image and a palm vein branch image by a convolution algorithm.

[0037] In one implementation, the convolution result feature map is calculated using the following formula:

[0038] in, is the convolution result feature map, is the pixel coordinate on the convolution result feature map, is the pixel coordinate of the binary palm vein image, For the preset convolution kernel The weight value of the position, is the intermediate process quantity of convolution; According to the convolution result feature map and the preset trunk length threshold, the palm vein trunk image and the palm vein branch image are separated. The size of the sliding window of the preset convolution kernel is 5×5.

[0039] It is worth noting that convolution is a mathematical operation that combines an image with a convolution kernel to produce an output image. In image processing, convolution is often used to achieve various filtering effects. Palmar vein trunks are usually wider and straighter than branch veins. Therefore, a filter can be designed to identify wide straight structures in the image, which correspond to the main veins. At the same time, another filter can be designed to identify thinner and more complex structures, which correspond to branch veins. When the convolution kernel scans the image, it calculates a weighted sum at each position. If the convolution kernel matches the main vein in the image, the output will be a high value; if it does not match, the output will be a low value. By thresholding the convolution result, the responses of the main vein and the branch vein can be separated. The high response area corresponds to the main vein, while the low response area corresponds to the branch vein. Further, the main vein and the branch vein have obvious differences in morphology, such as width, continuity and direction. These differences can be identified by a specific convolution kernel. To prevent the loss of trunk feature information, the size of the sliding window should not be too large, so the size of the convolution kernel sliding window of the present invention is set to 5×5.

[0040] In one embodiment, the convolution kernel is determined. Since the main vein is usually wider than the branch vein, the main vein appears as a straight and continuous path in the image, while the branch vein appears as a thin, tortuous path branching from the main vein, and the contrast between the vein and the surrounding tissue is different. Using a Gabor filter, lines in a specific direction can be detected, and the parameters can be adjusted to adapt to the direction and frequency of the vein.

[0041] In step S15, a palm vein reorganization code is obtained by a reorganization coding algorithm according to the palm vein trunk image, the palm vein branch image and the direction factor.

[0042] In one implementation, an XOR operation is performed on the palm vein trunk image and the palm vein branch image to obtain a palm vein trunk template image; The palm vein trunk template image is subjected to rotation correction according to the direction factor to obtain a reconstructed palm vein trunk template image, including: The reconstructed palmar vein trunk template image is calculated by the following formula: in, is the pixel coordinate of the palmar vein trunk template image, is the pixel coordinate of the translation during the process, is the pixel coordinate of the center point of the palm vein trunk template image, is the pixel coordinate of the reconstructed palmar vein trunk template image, is the direction factor; Binarizing the reconstructed palm vein trunk template image to obtain a palm vein reconstructed code; Wherein, the palm vein recombination code is The matrix of is the size of the reconstructed palmar vein trunk template image.

[0043] It is worth noting that the palm vein trunk image and branch image are merged into a single image through an exclusive OR (XOR) operation, in which the features of the trunk and branches are retained. For each pixel point, the pixel values ​​of the trunk image and the branch image are compared. If the pixel values ​​of the two images at this point are different (one is black and the other is white), the point is white (or 1) in the result image; if they are the same, it is black (or 0). The rotation correction is to adjust the palm vein trunk template image according to the direction factor to ensure that the direction of the image is consistent with the standard template, avoiding the waste of storage resources caused by multi-angle storage. The reconstructed palm vein trunk template image is converted into a binary image, further reducing the storage efficiency. Palm vein reconstructed coding, this matrix contains all the information of the palm vein image after rotation correction and binarization, which can be used in subsequent recognition and verification processes.

