Human ear vein image acquisition device and personnel identity recognition method

Abstract

The invention discloses a human ear vein image acquisition device and a personnel identity recognition method, and belongs to the technical field of biological feature image acquisition and biological feature recognition. The human ear vein image acquisition device comprises a head-mounted wearing piece; the light source assembly is connected to the head-mounted wearing piece and located behind the human ears. The camera is connected to the head-mounted wearing piece, and the camera is located in front of human ears; the adjusting assembly comprises a shell for accommodating a first knob and a second knob which are mutually nested and can independently rotate, and a first flexible connecting rod and a second flexible connecting rod of which one ends are respectively connected with the first knob and the second knob and the other ends are respectively connected with the light source assembly and the camera; the power supply is located in the shell and electrically connected with the light source assembly and the camera. According to the human ear vein image acquisition device, human ear vein images can be acquired through single-person operation, and texture structure features of the human ear vein images can be stored in electronic equipment through processing for subsequent identity recognition.

Description

Technical Field

[0001] The present invention relates to the technical fields of biometric image acquisition and biometric recognition. More specifically, the present invention relates to a human ear vein image acquisition device and a personnel identification method. Background Art

[0002] Biometric recognition is a science that uses a person's physiological or behavioral characteristics to identify the person's identity. The accurate identification and verification of user identities has always been one of the hot research issues in this field. The purpose is to use biometric recognition to replace the traditional identity verification method through user names and passwords, which makes identity confirmation more accurate, convenient, and secure. Biometric features mainly include fingerprints, palm prints, palm veins, dorsal hand veins, faces, irises, voices, human ears, and retinas, etc.

[0003] Human ear recognition is a type of biometric recognition technology. There have already been many studies on human ear recognition based on ear shape contours, but the research field of human ear veins is blank. Therefore, the present invention focuses on filling the gap in this field. The vein texture has many advantages as a method of human identity authentication. The uniqueness and complexity of human ear veins make recognition more secure. In addition, human ear veins are located under the skin, and their texture features are difficult to be forged by people.

[0004] However, the existing vein acquisition devices are mainly used for collecting finger and palm veins. There is a lack of human ear vein acquisition devices in the market, and there is also no method for using human ear veins to identify the identities of personnel. Summary of the Invention

[0005] In view of this, the present invention provides a human ear vein image acquisition device, a personnel identification method, and a human ear vein recognition system.

[0006] In one aspect of the present invention, a human ear vein image acquisition device is provided. In one embodiment of the present invention, the human ear vein image acquisition device of the present invention includes a head-mounted wearing member; a light source assembly, the light source assembly is connected to the head-mounted wearing member, and when the head-mounted wearing member is worn on a person's head, the light source assembly is located behind the human ear and is used for irradiating the back of the human ear; a camera, the camera is connected to the head-mounted wearing member, and when the head-mounted wearing member is worn on a person's head, the camera is located in front of the human ear and is used for capturing human ear vein images; and a power supply, the power supply is electrically connected to the light source assembly and the camera and is used for supplying power to the light source assembly and the camera.

[0007] In a further embodiment of the present invention, the human ear vein image acquisition device of the present invention further includes an adjustment assembly, the adjustment assembly includes a housing, a first flexible connecting rod, and a second flexible connecting rod. The housing is fixedly arranged at a position corresponding to the position of the human ear on the head-mounted wearing member. The light source assembly and the camera are respectively connected to the housing through the first flexible connecting rod and the second flexible connecting rod.

[0008] In a further embodiment of the present invention, the housing of the adjustment assembly includes a first knob and a second knob that are nested with each other and can rotate independently. The first flexible connecting rod is connected to the first knob, and the second flexible connecting rod is connected to the second knob. The power supply is arranged in the housing of the adjustment assembly. Wherein, wires are arranged in the first flexible connecting rod and the second flexible connecting rod, and the light source assembly and the camera are electrically connected to the power supply in the housing through the wires.

