Palm vein image quality detection method, system, medium and electronic equipment

Abstract

The present application provides a palm vein image quality detection method, system, medium, and electronic device. The method comprises the following steps: acquiring a palm vein image; performing an initial image screening on the palm vein image to acquire a palm vein image that is preliminarily determined to be qualified; and performing image quality detection on the palm vein image that is preliminarily determined to be qualified. The image quality detection comprises at least one of the following: image brightness assessment, palm edge assessment, image noise assessment, hand posture estimation, and palm area assessment. By performing initial image screening and image quality detection on the acquired palm vein image, the present application is able to acquire high-quality palm vein images, effectively resolving the problem in the prior art of poor palm vein recognition performance due to quality issues in the palm vein image, and improving the accuracy and reliability of subsequent palm vein recognition using the palm vein image.

Description

Technical Field

[0001] The present application belongs to the field of image processing technology and relates to a palm vein recognition system, and in particular to a palm vein image quality detection method, system, medium and electronic equipment. Background Art

[0002] In palm vein recognition systems, image quality significantly impacts palm vein recognition performance. A range of quality issues may occur during palm vein image capture, which can adversely impact the system's performance, leading to poor recognition performance. Summary of the Invention

[0003] The purpose of this application is to provide a palm vein image quality detection method, system, medium and electronic device to solve the problem of poor palm vein recognition performance due to quality problems of palm vein images in the prior art.

[0004] In a first aspect, the present application provides a palm vein image quality detection method, which includes the following steps: obtaining a palm vein image; performing an initial image screening on the palm vein image to obtain a palm vein image that is preliminarily determined to be qualified; performing image quality detection on the palm vein image that is preliminarily determined to be qualified; the image quality detection includes at least any one of the following: image brightness evaluation, palm edge evaluation, image noise evaluation, hand posture estimation, and palm area evaluation.

[0005] In this application, by performing preliminary image screening and image quality detection on the acquired palm vein images, it is possible to obtain high-quality palm vein images, effectively solving the problem in the prior art of poor palm vein recognition performance due to quality problems of the palm vein images, and improving the accuracy and reliability of subsequent palm vein recognition using the palm vein images.

[0006] In an implementation of the first aspect, the initial screening of the palm vein image includes the following steps: obtaining a first edge area and a second edge area of ​​preset width from the palm vein image; the first edge area and the second edge area are both located at the edge of the palm vein image and are relative to each other; calculating the average pixel value of the first edge area and the second edge area; if the average pixel value meets a preset pixel condition, preliminarily determining that the palm vein image is qualified; if the average pixel value does not meet the preset pixel condition, preliminarily determining that the palm vein image is unqualified.

[0007] In an implementation of the first aspect, the initial screening of the palm vein image includes the following steps: obtaining a central area of ​​the palm vein image; calculating an average grayscale value and a maximum grayscale value of the central area; if the average grayscale value and the maximum grayscale value meet a preset grayscale condition, preliminarily determining that the palm vein image is qualified; if the average grayscale value and the maximum grayscale value do not meet the preset grayscale condition, preliminarily determining that the palm vein image is unqualified.

[0008] In an implementation of the first aspect, the image quality detection of the palm vein image preliminarily determined to be qualified includes the following steps: performing image brightness evaluation on the palm vein image preliminarily determined to be qualified to obtain a first image whose image brightness satisfies a preset brightness condition; performing palm edge evaluation on the first image to obtain a second image whose palm edge satisfies a preset edge condition; performing image noise evaluation on the second image to obtain a third image whose image noise satisfies a preset noise condition; performing hand posture estimation on the third image to obtain a fourth image whose hand posture satisfies a preset hand posture condition; performing palm area evaluation on the fourth image to obtain a fifth image whose palm area satisfies the preset palm area condition; the fifth image is the palm vein image finally determined to be qualified.

[0009] In this implementation, the image quality of the palm vein image is detected in the order of image brightness assessment, palm edge assessment, image noise assessment, hand posture estimation, and palm area assessment. This allows for the rapid filtering of unqualified images before ultimately obtaining qualified, high-quality palm vein images, thereby significantly improving the user experience.

[0010] In an implementation of the first aspect, when the image quality detection includes hand posture estimation, hand posture estimation is performed on the palm vein image that is preliminarily determined to be qualified, including the following steps: judging whether a hand exists based on the palm vein image that is preliminarily determined to be qualified; when the hand exists, estimating multiple hand joints; judging the palm posture based on the multiple hand joints to determine whether the hand posture meets a preset hand posture condition; the palm posture judgment includes at least any one of the following: judging whether the palm is complete, judging whether the palm is a left palm or a right palm, and judging whether the palm is fully open; wherein, judging whether the palm is fully open includes: judging whether the curvature of each finger meets a preset curvature condition, and / or calculating the angle between two adjacent fingers to determine whether the angle meets a preset angle condition.

[0011] In an implementation of the first aspect, when the image quality detection includes palm area assessment, performing palm area assessment on the palm vein image that is preliminarily determined to be qualified includes the following steps: calculating a radius of the palm area based on a plurality of hand joint points; the palm area being the maximum inscribed circle of the palm; and determining whether the radius satisfies a preset radius condition, thereby determining whether the palm area satisfies the preset palm area condition.

[0012] In an implementation of the first aspect, when the image quality detection includes palm edge assessment, performing palm edge assessment on the palm vein image preliminarily determined to be qualified includes the following steps: acquiring a palm edge detection image based on the palm vein image preliminarily determined to be qualified; and determining whether the palm edge meets a preset edge condition based on the palm edge detection image and the palm vein image preliminarily determined to be qualified.