[0044] In another embodiment, the SIFT algorithm is used to extract local features of the palm vein image and encode them. This method can capture the detailed features of the vein while taking into account the rotation correction of the image. First, the scale space of the image is constructed, which is achieved by Gaussian blur and image pyramid. At each scale level, SIFT detects local extreme points, which appear as prominent feature points in the scale space of Gaussian difference as key points of the palm vein. For each key point, the SIFT algorithm generates a feature descriptor, which consists of the gradient direction and size of the pixels in the neighborhood around the key point. The descriptor is a vector that contains the feature information of the area around the key point. The palm vein image is encoded using the feature descriptor extracted by the SIFT algorithm. The descriptor of each key point can be regarded as a part of the image, and a description of the entire image can be formed through the collection of these parts. Since the SIFT feature descriptor is very large, quantization, clustering or hashing technology can be used to further compress it. In the palm vein recognition system, these descriptor vectors can be used as palm vein recombinant encoding and stored in the database. In the recognition phase, the new palm vein image is processed using the same SIFT feature extraction process and the extracted feature descriptors are matched with the feature descriptors stored in the database. By matching the number and consistency of key points, the user's identity can be verified.

[0045] In another embodiment, a posture prediction model is trained using machine learning or deep learning technology, with the input being the palm vein image and its related features, such as the trunk, branches, and directional factors, and the output being the palm vein distribution under a standard viewing angle. The model is capable of learning the changing patterns of palm veins at different angles and postures, and adjusting any input image to a standard view accordingly. Based on the palm vein trunk image, the palm vein branch image, and the directional factor, the standardized palm vein image is converted into a compact form representation, namely feature encoding, through the posture prediction model. Further data dimensionality reduction techniques may be used, such as PCA (principal component analysis), LDA (linear discriminant analysis), or other feature selection algorithms. By encoding the features, the required storage space can be significantly reduced while maintaining sufficient information for identity authentication.

[0046] In another embodiment, a linear discriminant analysis method is used to encode the palm vein image, the goal is to find the best representation of the data in a low-dimensional space, which can maximize the difference between classes while minimizing the difference within classes. For each class, the scatter matrix of its internal data points is calculated, where the inter-class scatter matrix measures the degree of separation between different classes; the generalized eigenvalue problem is solved; according to the size of the eigenvalue, the eigenvector corresponding to the largest eigenvalue is selected, and these eigenvectors define the best low-dimensional space; the original data is projected into the low-dimensional space defined by these optimal eigenvectors; the obtained projection vector is used as the recombinant encoding of the palm vein image.

[0047] In step S16, the palm vein recombinant code and the user information corresponding to the pre-acquired palm vein image are stored in a database.

[0048] In one embodiment, the database includes multiple tables, such as a user information table, a palm vein code table, and a user authority table. The palm vein image of the user is processed using the method described above to generate a palm vein reconstruction code. Before the palm vein reconstruction code and user information are stored in the database, necessary security processing, such as data encryption, is performed to protect user privacy and data security. During subsequent access, the user makes an access request through the interface of the palm vein recognition system. The system collects the palm vein image of the user. The user places the palm on the palm vein collection device, and the device captures the palm vein image through an infrared camera. The collected palm vein image is processed using the same algorithm as during registration to generate a palm vein reconstruction code for the current access. The system matches the palm vein reconstruction code of the current session with the palm vein code stored in the database to find the corresponding user information. If the palm vein reconstruction code successfully matches a code in the database, the system will verify the identity of the user and decide whether to authorize access based on the user's authority. If the match fails, the system will deny the access request and record the attempt for auditing.