[0009] In a further embodiment of the present invention, the head-mounted wearing member is made of an elastic material, and / or the head-mounted wearing member has an adjustment buckle. The adjustment buckle is in the form of a magic tape, a buckle, or an elastic band, and / or a sponge forehead guard is arranged on the head-mounted wearing member.

[0010] In a further embodiment of the present invention, the light source assembly includes an LED backplane and LED light sources arranged on the LED panel. The LED backplane has a curved shape corresponding to the back of the human ear.

[0011] In another aspect of the present invention, a method for identifying a person's identity using a human ear vein image is provided. In an embodiment of the present invention, the method includes: receiving a human ear vein image from the human ear vein image acquisition device according to the first aspect of the present invention; preprocessing the human ear vein image to obtain a human ear vein ROI image; performing feature extraction on the human ear vein ROI image to obtain human ear vein texture structure features; and comparing the human ear vein texture structure features with pre-stored human ear vein texture structure feature samples, so as to identify a person's identity based on the comparison result.

[0012] In a further embodiment of the present invention, the preprocessing includes: performing ROI extraction on the human ear vein image using an edge detection algorithm, an image binarization method, and a morphological dilation operation to obtain a human ear vein ROI image; performing filtering on the human ear vein ROI image using a median filter to obtain a filtered human ear vein ROI image; and performing a contrast enhancement operation on the filtered human ear vein ROI image using histogram equalization to obtain an enhanced human ear vein ROI image.

[0013] In a further embodiment of the present invention, the feature extraction includes: extracting the texture structure features of the enhanced human ear vein ROI image by using the MFRAT method to obtain the human ear vein texture structure features.

[0014] In a further embodiment of the present invention, the comparison includes comparing the human ear vein texture structure features with the human ear vein texture structure feature samples by using a matching algorithm, including: comparing the human ear vein texture structure features with the human ear vein texture structure feature samples based on the pixel area to obtain a matching score score; and comparing the matching score score with a preset matching threshold Tk. When score is greater than Tk, the personnel identity corresponding to the human ear vein texture structure feature sample is identified; when score is less than Tk, the personnel identity cannot be identified.

[0015] In another aspect of the present invention, a human ear vein recognition system is provided. In an embodiment of the present invention, the human ear vein recognition system includes: a human ear vein image acquisition device according to the first aspect of the present invention; and an electronic device, the electronic device includes a processor and a memory, and the memory stores instructions, and the instructions execute the method for identifying personnel identity by using a human ear vein image according to the second aspect of the present invention when executed by the processor.

[0016] The beneficial effects of this application are as follows:

[0017] The human ear vein image acquisition device of the present invention uses ordinary LED lighting to obtain vein texture features in a transmission mode. Compared with using near-infrared lighting to obtain vein texture features, the cost is reduced. In addition, the human ear vein image acquisition device of the present invention extracts human ear veins in a non-contact manner, meeting the user's requirements for hygiene.

[0018] The method for identifying personnel identity by using a human ear vein image of the present invention can adaptively cope with different image qualities and lighting conditions, ensuring that the ear vein blood vessel area can be stably and accurately extracted under variable environments. In addition, this method effectively retains the details of the auricle and suppresses the interference of noise by enhancing the contrast, detecting edges, and performing morphological operations, thereby improving the accuracy of the extraction results. Description of the Drawings

[0019] Figure 1 is a schematic diagram of an embodiment of the human ear vein image acquisition device of the present invention;

[0020] Figure 2 is a schematic diagram of another embodiment of the human ear vein image acquisition device of the present invention;

[0021] Figure 3 It is a schematic diagram of an embodiment of the human ear vein recognition system of the present invention;

[0022] Figure 4 It is a block diagram showing the relationship between the main units of the electronic device in the human ear vein recognition system of the present invention;

[0023] Figure 5 It is a flowchart of the method for recognizing a person's identity by using a human ear vein image according to the present invention.

