Palmprint collection method, device, equipment and storage medium

Abstract

The present application discloses a palm print collection method, device, equipment and storage medium, and relates to the field of palm print payment. The device includes crisscross conveyor belts and a support frame; a palm print collection component is fixedly installed at the center of the transverse conveyor belt; the belt body of the transverse conveyor belt is sleeved in a first sliding groove, and the first sliding groove is fixedly installed on the longitudinal conveyor belt; the belt body of the longitudinal conveyor belt is sleeved in a second sliding groove, and the second sliding groove is fixedly installed on the support frame; the transverse conveyor belt and the longitudinal conveyor belt are respectively controlled by a first stepper motor and a second stepper motor to achieve transverse and longitudinal sliding. The device can identify the collected palm shape and determine the palm orientation and palm deflection angle according to the palm integrity and key points. In the case of left-right deviation of the palm, the aperture can be moved to the position corresponding to the palm by moving the transverse conveyor belt of the device left-right. In the case of up-down deviation of the palm, the aperture is moved to the position facing the palm by moving the longitudinal conveyor belt of the device up-down.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of palmprint payment, and in particular to a palmprint collection method, device, equipment and storage medium. Background Art

[0002] Fingerprints and palm prints are limited by the development of camera technology. They usually use contact capacitive sensors to scan images and then process them digitally. Non-contact image scanning often loses a lot of details, especially on palm prints. Low-resolution cameras cannot capture clearer palm print images, while high-resolution cameras are too expensive.

[0003] In the related art, the method for the camera to collect palm prints is that the user moves the distance and angle between the user and the camera. When the palm is moved to a suitable angle with the camera, a clear image will be collected. The reason for this application scenario is that most of the current palm print collection cameras collect in a fixed direction and will not move in a direction to actively find a suitable angle. When the palm of the hand is tilted in a certain direction and is not facing the camera, the palm print image will be collected unsatisfactorily. Although this kind of palm deviation can be corrected by an algorithm, this correction also has the risk of deformity, so it is possible to consider solving the problem of palm deviation from the device. Summary of the invention

[0004] The embodiments of the present application provide a palmprint collection method, device, equipment and storage medium to solve the problems of image distortion, deformity and offset when collecting images for palmprint payment.

[0005] On the one hand, the present application provides a palm print collection device, the device comprising a longitudinal conveyor belt, a transverse conveyor belt and a support frame; a palm print collection component is fixedly installed at the center of the transverse conveyor belt; the belt body of the transverse conveyor belt is sleeved in a first sliding groove, and the first sliding groove is fixedly installed on the longitudinal conveyor belt; the belt body of the longitudinal conveyor belt is sleeved in a second sliding groove, and the second sliding groove is fixedly installed on the support frame;

[0006] The transverse conveyor belt is controlled to slide transversely by a first stepper motor, and the longitudinal conveyor belt is controlled to slide longitudinally by a second stepper motor;

[0007] When the palm print image collected by the palm print collection component is offset, the longitudinal conveyor belt and / or the transverse conveyor belt are respectively controlled to move in corresponding directions to adjust the shooting position.

[0008] Specifically, the longitudinal conveyor belt and the transverse conveyor belt are arc-shaped structures respectively, and are respectively inserted into the arc-shaped first sliding groove and the second sliding groove.

[0009] Specifically, the legs of the first sliding slot are fixedly mounted on the longitudinal conveyor belt, an inner cavity is provided between the slot and the legs, and a first gear drivingly connected to the first stepper motor is installed in the inner cavity;

[0010] The bottom of the transverse conveyor belt is provided with a thread, and the first gear is meshed with the thread of the transverse conveyor belt to drive the transverse conveyor belt to move transversely.

[0011] Specifically, the legs of the second sliding slot are fixedly mounted on the support frame, an inner cavity is provided between the slot and the legs, and a second gear drivingly connected to the second stepping motor is installed in the inner cavity;

[0012] A thread is provided at the bottom of the longitudinal conveyor belt, and the second gear is meshed with the thread of the longitudinal conveyor belt to drive the longitudinal conveyor belt to move longitudinally.

[0013] Specifically, the palm print collection component includes a cylindrical shell, a prism is installed at the axis position of the bottom of the shell, and a camera is installed on the side wall of the shell. Light is incident along the axis of the shell, and is reflected to the camera after passing through the prism to collect the palm print image.

[0014] An aperture assembly is fixedly arranged on the outside of the prism, and an aperture gear capable of adjusting the aperture size is sleeved on the outside of the aperture assembly; a convex lens is arranged above the aperture assembly, and the convex lens is embedded at the entrance of the aperture assembly and the housing;

[0015] When the palm print image collected by the camera does not meet the recognition requirements, the aperture gear is controlled to rotate forward or reverse according to the size of the palm print image to adjust the aperture focal length and depth of field.

[0016] On the other hand, the present application provides a palmprint collection method, which is used in a palmprint collection device, and the method includes:

[0017] Collect and extract palm print images according to the image acquisition instruction, identify the pixel information of the palm print images, and determine the number of palm key points and the pixel coordinate data of the palm key points; the palm key points include finger tip end points, finger joint end points and wrist joint end points;

[0018] Calculate the completeness of the palm image and the angle orientation of the palm center according to the number of the palm key points and the pixel coordinate values;

[0019] When the palm print image collected by the camera does not meet the recognition requirements, the adjustment parameters of each gear in the palm print collection device are calculated according to the palm angle offset position, and the aperture collection position is adjusted to obtain the palm print image of the target angle.