[0049] In summary, the present invention discloses a palm vein image processing method, system and storage medium, focusing on improving the accuracy and stability of palm vein recognition technology, and solving the problem of low efficiency of data storage in the prior art. The method obtains palm vein images of the palm through an infrared camera, uses a binarization algorithm to distinguish the trunk and branch veins in the palm vein image, and cuts and retains the key branch end trunk part as branch vein information. The palm vein trunk is encoded by a convolution algorithm, and the trunk features are extracted by Fourier transform and filtering to generate a palm vein trunk feature image. At the same time, the branch vein image is encoded and XORed with the trunk image to form a palm shape template containing the palm vein trunk and branches. In order to improve the flexibility and accuracy of recognition, the method also proposes a method for judging the direction factor based on the posture of the finger on the touch screen, and uses this direction factor to reorganize the palm vein trunk and branch veins to generate a palm vein reorganization code. Furthermore, the method can also use the SIFT algorithm to analyze and identify the palm vein feature image, and match it with the palm vein information in the database to achieve efficient identity authentication. In terms of data storage, this method significantly reduces the storage capacity of palm vein information and reduces the computational complexity by combining the palm vein trunk feature vector with the branch vein information. Through this processing method, the system can not only reduce information redundancy, but also effectively avoid the interference of pseudo-vein images, and improve the stability and accuracy of palm vein recognition. In short, this method can improve the efficiency of palm vein image processing and storage by considering the direction of palm entry and the fusion coding of palm vein trunk and branches.

[0050] Reference Figure 2 A second embodiment of the present invention provides a palm vein image processing system, comprising: An image acquisition module, used to acquire a palm vein image to be identified; A direction factor module, used to calculate the direction factor according to the palm vein image to be identified by using a hand shape judgment algorithm; A binarization module, used for performing binarization processing on the palm vein image to be identified to obtain a binarized palm vein image; A trunk separation module is used to separate the binary palm vein image into a palm vein trunk image and a palm vein branch image by a convolution algorithm; A coding module, used for obtaining palm vein reorganization coding by a reorganization coding algorithm according to the palm vein trunk image, the palm vein branch image and the direction factor; The storage module is used to store the palm vein reorganization code and the user information corresponding to the pre-acquired palm vein image in a database.

[0051] Preferably, the image acquisition module is used to: Obtain the palm vein image to be identified.

[0052] Preferably, the direction factor module is used to: According to the palm vein image to be identified, the direction factor is calculated by a hand shape judgment algorithm, including: Extracting the finger touch point coordinates and the palm center coordinates according to the palm vein image; Calculate a first finger vector, a second finger vector and a third finger vector according to the finger touch point coordinates and the palm center coordinates; According to the first finger vector, the second finger vector and the third finger vector, respectively calculate the cosine values ​​of the angles between the three vectors; Calculate the direction factor according to the cosine value of the angle and a preset cosine value threshold; When all the cosine values ​​of the included angles are smaller than a preset cosine value threshold, the direction factor is 0.

[0053] Preferably, the binarization module is used for: Binarizing the palm vein image to be identified to obtain a binary palm vein image, including: The palm vein image to be identified is converted into a grayscale image by the following formula: in, is the pixel gray value of the grayscale image, R, G and B Respectively represent the red, green and blue channel intensities of the pixels in the palm vein image to be identified; The grayscale image is converted into a binary palm vein image by the following formula: in, is the pixel value of the binary palm vein image, is the preset grayscale threshold.

[0054] Preferably, the trunk separation module is used to: The binary palm vein image is separated into a palm vein trunk image and a palm vein branch image by a convolution algorithm, including: The convolution result feature map is calculated by the following formula: in, is the convolution result feature map, is the pixel coordinate on the convolution result feature map, is the pixel coordinate of the binary palm vein image, For the preset convolution kernel The weight of the position, is the intermediate process quantity of convolution; According to the convolution result feature map and a preset trunk length threshold, a palm vein trunk image and a palm vein branch image are separated and obtained.

[0055] Preferably, the size of the sliding window of the preset convolution kernel is 5×5.