[0024] Explanation of the reference numerals of the elements in the drawings

[0025] 1. Head-mounted wearing member; 2. Adjusting assembly; 21. First knob; 22. Second knob; 23. First flexible connecting rod; 24. Second flexible connecting rod; 3. Light source assembly; 31. LED light source; 32. LED backplane; 4. Camera; 5. Adjusting buckle; 6. Forehead guard. Detailed implementation manners

[0026] As Figure 1 shown, in an embodiment of the present invention, the present invention provides a human ear vein image acquisition device, which includes: a head-mounted wearing member; a light source assembly 3, the light source assembly 3 is connected to the head-mounted wearing member, and when the head-mounted wearing member is worn on a human head, the light source assembly 3 is located behind the human ear and is used to irradiate the back of the human ear; a camera 4, the camera 4 is connected to the head-mounted wearing member, and when the head-mounted wearing member is worn on a human head, the camera 4 is located in front of the human ear and is used to capture a human ear vein image; and a power source, the power source is electrically connected to the light source assembly 3 and the camera 4 and is used to supply power to the light source assembly 3 and the camera 4. Through the human ear vein image acquisition device of the present invention, the acquisition of the human ear vein image of oneself can be completed by a single person's operation.

[0027] As Figure 2 shown, in a further embodiment of the present invention, the human ear vein image acquisition device further includes an adjusting assembly 2, the adjusting assembly 2 includes a housing and a first flexible connecting rod 23 and a second flexible connecting rod 24, and the flexible connecting rod can be made of, for example, a fiber-reinforced composite material. The housing is fixedly arranged at a position corresponding to the position of the human ear on the head-mounted wearing member, for example, by sewing or bonding. The light source assembly 3 and the camera 4 are respectively connected to the housing through the first flexible connecting rod 23 and the second flexible connecting rod 24. In this way, the distance and angle between the light source assembly 3 and the camera 4 and the human ear can be adjusted by adjusting the flexible connecting rod, so that the camera 4 can capture a clear image.

[0028] Still referring to Figure 2, in an embodiment of the present invention, the housing of the adjusting assembly 2 includes a first knob 21 and a second knob 22 that are nested with each other and can rotate independently. The outer diameter of the second knob 22 is smaller than the inner diameter of the first knob 21. The first knob 21 and the second knob 22 are respectively mounted on a fixed shaft located at the central position through their respective bearings. The inner ring of the bearing cooperates with the fixed shaft, and the outer rings are respectively connected to the first knob 21 and the second knob 22. Alternatively, an annular track can be machined on the inner side of the first knob 21. The track can be in the shape of a dovetail groove, a T-shaped groove, a circular groove, or the like. A slider that matches the track is installed on the outer side of the second knob 22. The slider can be made of metal or plastic and is precisely machined to slide smoothly within the track while restricting the radial movement of the second knob 22 to ensure that it can rotate around the central axis. Alternatively, rollers can be used instead of the slider, and multiple rollers are installed on the outer side of the second knob 22. The rollers are in rolling contact with the annular track of the first knob 21. This method has less friction and is more flexible to rotate.

[0029] Optionally, in order to give the knob a clear sense of position during rotation, a positioning structure can be set. For example, a positioning disk with multiple grooves is set on the rotating shaft of the knob, and an elastic positioning pin is installed on another corresponding knob. When the knob rotates, the positioning pin jumps between the grooves, generating a "click" sound to prompt the user of the rotation position and also facilitating precise adjustment.

[0030] Optionally, in order to prevent the knob from over-rotating and damaging the internal structure or exceeding the adjustment range, a limiting block can be set on the rotation path of the knob. When the knob rotates to a certain angle, the limiting block contacts the housing or other fixed components to prevent further rotation.