[0020] Specifically, the calculating the completeness of the palm image and the palm center angle orientation according to the number of the palm key points and the pixel coordinate values ​​includes:

[0021] Define the key point A (Xa, Ya) at the base of the little finger, the key point B (Xb, Yb) at the tip of the thumb, the key point C (Xc, Yc) at the tip of the index finger, the key point D (Xd, Yd) at the tip of the middle finger, the key point E (Xe, Ye) at the tip of the ring finger, and the key point F (Xf, Yf) at the end of the middle finger joint;

[0022] After extracting the palmprint image, identifying all key points and corresponding coordinate information according to pixel information;

[0023] When the palm image contains six key points, indicating that the palm image is complete, the coordinates of the intersection point Qe formed by the line connecting key points A and D and the line connecting key points B and E, and the coordinates of the intersection point Qc formed by the line connecting key points A and C and the line connecting key points B and D are calculated respectively;

[0024] Determine the palm orientation and palm deflection angle according to the coordinates of the intersection point Qe and the coordinates of the intersection point Qc;

[0025] When the palm orientation and palm deflection angle do not meet the recognition requirements, the horizontal axis offset parameter and / or the vertical axis offset parameter are determined according to the palm orientation and the palm deflection angle, and the posture adjustment information is generated according to the offset parameters, and the stepper motor is controlled to adjust the horizontal conveyor belt and / or the vertical conveyor belt and then re-capture the image.

[0026] Specifically, the horizontal axis offset parameter is determined according to the palm orientation and the palm deflection angle, including:

[0027] When it is identified that the coordinate of the intersection point Qc falls within the pixel area of ​​the index finger, and the coordinate of the intersection point Qe falls within the pixel area of ​​the middle finger, it is determined that the palm is deviated to the left relative to the aperture; the transverse conveyor belt is controlled to drive the palm print collection component to move left, and then the camera angle is adjusted to the right;

[0028] When it is identified that the coordinate of the intersection point Qc falls within the pixel area of ​​the ring finger, and the coordinate of the intersection point Qe falls within the pixel area of ​​the middle finger, it is determined that the palm is right-biased relative to the aperture; the transverse conveyor belt is controlled to drive the palm print collection component to move right, and then the camera angle is adjusted to the left;

[0029] The longitudinal axis offset parameters are determined according to the palm orientation and palm deflection angle, including:

[0030] Set the coordinates of the center point of the line connecting key points A and B to Qz, and the coordinates are expressed as Qz((Xa+Xb) / 2,(Ya+Yb) / 2); set the first distance Sdf between key points D and F, expressed as follows:

[0031]

[0032] Set the second distance Sqf between the key points Qz and F, expressed as follows:

[0033]

[0034] The ratio P of the palm length to the middle finger length is determined according to the ratio of the first distance to the second distance, which is expressed as follows:

[0035] P=Sqf / Sdf

[0036] When P>1.2, it indicates that the length of the palm is greater than the length of the middle finger, and the palm is above the aperture; the longitudinal conveyor belt is controlled to drive the palm print collection component to move upward, and then the camera angle is controlled to adjust downward;

[0037] When P<1, it indicates that the length of the palm is less than that of the middle finger, and the palm is lower than the aperture; the longitudinal conveyor belt is controlled to drive the palm print collection component downward, and then the camera angle is controlled to adjust upward.

[0038] On the other hand, the present application provides a computer device, comprising a processor and a memory, wherein the memory stores at least one instruction, at least one program, a code set or an instruction set, and the at least one instruction, the at least one program, the code set or the instruction set is loaded and executed by the processor to implement the palmprint collection method described in any of the above aspects.

[0039] On the other hand, the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores at least one instruction, at least one program, a code set or an instruction set, and the at least one instruction, the at least one program, the code set or the instruction set is loaded and executed by a processor to implement the palmprint collection method described in any of the above aspects.

[0040] The beneficial effects brought about by the technical solution provided by the embodiment of the present application include at least: the algorithm and device of the present application can first identify the palm shape in the collected picture, and determine the palm orientation and palm deviation angle according to the palm integrity and key point identification. In the case of the palm being offset to the left or right, the aperture can be moved to the position corresponding to the palm by moving the horizontal conveyor belt of the device to the left or right to achieve angle correction. In the case of the palm being offset up or down, the aperture can be moved to the position facing the palm by moving the longitudinal conveyor belt of the device up or down. Because of the design of the arc-shaped conveyor belt, posture correction can be achieved in a large range, ensuring that a complete and angle-accurate palm print image is captured.

[0041] Especially for the scene where the palm deviates seriously and the key points are not fully recognized, the stepper motor in the palm print acquisition component can be used alone to control the focus. When the palm size is judged to exceed the image size, it means that the palm is too close to the camera. At this time, the camera aperture is controlled to be reduced, and the depth of field becomes farther, and the palm size will match the image size. When the palm size is smaller than the image size, the camera aperture is increased according to demand, and the depth of field becomes closer, until the palm size matches the image size. This solution can adaptively adjust and collect palm images by controlling the focus to the greatest extent, thereby improving the recognition success rate. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 is a front view of a palmprint collection device provided in an embodiment of the present application;

[0043] Figure 2 is a cross-sectional view of the side of the palmprint collection device;

[0044] Figure 3 A partial enlarged view of the palm print collection component is shown;

[0045] Figure 4 What is shown is a schematic diagram of the palm print collection component moving with respect to the palm on the horizontal and vertical axes;

[0046] Figure 5 is a flow chart of a palmprint collection method provided in an embodiment of the present application;

[0047] Figure 6 A schematic diagram of selecting key points of the palm is shown.

[0048] Figure 7 Schematic diagrams showing scenes with the palm tilted to the left and the palm turned up;

[0049] Figure 8 An algorithm flow chart of a palmprint collection method provided in an embodiment of the present application is shown;

[0050] Fig. 9 Schematic diagrams of different palm placement scenarios are given;

[0051] Fig.10 A flow chart of posture adjustment strategies for different palm placement scenarios is given.