[0056] Preferably, the encoding module is used for: According to the palm vein trunk image, the palm vein branch image and the direction factor, a palm vein reorganization code is obtained by a reorganization coding algorithm, including: Performing an XOR operation on the palm vein trunk image and the palm vein branch image to obtain a palm vein trunk template image; The palm vein trunk template image is subjected to rotation correction according to the direction factor to obtain a reconstructed palm vein trunk template image, including: The reconstructed palmar vein trunk template image is calculated by the following formula: in, is the pixel coordinate of the palmar vein trunk template image, is the pixel coordinate of the translation during the process, is the pixel coordinate of the center point of the palm vein trunk template image, is the pixel coordinate of the reconstructed palmar vein trunk template image, is the direction factor; Binarizing the reconstructed palm vein trunk template image to obtain a palm vein reconstructed code; Wherein, the palm vein recombination code is The matrix of is the size of the reconstructed palmar vein trunk template image.

[0057] Preferably, the storage module is used for: The palm vein recombinant code and user information corresponding to the pre-acquired palm vein image are stored in a database.

[0058] It should be noted that a palm vein image processing system provided in an embodiment of the present invention is used to execute all process steps of a palm vein image processing method in the above embodiment, and the working principles and beneficial effects of the two correspond one to one, so they will not be repeated here.

[0059] The embodiment of the present invention further provides an electronic device. The electronic device includes: a processor, a memory, and a computer program stored in the memory and executable on the processor, such as a data acquisition program. When the processor executes the computer program, the steps in the above-mentioned palm vein image processing method embodiments are implemented, such as Figure 1 Alternatively, when the processor executes the computer program, the functions of the modules / units in the above-mentioned device embodiments are realized, such as the data acquisition module.

[0060] Exemplarily, the computer program may be divided into one or more modules / units, which are stored in the memory and executed by the processor to implement the present invention. The one or more modules / units may be a series of computer program instruction segments capable of implementing specific functions, which are used to describe the execution process of the computer program in the electronic device.

[0061] The electronic device may be a computing device such as a desktop computer, a notebook, a PDA, and a smart tablet. The electronic device may include, but is not limited to, a processor and a memory. Those skilled in the art will appreciate that the above components are merely examples of electronic devices and do not constitute a limitation on the electronic device. The electronic device may include more or fewer components than the above components, or may combine certain components, or different components. For example, the electronic device may also include input and output devices, network access devices, buses, etc.

[0062] The processor may be a central processing unit (CPU), other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc. The processor is the control center of the electronic device, and uses various interfaces and lines to connect various parts of the entire electronic device.

[0063] The memory can be used to store the computer program and / or module, and the processor realizes various functions of the electronic device by running or executing the computer program and / or module stored in the memory, and calling the data stored in the memory. The memory can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, an application required for at least one function (such as a sound playback function, an image playback function, etc.), etc.; the data storage area can store data created according to the use of the mobile phone (such as audio data, a phone book, etc.), etc. In addition, the memory can include a high-speed random access memory, and can also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), at least one disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0064] Wherein, if the module / unit integrated in the electronic device is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, the present invention implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Wherein, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium. It should be noted that the content contained in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media do not include electric carrier signals and telecommunication signals.

[0065] It should be noted that the device embodiments described above are merely schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. In addition, in the accompanying drawings of the device embodiments provided by the present invention, the connection relationship between the modules indicates that there is a communication connection between them, which may be specifically implemented as one or more communication buses or signal lines. A person of ordinary skill in the art may understand and implement it without paying any creative effort.

[0066] The specific embodiments described above further illustrate the purpose, technical solutions and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. It is particularly pointed out that for those skilled in the art, any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the scope of protection of the present invention.

Claims

1. A palm vein image processing method, characterized in that: include: Acquire a palm vein image to be identified; According to the palm vein image to be identified, a direction factor is calculated by a hand shape determination algorithm; Binarizing the palm vein image to be identified to obtain a binary palm vein image; Separating the binary palm vein image into a palm vein trunk image and a palm vein branch image by a convolution algorithm; Obtaining a palm vein reorganization code by a reorganization coding algorithm according to the palm vein trunk image, the palm vein branch image and the direction factor; The palm vein recombinant code and user information corresponding to the pre-acquired palm vein image are stored in a database.