[0031] Still referring to Figure 2 , in the above embodiment, the first flexible connecting rod 23 is connected to the first knob 21, the second flexible connecting rod 24 is connected to the second knob 22, and the power supply is arranged in the housing of the adjusting assembly 2. Wires are provided in the first flexible connecting rod 23 and the second flexible connecting rod 24. The light source assembly 3 and the camera 4 are electrically connected to the power supply in the housing through the wires. In this way, the distance and angle between the light source assembly 3 and the camera 4 and the human ear can be adjusted in two ways, namely, through the knob and the flexible connecting rod, making the adjustment operation more flexible and the adjustment range wider, so that the light source assembly 3, the human ear, and the camera 4 are located on the same straight line, thereby taking clear images.

[0032] As Figure 3As shown, in an embodiment of the present invention, the camera 4 can be connected to an electronic device such as a computer in a wired or wireless manner. When light passes through the auricle, darker textures will be obtained when encountering the venous blood vessels, and these textures can be regarded as ear veins. After obtaining the ear vein image, the camera 4 can be set to automatically transmit the captured ear vein image to the computer for the next image processing operation. When in use, turn on the switch of the light source component 3. After the light emits from the light source component 3 and penetrates the auricle, the staff uses the computer to manipulate the camera 4 to take pictures. Then, the image captured by the camera 4 is automatically transmitted to the computer. In the ear vein image acquisition stage, according to the image displayed on the computer display screen, it can be checked whether the ear vein image meets the requirements. If not, rotate the knob of the adjustment component 2 and adjust the flexible connecting rod until it meets the requirements. In practice, the ear vein image can be collected several times.

[0033] Return to refer Figure 2 , in an embodiment of the present invention, the head-mounted wearing piece is made of elastic materials such as rubber, thermoplastic elastomer, polyurethane elastomer, spandex, and elastic knitted fabric, so that it can be more easily fitted to a person's head and fixedly worn on the person's head. Additionally or optionally, the head-mounted wearing piece has an adjustment buckle 5, and the adjustment buckle 5 is in the form of a magic tape, a buckle, or an elastic band, so that it can adapt to different head shapes of people. Additionally or optionally, a soft forehead guard 6 made of materials such as sponge is provided on the head-mounted wearing piece, so that the comfort of the head-mounted wearing piece can be increased.

[0034] In an embodiment of the present invention, the light source component 3 includes an LED backplane 32 and LED light sources 31 arranged on the LED panel, and the LED backplane 32 has a curved shape corresponding to the back of the human ear. In this way, the LED light sources 31 on the LED backplane 32 can achieve more uniform illumination of the back of the human ear. More specifically, the present invention uses ordinary LED lights for illumination and adopts a transmission mode to obtain ear vein images, and the cost is reduced compared with using near-infrared illumination to obtain venous texture features. In addition, the ear vein image acquisition device of the present invention extracts ear veins in a non-contact manner, meeting the user's requirements for hygiene.

[0035] In an embodiment, the light source component 3 includes three LED lights, and the three LED lights can work simultaneously when the switch is turned on. Optionally, the light source component 3 can include more or fewer LED lights to adapt to different usage conditions.

[0036] The present invention also provides a method for identifying a person's identity using ear vein images. Figure 5The flowchart of the method is shown, and the method includes: receiving a human ear vein image from the human ear vein image acquisition device of the present invention; preprocessing the human ear vein image to obtain a human ear vein ROI image. The preprocessing includes: performing ROI extraction on the human ear vein image by using an edge detection algorithm, an image binarization method, and a morphological dilation operation, so that feature extraction pays more attention to ear vein features, thereby obtaining a human ear vein ROI image; performing filtering on the human ear vein ROI image by using a median filter to eliminate some unnecessary noises, thereby obtaining a filtered human ear vein ROI image; and performing a contrast enhancement operation on the filtered human ear vein ROI image by using histogram equalization to obtain an enhanced human ear vein ROI image.