[0052] Figure numerals: 10-longitudinal conveyor belt; 20-transverse conveyor belt; 30-first sliding groove; 40-second sliding groove; 50-first stepper motor; 60-palmprint collection assembly; 61-convex lens; 62-aperture gear; 63-stepper motor; 65-aperture assembly; 66-sensor; 67-prism; 68-camera; 70-support frame. DETAILED DESCRIPTION

[0053] In order to make the objectives, technical solutions and advantages of the present application clearer, the implementation methods of the present application will be further described in detail below with reference to the accompanying drawings.

[0054] The term "multiple" as used herein refers to two or more than two. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone. The character " / " generally indicates that the related objects are in an "or" relationship.

[0055] Figure 1 is a front view of a palmprint collection device provided in an embodiment of the present application, Figure 2 : is a cross-sectional view of the side of the palm print collection device. The device as a whole is a multi-angle posture adjuster erected by a support frame 70, including a longitudinal conveyor belt 10, a transverse conveyor belt 20 and a support frame 70; a palm print collection component 60 is fixedly installed at the center of the transverse conveyor belt 20. The belt body of the transverse conveyor belt 20 is entirely or partially sleeved in the first sliding groove 30. The first sliding groove 30 is fixedly installed on the longitudinal conveyor belt 10. The belt body of the longitudinal conveyor belt 10 is entirely or partially sleeved in the second sliding groove 40, and the second sliding groove 40 is fixedly installed on the support frame 70. The transverse conveyor belt 20 is controlled to slide horizontally by the first stepper motor 50, and the longitudinal conveyor belt 10 is controlled to slide longitudinally by the second stepper motor. The function of the device is to control the longitudinal conveyor belt 10 and / or the transverse conveyor belt 20 to move in the corresponding direction and adjust the shooting position when the palm print image collected by the palm print collection component 60 is offset.

[0056] In the embodiment of the present application, the longitudinal conveyor belt 10 and the transverse conveyor belt 20 are respectively designed as arc structures, and are respectively inserted into the arc-shaped first sliding groove 30 and the second sliding groove 40. The overall transverse and longitudinal belt bodies are perpendicular to each other, and the two arc structures are specifically set according to actual conditions. The two arcs are generally kept the same, and the curved surface formed by the arc-shaped curved structure is similar to a "small sun heater".

[0057] See also Figure 3 The enlarged partial view of the palm print collection assembly is shown, the legs of the first sliding slot 30 are fixedly mounted on the longitudinal conveyor belt 10, an inner cavity is provided between the slot and the legs, and a first gear 69 is installed in the inner cavity which is transmission-connected to the first stepper motor 50. The bottom of the transverse conveyor belt 20 is provided with a thread, and the first gear 69 meshes with the thread of the transverse conveyor belt 20 to drive the transverse conveyor belt 20 to move transversely.

[0058] The legs of the second sliding slot 40 are fixedly mounted on the support frame 70, and an inner cavity is provided between the slot and the legs, in which a second gear 41 connected to the second stepping motor is installed. Correspondingly, a thread is provided at the bottom of the longitudinal conveyor belt 10, and the second gear 41 meshes with the thread of the longitudinal conveyor belt 10 to drive the longitudinal conveyor belt 10 to move longitudinally.

[0059] In some embodiments, because the palm print collection component 60 moves on an arc surface, it is necessary to have a highly controllable focusing capability when the shooting angle and distance change. Therefore, the palm print collection component 60 provided in the present application is also designed specifically. The palm print collection component 60 is a cylindrical device shell as a whole, and a prism 67 is installed at the axis position of the bottom of the shell. A camera 68 for collecting palm print images is installed on the inner side wall of the shell. When shooting, the palm print light is incident inward along the axis of the shell, and is reflected to the camera 68 after passing through the prism 67 to capture the palm print image.

[0060] An aperture assembly 65 is fixedly arranged above the prism 67, and an aperture gear 62 with adjustable aperture size is sleeved on the outside of the aperture assembly 65, that is, the aperture parameters can be changed by rotating the aperture gear 62. A convex lens 61 is arranged above the aperture assembly 65, and the convex lens 61 is embedded at the entrance of the aperture assembly 65 and the housing.

[0061] A sensor 66 is installed on the inner wall of the housing where the aperture gear 62 is located to collect the aperture tooth pattern in real time. A stepper motor 63 is installed on the inner side of the bottom of the housing, and a gear is installed on the output shaft of the stepper motor 63, which meshes with the aperture gear 62 on the outer edge of the aperture assembly 65. When the palm print image collected by the camera 68 is incomplete, the aperture gear 62 can be controlled to rotate forward or reverse through the motorized assembly to adjust the aperture focal length and depth of field. For example, the aperture side rotates one circle, and the aperture is adjusted from maximum to minimum.

[0062] In a feasible implementation, the focal length of the convex lens 61 is set to 40mm, and the material is quartz glass. It is divided into two symmetrical convex lenses 61, the upper half of the convex lens 61 is set outward and installed on the external housing, and the lower half is set inward and set at the aperture. There is a gap between the planes of the two convex lenses 61, and the gap size is 1.8mm. The aperture thickness varies with the adjusted overlap degree, with a minimum of 1.3mm and a maximum of 1.5mm.

[0063] In the present application, the cross section of the prism 67 is a right triangle, with two right angles of 45 degrees. The upper right angle is the light entrance, receiving the external light from the convex lens 61. The hypotenuse of the prism 67 is the light reflection point, and the hypotenuse has a reflective coating, and also absorbs floodlight, so as to enhance the image contrast. The incident light is reflected by the oblique surface and horizontally transmitted to the camera 68, which is the CMOS image sensor 66. In a feasible embodiment, the focal length of the CMOS image sensor 66 is 40mm, the pixel is 200W, and the imaging size is 640*480.

