CN111563454A - Hand vein identification method and device for double in-vivo verification - Google Patents

Abstract

The invention discloses a hand vein identification method and device for double living verification, and relates to the field of identity verification. The method includes: acquiring physiological parameters of the hand to be verified; performing in vivo verification of action features when the physiological parameters are greater than a first preset threshold; successfully calculating the palm similarity and the back of the hand similarity in the in vivo verification; If the preset threshold is set, the hand vein recognition is successful. The present invention adopts the combination of action feature and physiological parameter detection to perform double in vivo verification, extracts palm vein images and dorsal hand vein images respectively for vein identification, and improves reliability through double in vivo verification and dual vein identification of palm vein images and dorsal hand vein images. It is especially suitable for occasions with particularly high reliability requirements.

Description

A method and device for recognizing hand veins for double living verification

technical field

The present invention relates to the field of identity verification, in particular to a method and device for identifying hand veins for double living verification.

Background technique

Vein recognition is to obtain images of hand veins through a near-infrared camera, and store the digital images of veins in a computer system to achieve feature value storage. When comparing veins, real-time vein maps are taken, and advanced filtering, image binarization and refinement methods are used to extract features from digital images, and complex matching algorithms are used to compare and match the vein feature values stored in the computer system host. Thereby identifying and confirming the identity of the individual. Veins are located inside the human body and are not affected by rough skin and external environment. The use of vein identification has the advantages of high accuracy, difficult to copy, safe and convenient, and has been used in access control, social security and other fields.

Most of the devices for vein recognition in the market use a single vein recognition method in the back of the hand, palm or fingers, with relatively few feature points, and the recognition reliability and accuracy are relatively low. Moreover, although the veins are hidden in the human skin, it is not easy to leave traces, and it is more difficult to forge than other biometric methods, but it can still be forged and cracked. The most common cracking method is to draw lines on paper, wear rubber gloves, and wear rubber gloves. For example, at the Chaos Communication Congress hacking conference held in Leipzig, Germany in 2018, researchers Jan Krissler and Julian Albrecht successfully deceived the vein authentication system through a wax hand model; the easiest way to crack is to use a vein collection device to obtain The vein pattern map, copy the vein pattern map of the finger, and use the picture to crack. Therefore, the existing vein identification devices have the problems of low identification reliability and easy cracking.

SUMMARY OF THE INVENTION

The purpose of the present invention is to provide a method and device for hand vein identification with double living verification, which solves the problems of low identification reliability and easy cracking of the existing vein identification equipment.

For achieving the above object, the present invention provides the following scheme:

A dual-liveness verification method for identifying hand veins, comprising:

Obtain the physiological parameters of the hand to be verified detected by the physiological parameter detection sensor;

If the physiological parameter is greater than the first preset threshold, performing in vivo verification of the motion characteristics of the hand to be verified to obtain a verification result;

If the verification result indicates that the in vivo verification is successful, acquiring the palm vein image and the dorsal hand vein image of the hand to be verified;

Obtain pre-stored palm vein storage images and dorsal vein storage images;

respectively calculating the palm similarity between the palm vein image and the palm vein stored image, and the palm similarity between the palm vein image and the palm vein stored image;

If both the similarity of the palm and the similarity of the back of the hand are greater than the second preset threshold, the hand vein identification of the hand to be verified is successful.

Optionally, the obtaining of the physiological parameters of the hand to be verified detected by the physiological parameter detection sensor specifically includes:

Acquire the pulse blood oxygen value of the hand to be verified detected by the blood oxygen heart rate sensor.

Optionally, the in vivo verification of the motion characteristics of the to-be-verified hand to obtain a verification result specifically includes:

obtaining the fist image of the hand to be verified;

If the to-be-verified hand in the clenched fist image is in a clenched fist state, acquiring a half-clenched fist image of the to-be-verified hand;

If the to-be-verified hand in the half-clenched fist image is in a half-clenched fist state, acquiring hand images of the to-be-verified hand in different clenched fist states between the clenched fist state and the half-clenched fist state;

Use the optical flow method to detect the optical flow feature of the hand image, and compare the similarity between the optical flow feature of the hand image and the pre-stored optical flow feature;

If the similarity is greater than the third preset threshold, the in vivo verification is successful;

If the similarity is less than or equal to the third preset threshold, the living body verification fails.