[0037] In a preferred embodiment of the present invention, the Canny edge detection algorithm is used to process the image I 1 to extract the significant edges in the image. The Canny edge detection includes three steps: calculating the gradient of the image, non-maximum suppression, and double-threshold detection. First, calculate the image gradient:

[0038]

[0039] where G x and G y are the gradients of the image in the horizontal and vertical directions respectively, which can be calculated by the Sobel operator:

[0040]

[0041]

[0042] where S x and S y are the convolution kernels of the horizontal and vertical Sobel operators respectively. Then, perform non-maximum suppression, retain the significant edges, and determine the intensity of the edges according to the double threshold, and finally obtain the image I 2 .

[0043] In a preferred embodiment of the present invention, in order to enhance the connectivity of the auricle region, a morphological dilation operation is applied. The mathematical expression of the dilation operation is:

[0044]

[0045] where N(x, y) is the neighborhood of the pixel (x, y), and max is the maximum value of the pixel values in the neighborhood. The dilation operation can effectively expand the edges, enhance the connectivity of the auricle region, and reduce the region missing caused by incomplete edge detection.

[0046] In one embodiment of the present invention, the Otsu threshold method is adopted to further segment the ear vein blood vessel area. The Otsu method automatically calculates the optimal threshold T by maximizing the between-class variance. Its goal is to divide the image into two classes: the blood vessel area and the background area. The formula for the between-class variance in the Otsu method is:

[0047]

[0048] where μ 1 (T) and μ 2 (T) are the average gray values of the pixels below and above the threshold T respectively, are the variances of the corresponding regions. Select the threshold T* that maximizes the between-class variance, and then perform binary processing on the image:

[0049]

[0050] The above operations can obtain the edge contour of only the ear vein blood vessel area. Then, according to the auricle contour information, the leftmost, rightmost, topmost, and bottommost four points of the contour can be obtained by scanning in the x and y directions. These 4 points can be applied to the initial rgb image to obtain a rectangular ROI area I ROI .

[0051] In one embodiment of the present invention, for the ear vein ROI image I 4 (x, y), each pixel contains the values of the RGB (red, green, blue) channels. First, it is converted into a grayscale image I 5 by using the weighted average method:

[0052] I 5 (x, y)=0.299·R(x, y)+0.587·G(x, y)+0.144·B(x, y)

[0053] where R(x, y), G(x, y), and B(x, y) are the pixel values of the red, green, and blue channels in the input image I4(x, y) respectively, and I 5 (x, y) is the converted grayscale value, with a range of [0, 255].

[0054] Subsequently, a median filter is used to remove the noise in the ROI image I 5 (x, y) to obtain the filtered ROI image I 6 (x, y):

[0055] I 6 (x, y)=median(I 1 (x′, y′)|(x′-x) 2 +(y′-y) 2 ≤r2 )

[0056] Here, r is the radius of the window, and median is the operation of taking the median value among the neighborhood pixel values. This step can effectively remove the salt-and-pepper noise in the image while retaining the details of the image edges.

[0057] To enhance the contrast of the image, the image I 6 (x, y) can be processed by the histogram equalization method. The goal of histogram equalization is to adjust the gray-level distribution of the image so that the occurrence frequency of each gray level is as uniform as possible. Its process is given by the following formula:

[0058]

[0059] where r k represents the gray value, p(r i ) is the probability density of the gray value r i , and H(r k ) is the cumulative distribution function (CDF) of the gray value r k . The ROI image I 7 (x, y) after histogram equalization is calculated by the following formula:

[0060]

[0061] where L is the number of gray levels of the image (usually 256). Through this operation, the gray-level distribution of the ROI image becomes more uniform, thereby enhancing the details of the ear vein ROI image.

[0062] In an embodiment of the present invention, the method for identifying a person's identity using a human ear vein image further includes performing feature extraction on the preprocessed human ear vein image to obtain the human ear vein texture structure features. The feature extraction includes: using the MFRAT method to extract the texture structure features of the enhanced human ear vein ROI image to obtain the human ear vein texture structure features.