[0013] In a second aspect, the present application provides a palm vein image quality detection system, which includes: an image acquisition module for acquiring a palm vein image; an image screening module for performing an initial image screening on the palm vein image to obtain a palm vein image that is preliminarily determined to be qualified; an image detection module for performing image quality detection on the palm vein image that is preliminarily determined to be qualified; the image quality detection includes at least any one of the following: image brightness evaluation, palm edge evaluation, image noise evaluation, human hand posture estimation, and palm area evaluation.

[0014] In a third aspect, the present application provides an electronic device, comprising: a memory for storing a computer program; and a processor for executing the computer program so that the electronic device performs the above-mentioned palm vein image quality detection method.

[0015] In a fourth aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, which implements the above-mentioned palm vein image quality detection method when executed by an electronic device.

[0016] As described above, the palm vein image quality detection method, system, medium and electronic device described in this application have the following features:

[0017] Beneficial effects:

[0018] Compared with the existing technology, the present application provides a palm vein image quality detection method, which realizes the quality assessment of palm vein images in the preprocessing stage of the palm vein recognition system, can filter out palm vein images of unqualified quality and obtain qualified, high-quality palm vein images, thereby improving the accuracy and reliability of subsequent palm vein recognition; at the same time, high-quality palm vein images are also conducive to improving the efficiency of palm vein recognition. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1a Shown is a schematic diagram of an image of a palm involved in the initial image screening described in an embodiment of the present application.

[0020] Figure 1b to Figure 1g Shown is a schematic diagram of an image involved in the image brightness and darkness evaluation described in an embodiment of the present application.

[0021] Figure 1h to Figure 1k Shown is a schematic diagram of an image involved in palm edge assessment according to an embodiment of the present application.

[0022] Figure 11 and Figure 1m Shown is a schematic diagram of an image involved in image noise evaluation according to an embodiment of the present application.

[0023] Figure 1n Shown is a schematic diagram of the 21 hand joints described in an embodiment of the present application.

[0024] Figure 1o and Figure 1p Shown is a schematic diagram of images involved in palm area assessment according to an embodiment of the present application.

[0025] Figure 1q Shown is a block diagram of the working principle of the palm vein image quality detection method described in an embodiment of the present application.

[0026] Figure 2 Shown is a flowchart of the palm vein image quality detection method described in an embodiment of the present application.

[0027] Figure 3 Shown is a flowchart of obtaining palm vein images according to an embodiment of the present application.

[0028] Figure 4 Shown is a flowchart of performing preliminary screening of palm vein images according to an embodiment of the present application.

[0029] Figure 5 Shown is a flowchart of performing preliminary screening of palm vein images according to another embodiment of the present application.

[0030] Figure 6 Shown is a flowchart of performing image quality detection on a palm vein image that is preliminarily determined to be qualified, as described in an embodiment of the present application.

[0031] Figure 7 Shown is a flowchart of performing palm edge assessment on a palm vein image that is preliminarily determined to be qualified, as described in an embodiment of the present application.

[0032] Figure 8Shown is a flowchart of acquiring a palm edge detection image based on a palm vein image that is preliminarily determined to be qualified, as described in an embodiment of the present application.

[0033] Figure 9 Shown is a flowchart of performing image noise evaluation on a second image and obtaining a third image whose image noise satisfies a preset noise condition, as described in an embodiment of the present application.

[0034] Figure 10 Shown is a flowchart of estimating a human hand posture based on a palm vein image that is preliminarily determined to be qualified, as described in an embodiment of the present application.

[0035] Figure 11 Shown is a flowchart of performing palm area evaluation on a palm vein image that is preliminarily determined to be qualified, as described in an embodiment of the present application.

[0036] Figure 12 Shown is a structural schematic diagram of the palm vein image quality detection system described in an embodiment of the present application. DETAILED DESCRIPTION

[0037] The following describes the embodiments of the present application through specific examples. Those skilled in the art can easily understand the other advantages and effects of the present application from the content disclosed in this specification. The present application can also be implemented or applied through other different specific embodiments. The details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present application. It should be noted that the following embodiments and features in the embodiments can be combined with each other unless they conflict.

[0038] It should be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present application. Therefore, the illustrations only show components related to the present application and are not drawn according to the number, shape and size of components in actual implementation. In actual implementation, the type, quantity and proportion of each component can be changed at will, and the component layout type may also be more complicated.

[0039] See Figures 1a to 1q ,and Figures 2 to 12 . The following embodiments of the present application provide a palm vein image quality detection method, system, medium and electronic device. Compared with the existing technology, the present application provides a palm vein image quality detection method, which realizes the quality assessment of the palm vein image in the preprocessing stage of the palm vein recognition system, can filter out palm vein images of unqualified quality, and obtain qualified, high-quality palm vein images, thereby improving the accuracy and reliability of subsequent palm vein recognition; at the same time, high-quality palm vein images are also conducive to improving the efficiency of palm vein recognition.

[0040] In one embodiment, the palm vein image quality detection method provided in this application is applied to an electronic device.

[0041] As shown in FIG1 , in this embodiment, the working principle of the palm vein image quality detection method is as follows:

[0042] 1. Initial Image Screening

[0043] Since the camera (camera collects the original palm image, it is necessary to grayscale process the original palm image to obtain Figure 1a The palm vein image shown in the figure is collected very quickly, and the image screening function can be used to roughly eliminate images with unqualified image quality. This not only allows qualified images to be obtained as quickly as possible, but also improves the success rate of subsequent processing. Specifically,

[0044] (11) Figure 1a As shown, edge regions with a width of 2 pixels are obtained at the left and right edges of the image, and the average pixel values ​​of the two edge regions are calculated.

[0045] When no palm enters the two edge areas and the influence of environmental factors is ignored, the average pixel value of the two edge areas is less than 15; if the palm partially enters the two edge areas, the average pixel value of the two edge areas will increase.

[0046] A threshold value is set. If the average pixel value of the two edge areas is greater than the threshold, the palm is judged to be outside the image range; conversely, if the average pixel value of the two edge areas is not greater than the threshold, the palm is judged to be within the image range.