[0064] In the embodiment of the present application, the lengths of the transverse and longitudinal transmission belts are limited, such as Figure 4 The maximum angle of the palm print collection component on the horizontal and vertical axes for palm movement is 76°. This angle is measured based on the usage scenario. Generally, there will not be an extreme scenario where the palm is flat and only the side of the palm is facing the camera. Affected by the palm bone structure, in most cases, the angle of deflection to the camera is between 0° and 34° when the user's palm is facing down, and between 0° and 22° when the palm is facing up. When the palm is facing left or right, the angle of deflection to the camera will be between 0° and 30°. Therefore, the transmission belt in the horizontal and vertical axis direction can keep the camera moving between -38° and positive 38° to meet most application scenarios.

[0065] Figure 5 : is a flow chart of a palmprint collection method provided in an embodiment of the present application, which is applied to the above-mentioned palmprint collection device. The method comprises the following steps:

[0066] S1, collecting and extracting palmprint images according to the image acquisition instruction, identifying the pixel information of the palmprint images, and determining the number of palm key points and the pixel coordinate data of the palm key points;

[0067] Based on the above device, the PC sends the image acquisition command to the image acquisition device through the USB interface. After receiving the image acquisition command, the main control chip on the device turns on the CMOS image sensor and LED lamp beads to start collecting palm prints. At this time, the palm should be placed in front of or above the convex lens to capture a frame of the captured picture.

[0068] Then, the histogram of this frame of image is calculated through the image processing algorithm to obtain the palm threshold. The image is binarized according to the obtained threshold to distinguish the palm from the background and extract the palmprint image.

[0069] When the palmprint image contains the image content of the palm, it can be identified through pixel analysis.

[0070] In the embodiment of the present application, the finger endpoints, knuckle endpoints and wrist endpoints can be set as key points, and the position coordinates of these key points can be determined for subsequent position judgment.

[0071] S2, calculating the completeness of the palm image and the angle orientation of the palm center according to the number of palm key points and pixel coordinate values;

[0072] The palm integrity is determined by the distance between the palm and the convex lens. When the palm distance is appropriate, the complete palm image can be captured. When the palm distance is far, the complete palm image cannot be captured, which will result in the loss of key points. The position of the palm can be determined based on the lost key points. The main consideration for the palm angle orientation is the scene where the shooting screen contains all the key points. That is to say, under the premise of the complete palm image, the angle formed between the palm and the aperture (the light inlet is used as the recognition center point) is determined to determine the palm angle orientation. Only under the premise of a reasonable angle orientation can the palm print image be clear and accurately recognized.

[0073] S3. When the palm print image collected by the camera does not meet the recognition requirements, the adjustment parameters of each gear in the palm print collection device are calculated according to the palm angle offset position, and the aperture collection position is adjusted to obtain the palm print image of the target angle.

[0074] For incomplete palm print images, appropriate adjustment parameters can be calculated based on the palm position. At the same time, for complete palm print parameters, if the image clarity is insufficient due to distance, the adjustment parameters for the aperture gear are also calculated based on the palm position. For example, when the palm print is slightly deviated to the left, the aperture is directly controlled to reacquire the image and include the palm; and when the palm print is seriously deviated to the left, the voice broadcast reminds to move to the right. When the palm image is small due to the distance of the palm, the aperture is reduced to shorten the depth of field and then take the best palm print image.

[0075] When a complete palm print image is obtained but the palm angle is offset, posture adjustment is also required. For example, if the hand is tilted to the left or right, the movement of the horizontal conveyor belt can be controlled, which indirectly drives the overall movement of the palm print collection component and enables the palm to be aligned with the aperture entrance.

[0076] This application focuses on the discussion of angle and posture correction adjustment in the scenario of capturing a complete palm print image.

[0077] In some embodiments, the completeness of the palm image and the palm center angle orientation are calculated according to the number of palm key points and pixel coordinate values, including:

[0078] a. Define the key point A (Xa, Ya) of the base of the little finger, the key point B (Xb, Yb) of the thumb tip, the key point C (Xc, Yc) of the index finger tip, the key point D (Xd, Yd) of the middle finger tip, the key point E (Xe, Ye) of the ring finger tip, and the key point F (Xf, Yf) of the middle finger joint endpoint; Figure 6The left side shows a schematic diagram of the palm selection key points.

[0079] b. After extracting the palmprint image, identify all key points and corresponding coordinate information based on pixel information;

[0080] c. When the palm image contains six key points, indicating that the palm image is complete, the coordinates of the intersection point Qe formed by the line connecting key points A and D and the line connecting key points B and E, and the coordinates of the intersection point Qc formed by the line connecting key points A and C and the line connecting key points B and D are calculated respectively;

[0081] from Figure 6 As can be seen on the right, in an ideal state, the palm is facing the camera. Under normal circumstances, the middle finger of a human finger is approximately on the vertical axis of the palm. That is to say, connecting key points A, B, and D will result in an approximate isosceles triangle, with AB as the base of the triangle and AD and BD as the two isosceles sides of the triangle. The index finger and ring finger are the two fingers closest to the middle finger. We connect key point A and key point C to get line AC, and connect key point B and key point E to get line BE. At this point, you can see that the two lines AC and BD have an intersection, which is in the background area between the index finger and the middle finger that is not the palm, so we call it point Qc. Similarly, we can get that the two lines BE and AD have an intersection, which is in the background area between the ring finger and the middle finger that is not the palm, so we call it point Qe.