[0063] Specifically, using the MFRAT method to extract the texture structure features of the enhanced human ear vein ROI image includes the following three steps: scale selection, fractal dimension calculation, and feature extraction and representation.

[0064] 1. Scale selection

[0065] Select scale parameters s = 1, 2, 4, 8, etc. These scale values represent the analysis of the image at different granularities. Smaller scales are used to capture the fine structure of the ear vein, such as tiny vein branches; larger scales are helpful for analyzing the macroscopic distribution pattern of the veins.

[0066] 2. Fractal dimension calculation

[0067] At each selected scale s, the fractal dimension is calculated for the human ear vein ROI image. The box-counting dimension method is adopted in the present invention. That is, the image region is divided into grids of size ∈×i (where ∈ is related to the scale s), and the number of grids N(∈) covering the target texture (human ear vein) in the image is counted. The fractal dimension D is calculated by the formula In a discrete image, a straight line is fitted through the N(∈) values at different scales ∈, and its slope is approximately the fractal dimension.

[0068] Among them, the corresponding grid size ∈ is determined according to the scale s, the human ear vein ROI image is divided into grids, and the number of grids N(∈) covering the vein texture is counted. This process is repeated to obtain the N(∈) values at different scales, and then the fractal dimension at this scale is calculated.

[0069] 3. Feature Extraction and Representation

[0070] The fractal dimensions calculated at different scales are used as texture structure features. These feature values form a feature vector F, F = [D 1 , D 2 , D 4 , D 8 …], where D S represents the fractal dimension at scale s. This feature vector comprehensively reflects the texture complexity and structural information of the human ear vein image at different scales. If the fractal dimension at a certain scale is relatively high, it indicates that the complexity of the human ear vein texture at this scale is relatively high, and there may be more details or irregular structures.

[0071] In one embodiment of the present invention, the method for identifying a person using the human ear vein image of the present invention further includes comparing the human ear vein texture structure features with the pre-stored human ear vein texture structure feature samples, so as to identify the person's identity based on the comparison result. For the human ear vein texture structure feature samples, they will be stored in a dedicated database, and encryption technology is used to encrypt and store the samples to ensure the security and privacy of the data, prevent data leakage, and provide data support for subsequent identity recognition. The comparison includes using a matching algorithm to compare the human ear vein texture structure features and the human ear vein texture structure feature samples, including: comparing the human ear vein texture structure features and the human ear vein texture structure feature samples based on the pixel area to obtain a matching score score; and comparing the matching score score with a preset matching threshold Tk. In the case where score is greater than Tk, the person's identity corresponding to the human ear vein texture structure feature sample is identified; while in the case where score is less than Tk, the person's identity cannot be identified. In addition, the preset matching threshold Tk is usually determined through a large number of experiments and data statistical analysis according to specific application scenarios and requirements. In places with high security requirements, such as identity verification in financial institutions, the threshold will be set higher to reduce misidentification; while in some scenarios with higher requirements for convenience and relatively lower security requirements, such as ordinary attendance systems, the threshold can be appropriately reduced.

[0072] The use of the matching algorithm to compare the human ear vein texture structure features and the human ear vein texture structure feature samples based on the pixel area includes the following three steps: calculating the pixel area, calculating the similarity measure based on the pixel area, and calculating the comprehensive matching score.

[0073] 1. Calculate the pixel area

[0074] Calculate the pixel area of each feature unit (the local image area corresponding to the fractal dimension calculated at each scale) in the human ear vein texture structure feature sample.