[0047] (12) Figure 1a As shown, first, the middle position of the image height is found from top to bottom; then, at this middle height, along the X-axis direction of the image, at one-third of the image width, a square area with a length of one-third of the image width is obtained; finally, the average grayscale value and the maximum grayscale value of this square area are obtained.

[0048] If the average grayscale value is greater than 50 and the maximum grayscale value is less than 210, the image is appropriate, otherwise the image is too dark or overexposed.

[0049] After the above determination, if the palm does not exceed the image range and the image is appropriate, the palm vein image is preliminarily determined to be qualified.

[0050] The image initial screening function takes a very short time (a few milliseconds) to process the left and right edge areas and the middle square area. Therefore, it can quickly and effectively screen out more qualified images for the next function processing, thus saving time for subsequent processing.

[0051] After an initial image screening, a palm vein image that is preliminarily qualified is obtained. This preliminarily qualified palm vein image is then subjected to image quality testing. Specifically, the evaluation focuses on the following aspects: whether the image is overexposed or too dark, whether the palm edge is clear, whether it contains noise, whether the palm is intact, whether the palm is fully open, and whether the palm area is too small.

[0052] 2. Image brightness evaluation

[0053] Because each point in a scene has different colors and brightness, each point in a captured black and white image appears to have varying shades of gray. The relationship between white and black is divided into several levels, called grayscale, based on a logarithmic relationship. The grayscale range generally ranges from 0 to 255, with white at 255 and black at 0, so black and white images are also called grayscale images.

[0054] Overexposure or darkening of the palm vein image will affect the extraction and processing of palm vein features. Therefore, an image brightness evaluation algorithm is used to filter out overexposed and darkened images, leaving normal images.

[0055] like Figures 1b to 1g As shown in the figure, the grayscale value of the pixel points of the entire palm vein image is determined by the histogram; wherein, Figure 1b The grayscale value of the pixel in the image is as follows Figure 1c As shown; Figure 1d The grayscale value of the pixel in the image is as follows Figure 1e As shown; Figure 1f The grayscale value of the pixel in the image is as follows Figure 1g shown.

[0056] According to the histogram display of images with different brightness, the image pixels with grayscale values ​​less than 100 are classified as too dark pixels, and the pixels with grayscale values ​​greater than 200 are classified as too bright pixels.

[0057] Set the too bright threshold and too dark threshold, traverse all the pixels of the image, if the ratio of the too dark pixel to the total pixel is greater than the too dark threshold, the image is judged to be too dark (such as Figure 1b If the ratio of overbright pixels to total pixels is greater than the overbright threshold, the image is judged to be exposed (as shown in Figure 1f This allows you to filter out images that are judged to be too dark or overexposed.

[0058] 3. Palm Edge Assessment

[0059] If the palm vein image captured by the camera does not contain complete palm edge information, it will affect the palm vein recognition effect. Therefore, a palm edge evaluation algorithm is used to filter out inappropriate images.

[0060] like Figures 1h to 1k As shown, Figure 1hThe palm vein image I in the image is subjected to Sobel edge detection. The Sobel operator is a combination of Gaussian smoothing and differential operation. It has strong noise resistance and many uses. The Sobel operator uses fast convolution function for processing. It is simple, fast and effective, and the influence of the pixel position in the field is weighted. It has a good effect in reducing the degree of edge blur. The Sobel operator extracts the edge of the image in three steps: extraction Figure 1h The horizontal edge of the palm vein image in Figure 1i shown; extract Figure 1h The vertical edge of the palm vein image in Figure 1j As shown; the edge information of the two directions is weighted averaged to obtain the edge of the entire image, that is, the edge detection image E, as shown Figure 1k shown.

[0061] Calculate the grayscale mean of image E to be m E And the grayscale mean of palm vein image I is m I , calculate the Q value, the formula is as follows:

[0062]

[0063] Through experiments, it is found that the higher the Q value (the closer to 1), the clearer the edge.

[0064] The threshold k is set through experiments. If the Q value is less than k, it is determined that the edge of the palm is unclear.

[0065] IV. Image Noise Evaluation

[0066] If the captured image contains too much noise, it will affect subsequent image processing.

[0067] like Figure 11 and Figure 1m As shown, in this embodiment, first, the noise of the original acquired image L is removed by a filtering algorithm to obtain a filtered image K; then, the mean square error (MSE) of the filtered image K and the original acquired image L is calculated; if the MSE is smaller, the noise of the original acquired image L is smaller.

[0068] MSE stands for "Mean Square Error". It is one of the indicators for measuring image quality. Its principle is to sum and average the squares of the differences between the true value (here refers to the original acquired image L) and the predicted value (here refers to the filtered image K). The formula is as follows:

[0069]

[0070] Where M is the total number of pixels of the original acquired image L, and N is the total number of pixels of the filtered image K.

[0071] The smaller the MSE value, the more similar the images are, that is, the smaller the noise.

[0072] In this embodiment, different types of noise are artificially added, and the mean square error is calculated using the above method. The MSE threshold is selected as 4. If the calculated MSE value is greater than 4, it is determined that the image noise is too large.

[0073] 5. Hand Pose Estimation

[0074] The hand posture estimation algorithm is used to determine whether the hand exists, whether the palm is complete, whether each finger is complete, and whether the fingers are open.

[0075] Only images with a complete and fully open palm can correctly capture palm vein feature information.

[0076] After acquiring the palm vein image, first, it is determined whether a hand exists in the palm vein image; then, after determining that a hand exists in the palm vein image, the palm vein image is used to estimate the hand joint points.

[0077] like Figure 1n As shown, in this embodiment, the Mediapipe framework is used to predict and evaluate wrist joints.