[0082] d. Determine the palm orientation and palm deflection angle according to the coordinates of the intersection point Qe and the coordinates of the intersection point Qc;

[0083] e. When the palm orientation and palm deflection angle do not meet the recognition requirements, determine the horizontal axis offset parameter and / or the vertical axis offset parameter according to the palm orientation and the palm deflection angle, generate posture adjustment information according to the offset parameters, control the stepper motor to adjust the horizontal conveyor belt and / or the vertical conveyor belt, and then re-capture the image.

[0084] Figure 7 The diagram shows the scene of the palm tilted to the left and the palm turned up. The left picture shows the scene of the palm tilted to the left. It can be seen from the figure that the Qc point leaves the non-palm background area between the index finger and the middle finger and falls on the index finger of the palm, and the Qe point leaves the non-palm background area between the ring finger and the middle finger and falls on the middle finger. Therefore, we only need to determine the landing point of the two pixel coordinates of Qc and Qe to determine the palm direction and approximate deflection angle.

[0085] To find the formula for the coordinates of the two intersection points Qc and Qe, take Qc as an example, the coordinates of key points A (Xa, Ya), key points B (Xb, Yb), and C (Xc, Yc) are known. Therefore, the equation of the line AC can be obtained based on the two-point formula:

[0086] (Yb-Ya)*x-(Xb-Xa)*y+(Xb*Ya–Yb*Xa)=0

[0087] Similarly, the equation of line BD is:

[0088] (Yd-Yb)*x-(Xd-Xb)*y+(Xd*Yb–Yd*Xb)=0

[0089] So at this time we can form a set of two-variable linear equations to obtain the coordinates of the intersection point

[0090]

[0091] It should be noted that in most cases, the x and y obtained at this time are not integers, but the pixel coordinates are integers, so in the subsequent steps, x and y can be rounded.

[0092] In step d, different judgment and correction methods are used for the offsets in different directions and angles. Specifically, the horizontal axis offset parameters are determined according to the palm orientation and the palm deflection angle, including:

[0093] 1. When it is identified that the coordinate of the intersection point Qc falls within the pixel area of ​​the index finger, and the coordinate of the intersection point Qe falls within the pixel area of ​​the middle finger, it is determined that the palm is deviated to the left relative to the aperture;

[0094] In this case, the adjustment parameters can be determined based on the previously set relationship between the first stepper motor and the gear. For example, the first stepper motor is controlled to rotate forward 0.25 times, and the transverse conveyor belt drives the palm print collection component to move to the left, and then the camera is adjusted to the right, and then the picture is collected again.

[0095] 2. When it is identified that the coordinate of the intersection point Qc falls within the pixel area of ​​the ring finger, and the coordinate of the intersection point Qe falls within the pixel area of ​​the middle finger, it is determined that the palm is right-biased relative to the aperture;

[0096] In this case, the first stepper motor can be controlled to reverse 0.25 turns, the transverse conveyor belt drives the palm print collection component to move to the right, and then the camera angle is adjusted to the left, and then the picture is collected again.

[0097] For scenarios where the pitch angle adjustment is required, the longitudinal axis offset parameters are determined according to the palm orientation and palm deflection angle, including:

[0098] First of all, it is necessary to clarify a premise of this application that the ratio between the length of a person's palm and the length of the middle finger is approximately between 1:1 and 1.2:1. Figure 2 The right side shows the palm facing upward relative to the camera. The middle finger is obviously shorter than the palm, and the ratio does not meet the range of 1:1 to 1.2:1. Similarly, when the palm faces downward relative to the camera, the middle finger is obviously longer than the palm. In this way, the step becomes to find the ratio of the palm to the middle finger.

[0099] 1. Set the center point coordinates of the line connecting key points A and B as Qz, and the coordinate representation is Qz((Xa + Xb) / 2, (Ya + Yb) / 2); set the first distance Sdf between key points D and F, which is expressed as follows:

[0100]

[0101] Set the second distance Sqf between key points Qz and F, which is expressed as follows:

[0102]

[0103] 2. Determine the ratio P of the palm length to the middle finger length according to the ratio of the first distance to the second distance, which is expressed as follows:

[0104] P = Sqf / Sdf

[0105] 3. When P > 1.2, it indicates that the palm length is greater than the middle finger length, and the palm center is relatively above the aperture; control the longitudinal conveyor belt to drive the palmprint acquisition component to move up, and then control the camera angle to adjust downward;

[0106] 4. When P < 1, it indicates that the palm length is less than the middle finger length, and the palm center is relatively below the aperture; control the longitudinal conveyor belt to drive the palmprint acquisition component to move down, and then control the camera angle to adjust upward.

[0107] If the measurement result satisfies 1 < P < 1.2, it means that the angle of this image acquisition and detection is appropriate. It should be noted that there are also special groups of people whose ratio between the palm length and the middle finger length is not between 1:1 and 1.2:1. However, the ratio of this group of people is actually still close to this ratio. At this time, 1 < P < 1.2 is still used as the standard camera angle judgment value. In this case, the deviation angle of the captured image is very slight and does not affect the image quality.

[0108] Figure 8 The algorithm flow chart of the palmprint acquisition method provided based on the embodiment of the present application is shown. The image acquisition module is the acquisition and detection process of the aperture component and the camera module. After obtaining a frame of image, the palm shape is determined through the algorithm module of the main control chip, and then the key point information is extracted. After that, according to the positions of the Qc and Qe coordinate points, it is determined whether it is necessary to enable the first stepping motor (stepping motor A) for horizontal attitude adjustment, and whether it is necessary to enable the second stepping motor (stepping motor B) for vertical attitude adjustment.