[0075] 2. Similarity measure based on pixel area:

[0076] Calculate the Euclidean distance, and calculate the Euclidean distance between the pixel areas of the corresponding feature units of the human ear vein texture structure features and the sample features. Assume the feature vector F = [a 1 , a 2 ,...a n and the sample feature vector S = [b 1 , b 2 ,...b n , where a i and b i are respectively the pixel areas of the i-th feature unit, and the Euclidean distance The smaller the distance, the more similar the pixel areas of the corresponding feature units are.

[0077] 3. Calculate the comprehensive matching score

[0078] Perform a weighted sum on the similarity measurement results of all feature units to obtain the final matching score score. Optionally, if certain feature units are considered more important for human ear vein recognition, higher weights can be assigned to them. Let w i be the weight of the i-th feature unit, and the matching score If the matching score score is higher than the threshold Tk, it can be considered that the human ear vein texture structure features match the sample, that is, the human ear vein is recognized as belonging to the same object; if it is lower than the threshold, it is considered unmatched.

[0079] The present invention also provides a human ear vein recognition system, which includes: the human ear vein image acquisition device of the present invention; and an electronic device, the electronic device includes a processor and a memory, and the memory stores instructions, and the instructions execute the method of using human ear vein images to recognize the identity of a person according to the present invention when executed by the processor.

[0080] As Figure 4 shown, specifically, when the electronic device is a computer, the computer may include an image preprocessing unit 51, an image feature extraction unit 52, and an encoding and matching unit 53. When the computer receives the ear vein image captured by the camera 4, the image is preprocessed in the image preprocessing unit 51: first, ROI extraction is performed on the ear vein image. ROI extraction automatically detects and locates the specific position of the ear vein by using an edge detection algorithm, an image binarization method, and a morphological dilation operation, ensuring the efficiency and accuracy of subsequent steps. The human ear vein recognition system automatically selects the key area of the ear vein in the human ear vein image, eliminates the irrelevant background and noise parts, and ensures that subsequent processing focuses on the vein feature area; then, the image is smoothed by a median filter to eliminate some unwanted noises. After the computer preprocesses the ear vein image, in the image feature extraction unit 52, the texture features of the ear vein are extracted by using a feature extraction filter; the extracted texture features are stored in the database of the computer for future identity authentication. When performing identity authentication, the staff executes an identity authentication instruction through the computer. In the encoding and matching unit 53, the similarity degree between the newly obtained human ear vein texture structure features and the human ear vein texture structure features stored in the database is compared, and a matching score score is obtained. The matching score score is compared with a preset matching threshold Tk. When score is greater than Tk, the identity of the person corresponding to the human ear vein texture structure feature sample is recognized; when score is less than Tk, it indicates that the identity of the person cannot be recognized.

[0081] The method for identifying personnel using human ear vein images according to the present invention can adaptively cope with different image qualities and lighting conditions, ensuring stable and accurate extraction of the ear vein blood vessel region under variable environments. In addition, by enhancing contrast, detecting edges, and performing morphological operations, the method effectively retains the details of the auricle while suppressing noise interference, thereby improving the accuracy of the extraction results. Moreover, the method is not only applicable to the extraction of ear vein images, but also has good scalability and can be applied to other similar image processing tasks to meet different requirements.

[0082] The above description is only some preferred embodiments of the present disclosure and an explanation of the applied technical principles. Those skilled in the art should understand that the scope of the invention involved in the embodiments of the present disclosure is not limited to the technical solutions formed by the specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the above inventive concept. For example, the technical solutions formed by mutually replacing the above features with the technical features (but not limited to) disclosed in the embodiments of the present disclosure that have similar functions.

Claims

1. A human ear vein image acquisition device, comprising: A head-mounted wearing piece (1); A light source assembly (3), the light source assembly (3) being connected to the head-mounted wearing piece (1), and when the head-mounted wearing piece (1) is worn on a person's head, the light source assembly (3) is located behind the person's ear and is used to illuminate the back of the person's ear; A camera (4), the camera (4) being connected to the head-mounted wearable device (1), and when the head-mounted wearable device (1) is worn on a person's head, the camera (4) is located in front of the person's ear and is used to capture an image of the person's ear veins; and A power supply, the power supply is electrically connected to the light source assembly (3) and the camera (4), and is used to supply power to the light source assembly (3) and the camera (4).