[0078] MediaPipe is a data stream processing and machine learning application development framework developed and open-sourced by Google. It is a graph-based data processing pipeline used to build data sources in various forms. It solves these problems by abstracting various perception models into modules and connecting them into maintainable graphs. Among them, MediaPipe Hands is a high-fidelity hand and finger tracking solution that uses machine learning (ML) to infer 21 3D landmarks of the hand from a single frame.

[0079] After MediaPipe Hands estimates hand joint points for the entire image, its hand landmark model accurately locates 21 3D hand joint coordinates within the detected hand region through regression (i.e., direct coordinate prediction). This model learns a consistent internal hand pose representation that is robust even to partially visible hands and self-occlusion. The coordinate locations of the 21 detected points are shown in Figure 1n.

[0080] like Figure 1n As shown, the 21 hand joints are:

[0081] Point 0 is the wrist point;

[0082] {1,2,3,4} are the key points on the wrist to thumb skeleton;

[0083] {5,6,7,8} are the key points on the wrist to index finger skeleton;

[0084] {9,10,11,12} are the key points on the skeleton from wrist to middle finger;

[0085] {13,14,15,16} are the key points on the skeleton from wrist to ring finger;

[0086] {17,18,19,20} are the key points on the skeleton from wrist to little finger;

[0087] Among them, {4, 8, 12, 16, 20} are fingertip points, and {5, 9, 13, 17} are finger base points.

[0088] Each hand joint point corresponds to a coordinate point.

[0089] The 7 hand joint points 0, 5, 9, 13, 17, 4, and 20 are judged to be within the image range to determine whether the palm is complete; if one of the 7 hand joint points is not within the image range, the palm is determined to be incomplete.

[0090] By calculating the line between hand joint point 0 and hand joint point 9 and the x-axis ( Figure 1n The angle 1 between the two points (horizontally to the right in the figure) and the angle 2 between the hand joint points 0 and 4 and the x-axis; if it is an image of the palm of the left hand, the angle 1 is greater than the angle 2 and less than the angle 2 plus 90 degrees; if it is an image of the palm of the right hand, the angle 1 is less than the angle 2 and greater than the angle 2 minus 90 degrees.

[0091] When determining whether a palm is fully open, two main aspects are considered. The first is whether each finger is excessively bent. The second is to calculate the angle between each finger (index finger and middle finger, middle finger and ring finger, ring finger and pinky finger) and compare it with the finger opening threshold. If both conditions are met, the palm is judged to be fully open; otherwise, it will affect subsequent palm vein recognition.

[0092] In this embodiment, for a finger, two adjacent wrist joints are connected to form a line segment, and the angle between the two adjacent line segments is calculated. If the angle is less than 120°, the finger is defined as over-bent.

[0093] Such as, Figure 1nAs shown, for the middle finger, hand joint point 9 and hand joint point 10 are connected to form a first line segment; hand joint point 10 and hand joint point 11 are connected to form a second line segment; hand joint point 11 and hand joint point 12 are connected to form a third line segment; the angle between the first line segment and the second line segment is calculated (the angle is calculated according to the coordinate points of the hand joint points), and the angle between the second line segment and the third line segment are calculated; it is determined whether the two angles are both less than 120°. If there is an angle less than 120°, it is determined that the middle finger is over-bent; only when the two angles are not less than 120°, it is determined that the middle finger is not over-bent, that is, the middle finger is normal (the middle finger is fully open).

[0094] 6. Palm Area Assessment

[0095] If the palm is far away from the camera and the captured palm image is small, the extracted palm area will also be small. The smaller palm area does not contain rich feature information, which will affect the palm vein recognition effect. Therefore, in this embodiment, a palm area evaluation algorithm is used to filter out images with small palm areas.

[0096] like Figure 1o As shown, first, the palm area evaluation algorithm performs binary processing on the palm vein image to obtain Figure 1o The binary image shown; then, the palm area evaluation algorithm performs weighted calculation on hand joint point 0, hand joint point 5, hand joint point 9, hand joint point 13 and hand joint point 17 to obtain the palm area radius; if the radius is less than a certain threshold, it is determined that the palm area is too small.

[0097] like Figure 1p As shown, the maximum inscribed circle of the palm is drawn according to the radius of the palm area obtained by the above calculation.

[0098] like Figure 1q As shown, the palm vein image quality detection method provided in this embodiment performs image quality detection on the palm vein image (such as Figure 1q As shown in the dotted box, image brightness evaluation, palm edge evaluation, and image noise evaluation are first performed. Because these three evaluation modules have a fast processing speed, unqualified images can be quickly filtered out before finally obtaining qualified images. Then, the module enters the hand posture estimation module with a relatively slow processing speed, which can better improve the user experience. Finally, the palm area evaluation module is entered. This is because the palm area evaluation algorithm needs to rely on the hand joints estimated in the hand posture estimation algorithm.

[0099] The above evaluation algorithm enables the quality assessment of palm vein images during the preprocessing stage of the palm vein recognition system. This can filter out unqualified palm vein images and obtain qualified, high-quality palm vein images, thereby improving the accuracy and reliability of subsequent palm vein recognition. At the same time, high-quality palm vein images are also conducive to improving the efficiency of palm vein recognition.

[0100] The technical solutions in the embodiments of the present application will be described in detail below with reference to the accompanying drawings in the embodiments of the present application.

[0101] like Figure 2 As shown, this embodiment provides a palm vein image quality detection method; specifically, the palm vein image quality detection method includes the following steps:

[0102] Step S1: Acquire a palm vein image.

[0103] like Figure 3 As shown, in one embodiment, obtaining a palm vein image includes the following steps:

[0104] Step S11: Acquire the original palm image.

[0105] In one embodiment, the original palm image is acquired from a camera.

[0106] Step S12: grayscale processing is performed on the original palm image to obtain a palm vein image.