2. The palm vein image processing method according to claim 1, characterized in that: The step of calculating the direction factor based on the palm vein image to be identified by using a hand shape judgment algorithm includes: Extracting the finger touch point coordinates and the palm center coordinates according to the palm vein image; Calculate a first finger vector, a second finger vector and a third finger vector according to the finger touch point coordinates and the palm center coordinates; According to the first finger vector, the second finger vector and the third finger vector, respectively calculate the cosine values ​​of the angles between the three vectors; Calculate the direction factor according to the cosine value of the angle and a preset cosine value threshold; When all the cosine values ​​of the included angles are smaller than a preset cosine value threshold, the direction factor is 0.

3. The palm vein image processing method according to claim 1, characterized in that: The step of binarizing the to-be-identified palm vein image to obtain a binarized palm vein image includes: The palm vein image to be identified is converted into a grayscale image by the following formula: in, is the pixel gray value of the grayscale image, R, G and B Respectively represent the red, green and blue channel intensities of the pixels in the palm vein image to be identified; The grayscale image is converted into a binary palm vein image by the following formula: in, is the pixel value of the binary palm vein image, is the preset grayscale threshold.

4. The palm vein image processing method according to claim 1, characterized in that: The binary palm vein image is separated into a palm vein trunk image and a palm vein branch image by a convolution algorithm, including: The convolution result feature map is calculated by the following formula: in, is the convolution result feature map, is the pixel coordinate on the convolution result feature map, is the pixel coordinate of the binary palm vein image, For the preset convolution kernel The weight value of the position, is the intermediate process quantity of convolution; According to the convolution result feature map and a preset trunk length threshold, a palm vein trunk image and a palm vein branch image are separated and obtained.

5. The palm vein image processing method according to claim 4, characterized in that: The size of the sliding window of the preset convolution kernel is 5×5.

6. The palm vein image processing method according to claim 1, characterized in that: The step of obtaining palm vein reorganization coding by a reorganization coding algorithm according to the palm vein trunk image, the palm vein branch image and the direction factor includes: Performing an XOR operation on the palm vein trunk image and the palm vein branch image to obtain a palm vein trunk template image; Performing rotation correction on the palm vein trunk template image according to the direction factor to obtain a reconstructed palm vein trunk template image; Binarizing the reconstructed palm vein trunk template image to obtain a palm vein reconstructed code; Wherein, the palm vein recombination code is The matrix of is the size of the reconstructed palmar vein trunk template image.

7. The palm vein image processing method according to claim 6, characterized in that: The step of performing rotation correction on the palm vein trunk template image according to the direction factor to obtain a reconstructed palm vein trunk template image includes: The reconstructed palmar vein trunk template image is calculated by the following formula: in, is the pixel coordinate of the palmar vein trunk template image, is the pixel coordinate of the translation during the process, is the pixel coordinate of the center point of the palm vein trunk template image, is the pixel coordinate of the reconstructed palmar vein trunk template image, is the direction factor.

8. A palm vein image processing system, characterized in that: include: An image acquisition module, used to acquire a palm vein image to be identified; A direction factor module, used to calculate the direction factor according to the palm vein image to be identified by using a hand shape judgment algorithm; A binarization module, used for performing binarization processing on the palm vein image to be identified to obtain a binarized palm vein image; A trunk separation module, used for separating the binary palm vein image into a palm vein trunk image and a palm vein branch image by a convolution algorithm; A coding module, used for obtaining palm vein reorganization coding by a reorganization coding algorithm according to the palm vein trunk image, the palm vein branch image and the direction factor; The storage module is used to store the palm vein reorganization code and the user information corresponding to the pre-acquired palm vein image in a database.

9. An electronic device, characterized in that: The invention comprises a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein when the processor executes the computer program, the palm vein image processing method according to any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored computer program, wherein when the computer program is executed, the device where the computer-readable storage medium is located is controlled to execute the palm vein image processing method according to any one of claims 1 to 7.

Images