[0109] In addition, the present application also describes the premise of detecting the number of key points in step S2, the purpose of which is to ensure that the integrity of the palm meets the requirements, especially in the scene where the palm is too close to capture the whole picture. In this case, it is necessary to use the palm print collection component itself to adjust the focus to complete the panoramic shooting, and continue to measure the palm posture angle while ensuring that the full picture of the palm is captured.

[0110] In some implementations, calculating the integrity of the palm image according to the number of palm key points and pixel coordinate values ​​may include the following steps:

[0111] 1. Define the key point A (Xa, Ya) at the base of the thumb, the key point B (Xb, Yb) at the tip of the thumb, the key point C (Xc, Yc) at the tip of the index finger, the key point D (Xd, Yd) at the tip of the middle finger, the key point E (Xe, Ye) at the tip of the ring finger, and the key point F (Xf, Yf) at the tip of the little finger;

[0112] 2. After extracting the palmprint image, identify all key points and corresponding coordinate information based on pixel information;

[0113] 3. When the palm image contains six key points, it indicates that the palm image is complete, and the distance value S between endpoint A and endpoint D is calculated. AD ; means as follows:

[0114]

[0115] 4. When S AD When the value is within the first preset range, it indicates that the size of the palm is close to the size of the picture, and the picture is determined to be taken;

[0116] 5. When S AD When the value is within the first preset range but outside the second preset range, it indicates that the palm image is too small. AD The value determines the adjustment parameter and controls the stepper motor to rotate the aperture gear forward to shorten the depth of field.

[0117] 6. When there are no key points in the palm image, the offset parameters are determined according to the finger positions corresponding to the key points, and the posture adjustment information is generated according to the offset parameters. The stepper motor is controlled to correct the deviation and then the image is re-taken for recognition.

[0118] Specifically, for situations where key points are missing, the focus is on how to locate the palm position and how to make a judgment. Under normal logic, when all six key points are not detected, it indicates that the palm is close enough to exceed the set threshold and the size of the palm exceeds the image size; in this case, the palm can be prompted to retreat. Of course, this application can generate posture adjustment information such as reducing the aperture and extending the depth of field based on the number of missing key points. Of course, for those that exceed the adjustment range, the palm is prompted to retreat. When the detection result lacks key point F or key point B, the palm is instructed to shift to both sides; the user can be reminded to move to the corresponding side, and this application generates posture adjustment information such as reducing the aperture and extending the depth of field based on the number of missing key points. Similarly, when the detection result lacks key point A, or lacks key points A and B, or only contains key points A and B, it indicates that the palm is severely shifted up and down; this situation exceeds the aperture posture adjustment range, and the user can be reminded of the corresponding adjustment gesture.

[0119] In order to facilitate the description of the scene, this application further optimizes the algorithm and makes corresponding posture adjustment strategies according to different key points. Fig. 9 Schematic diagrams of different palm placement scenarios are given.

[0120] Key points A, B, C, D, E, and F have the following scenarios:

[0121] Scenario 1: The six key points are not detected. This means the size of the captured palm exceeds the image size of 640*480;

[0122] Scenario 2: The result does not detect key point F or the algorithm does not detect key point B. This means that the palm is not in the center of the image, and the palm is slightly tilted to the left or right;

[0123] Scenario 3: The result does not detect key points F and E, or the algorithm does not detect key points B and C. This means that the palm is not in the center of the image, and the palm is moderately tilted to the left or right;

[0124] Scenario 4: The result does not detect key points A, B, and C, or the algorithm does not detect key points D, E, and F. This means that the palm is not in the center of the image, and the palm is severely tilted to the left or right;

[0125] Scene 5: Only key points A and B are detected. This means that the palm is not in the center of the image, but slightly above it.

[0126] Scene 6: Keypoint A is not detected. This means the palm is not in the center of the image, but slightly below.

[0127] Scenario 7: Key points A and B are not detected, or only key point A is detected. This means that the palm is not in the center of the image, but is seriously tilted downward or upward.

[0128] Scenario Eight: All key points have been detected.

[0129] For the above eight scenarios, Fig.10 pose adjustment strategy flowcharts for different palm placement scenarios are given.

[0130] For Scenarios One, Two, Three, and Six, which are classified as one type where a certain part of the palm exceeds the picture size. After the algorithm in the main control chip determines that it belongs to these scenarios, the main control chip will first read the scale value in the sensor, then control the stepper motor to reverse half a turn, reduce the aperture, pull the depth of field farther, and then re-capture the image.

[0131] For Scenarios Four and Seven, since the palm posture is severely deviated, re-capturing the image is required. The user can adjust according to the prompt, or information such as payment failure can be reported.

[0132] For Scenario Five, the algorithm processes as follows:

[0133] Since the pixel coordinates of the key point to a point perpendicular to the bottom are (0, Ya), the distance from the key point A to the bottom pixel of the picture is Ya.

[0134] If Ya > 240, it means the palm image is severely biased upwards. The user can be prompted to adjust the palm downwards, and then it directly returns to the step of re-capturing the image.

[0135] If 120 < Ya < 240, it means the palm image is moderately biased upwards. The main control chip will first read the scale value in the sensor, then control the stepper motor to reverse one turn, reduce the aperture, pull the depth of field farther, and then return to the step of re-capturing the image.

[0136] If Ya < 120, it means the palm image is slightly biased upwards. The main control chip will first read the scale value in the sensor, then control the stepper motor to reverse half a turn, reduce the aperture, pull the depth of field farther, and then return to the step of re-capturing the image.

[0137] For Scenario Eight, the algorithm processes as follows:

[0138] Calculate the distance S between key point A and key point D AD ;

[0139] If 420 < S AD < 480, it means the size of the palm image is close to the picture size. The main control chip returns the current frame image to the host computer or PC via USB and prompts that the image capture is successful.