[0107] Step S2: performing a preliminary screening on the palm vein images to obtain palm vein images that are preliminarily determined to be qualified.

[0108] like Figure 4 As shown, in one embodiment, the performing of preliminary image screening on the palm vein image includes the following steps:

[0109] Step S21 : Acquire a first edge region and a second edge region of preset width from the palm vein image.

[0110] It should be noted that the preset width is pre-set, and its specific setting is not a condition to limit this application. In actual application, it can be set according to the specific application scenario; for example, in the above embodiment, the preset width is set to 2 pixels.

[0111] In one embodiment, the first edge region and the second edge region are both located at edges of the palm vein image and are opposite to each other.

[0112] Step S22: Calculate the average pixel value of the first edge area and the second edge area.

[0113] If the average pixel value meets the preset pixel condition, the palm vein image is preliminarily determined to be qualified; if the average pixel value does not meet the preset pixel condition, the palm vein image is preliminarily determined to be unqualified.

[0114] In one embodiment, the preset pixel condition is: the average pixel value is not greater than a preset pixel threshold.

[0115] It should be noted that the preset pixel threshold is pre-set, and its specific setting is not a condition to limit this application. In actual application, it can be set according to the specific application scenario.

[0116] like Figure 5 As shown, in one embodiment, the performing of preliminary image screening on the palm vein image includes the following steps:

[0117] Step S23: Acquire the central area of ​​the palm vein image.

[0118] In one embodiment, at the middle height of the palm vein image, along the X-axis direction of the palm vein image, at one-third of the width of the palm vein image, a square area with a length of one-third of the image width is obtained, which is the central area of ​​the palm vein image.

[0119] Step S24: Calculate the average grayscale value and the maximum grayscale value of the central area.

[0120] If the average grayscale value and the maximum grayscale value meet the preset grayscale condition, the palm vein image is preliminarily determined to be qualified; if the average grayscale value and the maximum grayscale value do not meet the preset grayscale condition, the palm vein image is preliminarily determined to be unqualified.

[0121] In one embodiment, the preset grayscale condition is: the average grayscale value is greater than a first preset grayscale threshold, and the maximum grayscale value is less than a second preset grayscale threshold.

[0122] It should be noted that the first preset grayscale threshold and the second preset grayscale threshold are both pre-set, and their specific settings are not used as conditions to limit this application (for example, in the above embodiment, the first preset grayscale threshold is 50 and the second preset grayscale threshold is 210). In actual applications, they can be set according to specific application scenarios; the second preset grayscale threshold is greater than the first preset grayscale threshold.

[0123] Specifically, if the average grayscale value and the maximum grayscale value meet the preset grayscale conditions, the palm vein image is determined to be appropriate; otherwise, the palm vein image is determined to be too dark or overexposed.

[0124] In one embodiment, the initial screening of the palm vein image includes the above-mentioned steps S21 to S24, that is, the palm vein image is preliminarily determined to be qualified only when the average pixel value meets the preset pixel condition and the average grayscale value and the maximum grayscale value meet the preset grayscale condition.

[0125] Step S3: performing image quality detection on the palm vein image that is initially determined to be qualified.

[0126] In one embodiment, the image quality detection includes at least but not limited to any one of the following: image brightness assessment, palm edge assessment, image noise assessment, hand posture estimation, and palm area assessment.

[0127] It should be noted that when performing image quality detection on a palm vein image that has been preliminarily determined to be qualified, the order of image brightness assessment, palm edge assessment, image noise assessment, hand posture estimation, and palm area assessment can be adjusted or swapped in actual application scenarios according to the specific application scenarios.

[0128] like Figure 6 As shown, in one embodiment, the image quality detection of the palm vein image that is preliminarily determined to be qualified includes the following steps:

[0129] Step S31 : performing image brightness evaluation on the palm vein image that is preliminarily determined to be qualified, and obtaining a first image whose image brightness satisfies a preset brightness condition.

[0130] In one embodiment, the palm vein image that is initially determined to be qualified is evaluated for brightness and darkness, and the grayscale value of each pixel of the palm vein image is obtained.

[0131] In one embodiment, if the grayscale value of a pixel is less than a first preset value, the pixel is determined to be an overly dark pixel.

[0132] In one embodiment, if the grayscale value of a pixel is greater than a second preset value, the pixel is determined to be an overbright pixel.

[0133] It should be noted that the first preset value and the second preset value are both pre-set, and their specific settings are not used as conditions to limit this application (for example, in the above embodiment, the first preset value is 100 and the second preset value is 200). In actual applications, they can be set according to specific application scenarios; the second preset value is greater than the first preset value.

[0134] In one embodiment, the preset brightness condition is: the ratio of too dark pixels in the image to the total pixels in the image is less than a preset too dark threshold, and the ratio of too bright pixels in the image to the total pixels in the image is less than a preset too bright threshold.

[0135] It should be noted that the preset too-dark threshold and the preset too-bright threshold are also pre-set. Their specific settings are not a condition to limit this application. In actual applications, they can be set according to specific application scenarios.

[0136] Step S32: perform palm edge evaluation on the first image to obtain a second image in which the palm edge meets a preset edge condition.

[0137] like Figure 7 As shown, in one embodiment, when the image quality detection includes palm edge assessment, performing palm edge assessment on the palm vein image that is initially determined to be qualified includes the following steps:

[0138] Step S321: Acquire a palm edge detection image based on the palm vein image that is preliminarily determined to be qualified.

[0139] like Figure 8 As shown, in one embodiment, obtaining a palm edge detection image based on the palm vein image that is preliminarily determined to be qualified includes the following steps:

[0140] Step S3211: Acquire a horizontal edge based on the palm vein image that is preliminarily determined to be qualified.

[0141] Step S3212: Acquire vertical edges based on the palm vein image that is preliminarily determined to be qualified.