2. The human ear vein image acquisition device according to claim 1, characterized in that: The invention also comprises an adjustment component (2), wherein the adjustment component (2) comprises a shell and a first flexible connecting rod (23) and a second flexible connecting rod (24); the shell is fixedly arranged at a position on the head-mounted wearable component (1) corresponding to the position of the human ear; the light source component (3) and the camera (4) are respectively connected to the shell via the first flexible connecting rod (23) and the second flexible connecting rod (24).

3. The human ear vein image acquisition device according to claim 2, characterized in that: The shell of the adjustment component (2) includes a first knob (21) and a second knob (22) which are nested in each other and can rotate independently, the first flexible connecting rod (23) is connected to the first knob (21), and the second flexible connecting rod (24) is connected to the second knob (22), and the power supply is arranged in the shell of the adjustment component (2), wherein wires are arranged in the first flexible connecting rod (23) and the second flexible connecting rod (24), and the light source component (3) and the camera (4) are electrically connected to the power supply in the shell through the wires.

4. The human ear vein image acquisition device according to claim 1, characterized in that: The head-mounted wearing piece (1) is made of elastic material, and / or the head-mounted wearing piece (1) has an adjustment buckle (5), the adjustment buckle (5) is in the form of Velcro, buckle or elastic band, and / or a soft forehead guard (6) is provided on the head-mounted wearing piece (1).

5. The human ear vein image acquisition device according to claim 1, characterized in that: The light source assembly (3) comprises an LED back panel (32) and an LED light source (31) arranged on the LED panel, wherein the LED back panel (32) has a curved shape corresponding to the back of a human ear.

6. A method for identifying a person using an ear vein image, comprising: Receiving a human ear vein image from the human ear vein image acquisition device according to any one of claims 1 to 5; Preprocessing the human ear vein image to obtain a human ear vein ROI image; Performing feature extraction on the human ear vein ROI image to obtain human ear vein texture structure features; and The human ear vein texture mechanism feature is compared with a pre-stored human ear vein texture mechanism feature sample, so as to identify the identity of the person based on the comparison result.

7. The method for identifying a person using an ear vein image according to claim 6, characterized in that: The pre-processing comprises: Performing ROI extraction on the human ear vein image by using an edge detection algorithm, an image binarization method, and a morphological dilation operation to obtain a human ear vein ROI image; Filtering the human ear vein ROI image using a median filter to obtain a filtered human ear vein ROI image; and A contrast enhancement operation is performed on the filtered human ear vein ROI image by using histogram equalization to obtain an enhanced human ear vein ROI image.

8. The method for identifying a person using an ear vein image according to claim 6, characterized in that: The feature extraction includes: extracting the texture structure features of the enhanced human ear vein ROI image using the MFRAT method to obtain the texture structure features of the human ear vein.

9. The method for identifying a person using an ear vein image according to claim 6, characterized in that: The comparison includes comparing the human ear vein texture structure feature with the human ear vein texture structure feature sample using a matching algorithm, including: Comparing the human ear vein texture structure feature with the human ear vein texture structure feature sample based on pixel area to obtain a matching score score; and The matching score score is compared with a preset matching threshold Tk. When the score is greater than Tk, the identity of the person corresponding to the human ear vein texture structure feature sample is identified; and when the score is less than Tk, the identity of the person cannot be identified.

10. A human ear vein recognition system, characterized in that: include: The human ear vein image acquisition device according to any one of claims 1 to 5; and An electronic device comprising a processor and a memory, wherein the memory stores instructions, and when the instructions are executed by the processor, the method for identifying a person by using a human ear vein image according to any one of claims 6 to 9 is executed.

Images