[0140] If 360 < S AD < 420, it means the palm image is slightly smaller. The main control chip will first read the scale value in the sensor, then control the stepper motor to rotate forward half a turn, increase the aperture, pull the depth of field closer, and then return to the step of re-capturing the image.

[0141] If 240 AD <360, it means that the palm image is too small. The main control chip will first read the scale value in the sensor, and then control the stepper motor to rotate forward one circle, increase the aperture, shorten the depth of field, and then return to the step of retaking the image.

[0142] If 0 AD <240, indicating that the palm image is too small. The main control chip will first read the scale value in the sensor, and then control the stepper motor to rotate forward twice, increase the aperture, shorten the depth of field, and then return to the step of retaking the image.

[0143] In particular, after each algorithm determines the scene, if the aperture needs to be adjusted, the main control chip will read the current scale through the sensor. If the current scale is at the forward rotation limit, and the algorithm requires forward rotation, it will directly return to the image acquisition failure. If the current scale is at the reverse rotation limit, and the algorithm requires reverse rotation, it will directly return to the image acquisition failure.

[0144] In summary, the algorithm and device of the present application can first identify the palm shape in the collected picture, and determine the palm orientation and palm deviation angle according to the palm integrity and key point identification. In the case of the palm shifting left and right, the aperture can be moved to the position corresponding to the palm by moving the horizontal conveyor belt of the device left and right to achieve angle correction. In the case of the palm shifting up and down, the aperture is moved to the position directly opposite the palm by moving the longitudinal conveyor belt of the device up and down. Because of the design of the arc-shaped conveyor belt, posture correction can be achieved in a large range, ensuring that a complete and angle-accurate palm print image is captured.

[0145] Especially for the scene where the palm deviates seriously and the key points are not fully recognized, the stepper motor in the palm print acquisition component can be used alone to control the focus. When the palm size is judged to exceed the image size, it means that the palm is too close to the camera. At this time, the camera aperture is controlled to be reduced, and the depth of field becomes farther, and the palm size will match the image size. When the palm size is smaller than the image size, the camera aperture is increased according to demand, and the depth of field becomes closer, until the palm size matches the image size. This solution can adaptively adjust and collect palm images by controlling the focus to the greatest extent, thereby improving the recognition success rate.

[0146] The palmprint collection device provided in the embodiment of the present application can be applied to the palmprint collection method provided in the above embodiment. For relevant details, refer to the above method embodiment. The implementation principle and technical effect are similar and will not be repeated here.

[0147] ​​It should be noted that the palmprint collection device provided in the embodiment of the present application is only illustrated by the division of the above-mentioned functional modules / functional units. In actual applications, the above-mentioned functions can be assigned to different functional modules / functional units as needed, that is, the internal structure of the palmprint collection device is divided into different functional modules / functional units to complete all or part of the functions described above. In addition, the implementation method of the palmprint collection method provided in the above method embodiment and the implementation method of the palmprint collection device provided in this embodiment belong to the same concept. The specific implementation process of the palmprint collection device provided in this embodiment is detailed in the above method embodiment, which will not be repeated here.

[0148] The embodiment of the present application also discloses a computer-readable storage medium. Specifically, the computer-readable storage medium is used to store a computer program, and when the computer program is executed by the processor, the method in the above-mentioned method implementation is implemented. Those skilled in the art can understand that the implementation of all or part of the process in the above-mentioned implementation method of the present application can be completed by instructing the relevant hardware through a computer program, and the program can be stored in a computer-readable storage medium, and when the program is executed, it can include the process of the implementation of the above-mentioned methods. Among them, the storage medium can be a disk, an optical disk, a read-only memory (ROM), a random access memory (RAM), a flash memory (Flash Memory), a hard disk (Hard Disk Drive, abbreviated as: HDD) or a solid-state drive (SSD), etc.; the storage medium can also include a combination of the above-mentioned types of memory.

[0149] This specific embodiment is merely an explanation of the present invention and is not a limitation of the present invention. After reading this specification, those skilled in the art may make non-creative modifications to the present embodiment as needed. However, as long as they are within the scope of the claims of the present invention, they are protected by the patent law.

Claims

1. A palmprint collection device, characterized in that: The device comprises a longitudinal conveyor belt (10), a transverse conveyor belt (20) and a support frame (70); a palm print collection assembly (60) is fixedly installed at the center of the transverse conveyor belt (20); the belt body of the transverse conveyor belt (20) is sleeved in a first sliding groove (30), and the first sliding groove (30) is fixedly installed on the longitudinal conveyor belt (10); the belt body of the longitudinal conveyor belt (10) is sleeved in a second sliding groove (40), and the second sliding groove (40) is fixedly installed on the support frame (70); The transverse conveyor belt (20) is controlled to slide transversely by a first stepper motor (50), and the longitudinal conveyor belt (10) is controlled to slide longitudinally by a second stepper motor; When the palm print image collected by the palm print collection component (60) is offset, the longitudinal conveyor belt (10) and / or the transverse conveyor belt (20) are respectively controlled to move in corresponding directions to adjust the shooting position.

2. The palmprint collection device according to claim 1, characterized in that: The longitudinal conveyor belt (10) and the transverse conveyor belt (20) are respectively arc-shaped structures and are respectively inserted into the arc-shaped first sliding groove (30) and the second sliding groove (40).

3. The palmprint collection device according to claim 2, characterized in that: The legs of the first sliding slot (30) are fixedly mounted on the longitudinal conveyor belt (10), an inner cavity is provided between the slot and the legs, and a first gear (69) drivingly connected to the first stepper motor (50) is installed in the inner cavity; The bottom of the transverse conveyor belt (20) is provided with a thread, and the first gear (69) is meshed with the thread of the transverse conveyor belt (20) to drive the transverse conveyor belt (20) to move laterally.