[0142] Step S3213: Perform weighted averaging on the horizontal edges and the vertical edges to obtain a palm edge detection image.

[0143] Step S322: Determine whether the palm edge meets a preset edge condition based on the palm edge detection image and the palm vein image that is initially determined to be qualified.

[0144] In one embodiment, the Q value is calculated based on the palm edge detection image and the palm vein image that is initially determined to be qualified. The specific calculation formula is as follows:

[0145]

[0146] Among them, m E Represents the grayscale mean of the palm edge detection image; m I Indicates the grayscale mean of the palm vein image or the first image that is preliminarily determined to be qualified.

[0147] In this embodiment, judging whether the palm edge meets a preset edge condition based on the palm edge detection image and the palm vein image that is preliminarily determined to be qualified includes: judging whether the Q value meets the preset edge condition.

[0148] In one embodiment, the preset edge condition is: the Q value is greater than a preset edge threshold.

[0149] Specifically, if the Q value is greater than the preset edge threshold, it is determined that the palm edge meets the preset edge condition, that is, the palm edge is clear; conversely, if the Q value is not greater than the preset edge threshold, it is determined that the palm edge does not meet the preset edge condition, that is, the palm edge is unclear.

[0150] It should be noted that the preset edge threshold is pre-set, and its specific setting is not a condition to limit this application. In actual application, it can be set according to the specific application scenario.

[0151] Step S33: performing image noise evaluation on the second image to obtain a third image whose image noise satisfies a preset noise condition.

[0152] like Figure 9 As shown, in one embodiment, performing image noise evaluation on the second image and obtaining a third image whose image noise satisfies a preset noise condition includes the following steps:

[0153] Step S331: Filter the second image to obtain a filtered image.

[0154] Step S332: Calculate the mean square error of the filtered image and the second image to obtain the mean square error of the filtered image and the second image.

[0155] Specifically, the formula for calculating the mean square error is as follows:

[0156]

[0157] Wherein, M represents the total number of pixels of the second image L; N represents the total number of pixels of the filtered image K; and MSE represents the mean square error between the filtered image K and the second image L.

[0158] Step S333: Determine whether the mean square error meets the preset noise condition.

[0159] In one embodiment, the preset noise condition is: a mean square error is less than a preset noise threshold.

[0160] It should be noted that the preset noise threshold is pre-set, and its specific setting is not a condition to limit this application. In actual application, it can be set according to the specific application scenario (for example, in the above embodiment, the corresponding preset noise threshold is 4).

[0161] Specifically, when the mean square error satisfies the preset noise condition, it is determined that the image noise satisfies the preset noise condition; when the mean square error does not satisfy the preset noise condition, it is determined that the image noise does not satisfy the preset noise condition.

[0162] Step S34: Estimating the hand posture of the third image to obtain a fourth image in which the hand posture meets a preset hand posture condition.

[0163] like Figure 10 As shown, in one embodiment, when the image quality detection includes human hand posture estimation, performing human hand posture estimation on the palm vein image that is initially determined to be qualified includes the following steps:

[0164] Step S341: Determine whether a hand exists based on the palm vein image that is preliminarily determined to be qualified.

[0165] When the hand exists, execute step S342.

[0166] Step S342: Estimate multiple hand joint points.

[0167] In one embodiment, the Mediapipe framework is used to estimate multiple wrist joints.

[0168] In one embodiment, in step S342, 21 hand joint points are estimated, as follows: Figure 1n shown.

[0169] Step S343: judging the palm posture based on the multiple hand joint points to determine whether the hand posture meets the preset hand posture condition.

[0170] In one embodiment, the palm posture determination includes at least but not limited to any one of the following: determining whether the palm is intact, determining whether the palm is the left palm or the right palm, and determining whether the palm is fully open.

[0171] In one embodiment, determining whether the palm is intact includes determining whether seven hand joints 0, 5, 9, 13, 17, 4, and 20 are within the image (herein referred to as the "third image") to determine whether the palm is intact.

[0172] If one of the seven hand joints is not within the image range, the palm is determined to be incomplete.

[0173] In one embodiment, determining whether the palm is the left palm or the right palm includes: calculating the line between the hand joint point 0 and the hand joint point 9 and Figure 1n The angle between the horizontal right direction is 1, the angle between the hand joint point 0 and the hand joint point 4 and Figure 1n The angle 2 between the horizontal right direction in the image; if angle 1 is greater than angle 2 and less than angle 2 plus 90 degrees, it is the palm image of the left hand; if angle 1 is less than angle 2 and greater than angle 2 minus 90 degrees, it is the palm image of the right hand.

[0174] In one embodiment, determining whether the palm is fully open includes: when determining whether the palm is fully open, two aspects are mainly considered. The first aspect is whether each finger is excessively bent; the second aspect is calculating the angle between each finger (index finger and middle finger, middle finger and ring finger, ring finger and pinky finger), and comparing it with the finger opening threshold; if both aspects meet the conditions, it is determined that the palm is fully open, otherwise it will affect the subsequent palm vein recognition.

[0175] In one embodiment, determining whether the palm is fully open includes: determining whether the curvature of each finger meets a preset curvature condition, and / or calculating the angle between two adjacent fingers to determine whether the angle meets a preset angle condition.

[0176] In one embodiment, the preset bending condition is: the bending degree is greater than the preset bending angle.

[0177] Specifically, if the curvature of the finger is greater than a preset curvature angle, it is determined that the finger is not over-bent; conversely, if the curvature of the finger is not greater than the preset curvature angle, it is determined that the finger is over-bent.

[0178] It should be noted that the preset bending angle is pre-set, and its specific setting is not a condition to limit this application. In actual application, it can be set according to the specific application scenario (for example, in the above embodiment, the corresponding preset bending angle is 120°).

[0179] In one embodiment, calculating the angle between two adjacent fingers includes respectively calculating the angles between the index finger and the middle finger, the middle finger and the ring finger, and the ring finger and the little finger.

[0180] In one embodiment, the preset angle condition is: the angle is greater than a preset angle threshold.

[0181] It should be noted that the preset angle threshold is pre-set, and its specific setting is not a condition to limit this application. In actual application, it can be set according to the specific application scenario.

[0182] Step S35: performing palm area evaluation on the fourth image to obtain a fifth image whose palm area satisfies a preset palm area condition.

[0183] In this embodiment, the fifth image is the palm vein image that is finally determined to be qualified.

[0184] Considering that image brightness assessment, palm edge assessment, and image noise assessment have relatively high processing speeds, and palm area assessment relies on the hand joints estimated in the hand posture estimation algorithm, in this embodiment, image quality detection of the palm vein image is performed in the order of image brightness assessment, palm edge assessment, image noise assessment, hand posture estimation, and palm area assessment. This allows for rapid filtering out of unqualified images before ultimately obtaining qualified, high-quality palm vein images, thereby significantly improving the user experience.

[0185] like Figure 11 As shown, in one embodiment, when the image quality detection includes palm area assessment, performing palm area assessment on the palm vein image that is initially determined to be qualified includes the following steps:

[0186] Step S351: Calculate the radius of the palm area based on the multiple hand joint points.

[0187] Specifically, weighted calculation is performed on hand joint point 0, hand joint point 5, hand joint point 9, hand joint point 13, and hand joint point 17 to obtain the radius of the palm area.

[0188] In this embodiment, the palm area is the largest inscribed circle of the palm.

[0189] Step S352: Determine whether the radius meets a preset radius condition, so as to determine whether the palm area meets a preset palm area condition.

[0190] In one embodiment, the preset palm area condition is: the radius of the palm area satisfies a preset radius condition.

[0191] In one embodiment, the preset radius condition is: the radius is greater than a preset radius threshold.

[0192] It should be noted that the preset radius threshold is pre-set, and its specific setting is not a condition to limit this application. In actual application, it can be set according to the specific application scenario.

[0193] In this embodiment, if the radius satisfies the preset radius condition, it is determined that the palm area satisfies the preset palm area condition; if the radius does not satisfy the preset radius condition, it is determined that the palm area does not satisfy the preset palm area condition.

[0194] The palm vein image quality detection method provided in this application is applied to a palm vein recognition system. It can pre-detect the quality of the palm vein image during the preprocessing stage of the palm vein recognition system, thereby filtering out unqualified, low-quality palm vein images and obtaining qualified, high-quality palm vein images. The high-quality palm vein images are then used for subsequent palm vein recognition, thereby improving the accuracy, reliability, and efficiency of palm vein recognition.

[0195] The protection scope of the palm vein image quality detection method described in the embodiment of the present application is not limited to the execution order of the steps listed in this embodiment. All solutions implemented by adding, subtracting, or replacing steps in the prior art based on the principles of the present application are included in the protection scope of the present application.

[0196] The embodiments of the present application also provide a palm vein image quality detection system, which can implement the palm vein image quality detection method described in the present application. However, the implementation device of the palm vein image quality detection method described in the present application includes but is not limited to the structure of the palm vein image quality detection system listed in this embodiment. Any structural deformation and replacement of the existing technology made according to the principles of the present application are included in the scope of protection of the present application.

[0197] like Figure 12 As shown, this embodiment provides a palm vein image quality detection system, the palm vein image quality detection system comprising:

[0198] The image acquisition module 121 is used to acquire palm vein images.

[0199] The image preliminary screening module 122 is used to perform preliminary screening on the palm vein images to obtain palm vein images that are preliminarily determined to be qualified.

[0200] The image detection module 123 is used to perform image quality detection on the palm vein image that is preliminarily determined to be qualified; the image quality detection includes at least any one of the following: image brightness assessment, palm edge assessment, image noise assessment, human hand posture estimation and palm area assessment.

[0201] It should be noted that the structures and principles of the image acquisition module 121, the image screening module 122 and the image detection module 123 correspond one-to-one to the steps (steps S1 to S3) in the above-mentioned palm vein image quality detection method, so they are not repeated here.

[0202] In the several embodiments provided in this application, it should be understood that the disclosed systems, devices or methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of modules / units is only a logical function division. There may be other division methods in actual implementation. For example, multiple modules or units can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or modules or units, which can be electrical, mechanical or other forms.

[0203] The modules / units described as separate components may or may not be physically separate, and the components displayed as modules / units may or may not be physical modules, that is, they may be located in one place or distributed across multiple network elements. Some or all of the modules / units may be selected according to actual needs to achieve the purpose of the embodiments of the present application. For example, the functional modules / units in the various embodiments of the present application may be integrated into a processing module, or each module / unit may exist physically separately, or two or more modules / units may be integrated into a single module / unit.

[0204] Those skilled in the art should further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the above description has generally described the composition and steps of each example according to function. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0205] This embodiment further provides an electronic device, comprising: a memory for storing a computer program; and a processor for executing the computer program so that the electronic device performs the above-mentioned palm vein image quality detection method.

[0206] An embodiment of the present application further provides a computer-readable storage medium on which a computer program is stored. When the program is executed by an electronic device, the above-mentioned palm vein image quality detection method is implemented.

[0207] Those skilled in the art will appreciate that all or part of the steps in the method for implementing the above embodiment can be performed by instructing a processor through a program, and the program can be stored in a computer-readable storage medium, which is a non-transitory medium, such as a random access memory, a read-only memory, a flash memory, a hard disk, a solid-state drive, a magnetic tape, a floppy disk, an optical disc, and any combination thereof. The above storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or a data center that includes one or more available media. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a digital video disc (DVD)), or a semiconductor medium (e.g., a solid-state drive (SSD)).