4. The palmprint collection device according to claim 2, characterized in that: The legs of the second sliding slot (40) are fixedly mounted on a support frame (70), an inner cavity is provided between the slot and the legs, and a second gear (41) drivingly connected to a second stepping motor is installed in the inner cavity; The bottom of the longitudinal conveyor belt (10) is provided with a thread, and the second gear (41) is meshed with the thread of the longitudinal conveyor belt (10) to drive the longitudinal conveyor belt (10) to move longitudinally.

5. The palmprint collection device according to any one of claims 1 to 4, characterized in that: The palm print collection component (60) comprises a cylindrical shell, a prism (67) is installed at the axis position of the bottom of the shell, and a camera (68) is installed on the side wall of the shell. Light is incident along the axis of the shell, passes through the prism (67), and then is reflected to the camera (68) to collect a palm print image. An aperture assembly (65) is fixedly arranged on the outside of the prism (67), and an aperture gear (62) capable of adjusting the aperture size is sleeved on the outside of the aperture assembly (65); a convex lens (61) is arranged above the aperture assembly (65), and the convex lens (61) is embedded at the entrance of the aperture assembly (65) and the housing; When the palm print image collected by the camera (68) does not meet the recognition requirements, the aperture gear (62) is controlled to rotate forward or reverse according to the size of the palm print image to adjust the aperture focal length and depth of field.

6. A palmprint collection method, characterized in that: The method is used for a palmprint collection device, and the method comprises: Collect and extract palm print images according to the image acquisition instruction, identify the pixel information of the palm print images, and determine the number of palm key points and the pixel coordinate data of the palm key points; the palm key points include finger tip end points, finger joint end points and wrist joint end points; Calculate the completeness of the palm image and the angle orientation of the palm center according to the number of the palm key points and the pixel coordinate values; When the palm print image collected by the camera does not meet the recognition requirements, the adjustment parameters of each gear in the palm print collection device are calculated according to the palm angle offset position, and the aperture collection position is adjusted to obtain the palm print image of the target angle.

7. The method according to claim 6, characterized in that The calculating the completeness of the palm image and the palm center angle orientation according to the number of the palm key points and the pixel coordinate values ​​comprises: Define the key point A (Xa, Ya) at the base of the little finger, the key point B (Xb, Yb) at the tip of the thumb, the key point C (Xc, Yc) at the tip of the index finger, the key point D (Xd, Yd) at the tip of the middle finger, the key point E (Xe, Ye) at the tip of the ring finger, and the key point F (Xf, Yf) at the end of the middle finger joint; After extracting the palmprint image, identifying all key points and corresponding coordinate information according to pixel information; When the palm image contains six key points, indicating that the palm image is complete, the coordinates of the intersection point Qe formed by the line connecting key points A and D and the line connecting key points B and E, and the coordinates of the intersection point Qc formed by the line connecting key points A and C and the line connecting key points B and D are calculated respectively; Determine the palm orientation and palm deflection angle according to the coordinates of the intersection point Qe and the coordinates of the intersection point Qc; When the palm orientation and palm deflection angle do not meet the recognition requirements, the horizontal axis offset parameter and / or the vertical axis offset parameter are determined according to the palm orientation and the palm deflection angle, and the posture adjustment information is generated according to the offset parameters, and the stepper motor is controlled to adjust the horizontal conveyor belt and / or the vertical conveyor belt and then re-capture the image.

8. The method according to claim 7, characterized in that The horizontal axis offset parameters are determined according to the palm orientation and palm deflection angle, including: When it is identified that the coordinate of the intersection point Qc falls within the pixel area of ​​the index finger, and the coordinate of the intersection point Qe falls within the pixel area of ​​the middle finger, it is determined that the palm is deviated to the left relative to the aperture; the transverse conveyor belt is controlled to drive the palm print collection component to move left, and then the camera angle is adjusted to the right; When it is identified that the coordinate of the intersection point Qc falls within the pixel area of ​​the ring finger, and the coordinate of the intersection point Qe falls within the pixel area of ​​the middle finger, it is determined that the palm is right-biased relative to the aperture; the transverse conveyor belt is controlled to drive the palm print collection component to move right, and then the camera angle is adjusted to the left; The longitudinal axis offset parameters are determined according to the palm orientation and palm deflection angle, including: Set the coordinates of the center point of the line connecting key points A and B to Qz, and the coordinates are expressed as Qz((Xa+Xb) / 2,(Ya+Yb) / 2); set the first distance Sdf between key points D and F, expressed as follows: Set the second distance Sqf between the key points Qz and F, expressed as follows: The ratio P of the palm length to the middle finger length is determined according to the ratio of the first distance to the second distance, which is expressed as follows: P=Sqf / Sdf When P>1.2, it indicates that the length of the palm is greater than the length of the middle finger, and the palm is above the aperture; the longitudinal conveyor belt is controlled to drive the palm print collection component to move upward, and then the camera angle is controlled to adjust downward; When P<1, it indicates that the length of the palm is less than that of the middle finger, and the palm is lower than the aperture; the longitudinal conveyor belt is controlled to drive the palm print collection component downward, and then the camera angle is controlled to adjust upward.

9. A computer device, characterized in that: The computer device includes a processor and a memory, wherein the memory stores at least one instruction, at least one program, a code set or an instruction set, and the at least one instruction, the at least one program, the code set or the instruction set is loaded and executed by the processor to implement the palmprint collection method according to any one of claims 6 to 8.

10. A computer-readable storage medium, characterized in that: The readable storage medium stores at least one instruction, at least one program, a code set or an instruction set, and the at least one instruction, the at least one program, the code set or the instruction set is loaded and executed by the processor to implement the palmprint collection method according to any one of claims 6 to 8.

Images