[0208] The descriptions of the processes or structures corresponding to the above figures have different emphases. For parts that are not described in detail in a certain process or structure, please refer to the relevant descriptions of other processes or structures.

[0209] The above embodiments are merely illustrative of the principles and effects of this application and are not intended to limit this application. Anyone skilled in the art may modify or alter the above embodiments without departing from the spirit and scope of this application. Therefore, all equivalent modifications or alterations made by one of ordinary skill in the art without departing from the spirit and technical concepts disclosed in this application shall be covered by the claims of this application.

Claims

1. A palm vein image quality detection method, characterized in that: The palm vein image quality detection method comprises the following steps: Acquire palm vein images; Performing a preliminary image screening on the palm vein image to obtain a palm vein image that is preliminarily determined to be qualified; obtaining a first edge region and a second edge region of preset width from the palm vein image; the first edge region and the second edge region are both located at the edge of the palm vein image and are positioned relative to each other; calculating an average pixel value of the first edge region and the second edge region; if the average pixel value meets a preset pixel condition, preliminarily determining that the palm vein image is qualified; if the average pixel value does not meet the preset pixel condition, preliminarily determining that the palm vein image is unqualified; Performing image quality detection on the palm vein image that is preliminarily determined to be qualified; performing image quality detection on the palm vein image that is preliminarily determined to be qualified includes: Performing image brightness evaluation on the palm vein image that is preliminarily determined to be qualified, and obtaining a first image whose image brightness satisfies a preset brightness condition; Performing palm edge evaluation on the first image to obtain a second image in which the palm edge satisfies a preset edge condition; performing image noise evaluation on the second image to obtain a third image whose image noise satisfies a preset noise condition; Performing hand posture estimation on the third image to obtain a fourth image in which the hand posture satisfies a preset hand posture condition; A palm area evaluation is performed on the fourth image to obtain a fifth image whose palm area meets a preset palm area condition; the fifth image is the palm vein image that is finally determined to be qualified.

2. The palm vein image quality detection method according to claim 1, characterized in that: The performing of preliminary image screening on the palm vein image comprises the following steps: Acquiring a central area of ​​the palm vein image; Calculating the average grayscale value and the maximum grayscale value of the central area; If the average grayscale value and the maximum grayscale value meet the preset grayscale condition, the palm vein image is preliminarily determined to be qualified; if the average grayscale value and the maximum grayscale value do not meet the preset grayscale condition, the palm vein image is preliminarily determined to be unqualified.

3. The palm vein image quality detection method according to claim 1, characterized in that: When the image quality detection includes human hand posture estimation, performing human hand posture estimation on the palm vein image that is preliminarily determined to be qualified comprises the following steps: determining whether a hand exists based on the palm vein image that is preliminarily determined to be qualified; When the hand exists, multiple hand joints are estimated; The palm posture is judged based on multiple hand joints to determine whether the hand posture meets the preset hand posture conditions; the palm posture judgment includes at least any one of the following: judging whether the palm is complete, judging whether the palm is the left palm or the right palm, and judging whether the palm is fully open; wherein, the judgment of whether the palm is fully open includes: judging whether the curvature of each finger meets the preset curvature condition, and / or calculating the angle between two adjacent fingers to determine whether the angle meets the preset angle condition.

4. The palm vein image quality detection method according to claim 3, characterized in that: When the image quality detection includes palm area assessment, performing palm area assessment on the palm vein image that is preliminarily determined to be qualified includes the following steps: Calculating the radius of the palm area according to the plurality of hand joint points; the palm area being the largest inscribed circle of the palm; It is determined whether the radius meets a preset radius condition, so as to determine whether the palm area meets a preset palm area condition.

5. The palm vein image quality detection method according to claim 1, characterized in that: When the image quality detection includes palm edge assessment, performing palm edge assessment on the palm vein image that is preliminarily determined to be qualified comprises the following steps: Acquire a palm edge detection image based on the palm vein image preliminarily determined to be qualified; It is determined whether the palm edge meets a preset edge condition based on the palm edge detection image and the palm vein image that is initially determined to be qualified.

6. A palm vein image quality detection system, characterized in that: The palm vein image quality detection system includes: An image acquisition module, used for acquiring palm vein images; an image preliminary screening module, configured to perform preliminary screening on the palm vein image to obtain a palm vein image that is preliminarily determined to be qualified; obtain a first edge region and a second edge region of preset width from the palm vein image; the first edge region and the second edge region are both located at the edge of the palm vein image and are positioned relative to each other; calculate an average pixel value of the first edge region and the second edge region; if the average pixel value meets a preset pixel condition, preliminarily determine that the palm vein image is qualified; if the average pixel value does not meet the preset pixel condition, preliminarily determine that the palm vein image is unqualified; An image detection module is used to perform image quality detection on the palm vein image that is preliminarily determined to be qualified; the image quality detection on the palm vein image that is preliminarily determined to be qualified includes: Performing image brightness evaluation on the palm vein image that is preliminarily determined to be qualified, and obtaining a first image whose image brightness satisfies a preset brightness condition; Performing palm edge evaluation on the first image to obtain a second image in which the palm edge satisfies a preset edge condition; performing image noise evaluation on the second image to obtain a third image whose image noise satisfies a preset noise condition; Performing hand posture estimation on the third image to obtain a fourth image in which the hand posture satisfies a preset hand posture condition; A palm area evaluation is performed on the fourth image to obtain a fifth image whose palm area meets a preset palm area condition; the fifth image is the palm vein image that is finally determined to be qualified.

7. An electronic device, characterized in that: The electronic device comprises: a memory for storing a computer program; A processor, wherein the processor is configured to execute the computer program so as to enable the electronic device to perform the palm vein image quality detection method according to any one of claims 1 to 5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by an electronic device, the palm vein image quality detection method according to any one of claims 1 to 5 is implemented.

Images