Optionally, the optical flow feature of the hand image is detected by an optical flow method, and the similarity between the optical flow feature of the hand image and the pre-stored optical flow feature is compared, specifically including:

Taking the hand image of the ith frame as the first comparison image, and extracting the region of interest of the first comparison image to obtain the first region of interest;

Taking the hand image of the i+1th frame as the second comparison image, and extracting the region of interest of the second comparison image to obtain the second region of interest;

Using the optical flow method to detect the positions in the first region of interest and the second region of interest where the fluctuation range is greater than a fourth preset threshold, to obtain optical flow characteristics;

Let i=i+1, return to the step "use the hand image of the i-th frame as the first comparison image, and extract the region of interest of the first comparison image to obtain the first region of interest", and obtain all the Optical flow features of hand images;

Get pre-stored optical flow features;

All the optical flow features are compared with corresponding pre-stored optical flow features respectively to obtain the similarity.

A hand vein identification device for double living verification, comprising: a top collection area, a hand placement panel area and a bottom collection area; the top collection area and the hand placement panel area are both connected to the bottom collection area;

The top collection area is located at the top of the hand vein identification device, and the top collection area is used to collect the dorsal hand vein image of the hand to be verified;

The hand placement panel area is located below the top collection area, and the hand placement panel area is used to collect the physiological parameters of the hand to be verified;

The bottom collection area is located at the bottom of the hand vein identification device and below the hand placement panel area, and the bottom collection area is used to obtain living body verification information, and process the living body verification information to obtain The hand vein identification result; the living verification information includes the dorsal hand vein image, the physiological parameter and the palm vein image of the hand to be verified.

Optionally, the top collection area specifically includes: a top camera, a top light source and a display screen;

The top camera is used to collect the dorsal vein image of the hand to be verified; the output end of the top camera is connected to the bottom acquisition area;

The top light source is arranged corresponding to the top camera, and the top light source is used to illuminate the back of the hand to be verified;

The display screen is connected to the output end of the bottom collection area, and the display screen is used to display the hand vein recognition result of the bottom collection area.

Optionally, the hand placement panel area specifically includes: a hand placement area, a hand texture map, a protrusion and a physiological parameter detection sensor;

The hand placement area is located just below the top camera, and the hand placement area is used to place the to-be-verified hand;

The hand texture map is engraved in the center of the hand placement area, the protrusion is located at the first joint of the middle finger of the hand texture map, and the hand texture map and the protrusion are used to display the to-be-verified the placement of the hand;

The physiological parameter detection sensor is located under the protrusion, and the physiological parameter detection sensor is used for detecting the physiological parameter.

Optionally, the bottom collection area specifically includes: a bottom camera, a bottom light source, a hand vein recognition system, and a power supply;

the bottom camera is used to collect the palm vein image of the hand to be verified; the output end of the bottom camera is connected to the input end of the hand vein identification system;

The bottom light source is arranged corresponding to the bottom camera, and the bottom light source is used to illuminate the palm of the hand to be verified;

The hand vein identification system is respectively connected with the top camera and the physiological parameter detection sensor; the hand vein identification system is used to obtain the living body verification information, process the living body verification information, and obtain the hand vein identification system. External vein identification results;

The power source is respectively connected with the top light source, the bottom light source, the physiological parameter detection sensor and the hand vein identification system, and the power source is used for the top light source, the bottom light source, the physiological parameter The parameter detection sensor and the hand vein identification system are powered.

Optionally, the hand vein identification system specifically includes:

a physiological parameter acquisition module, used for acquiring the physiological parameters of the hand to be verified detected by the physiological parameter detection sensor;

an action feature in vivo verification module, configured to perform action feature in vivo verification on the to-be-verified hand when the physiological parameter is greater than a first preset threshold to obtain a verification result;

a vein image acquisition module, configured to acquire the palm vein image and the dorsal vein image of the hand to be verified when the verification result indicates that the in vivo verification is successful;

The vein storage image acquisition module is used to acquire the pre-stored palm vein storage image and dorsal palm vein storage image;

a similarity calculation module, configured to calculate the palm similarity between the palm vein image and the palm vein stored image, and the palm similarity between the palm vein image and the palm vein stored image, respectively;

A hand vein identification module, configured to successfully identify the hand veins of the to-be-verified hand when both the palm similarity and the hand back similarity are greater than a second preset threshold.

Optionally, the action feature liveness verification module specifically includes:

a clenched fist image obtaining unit, used for obtaining the clenched fist image of the hand to be verified;

a half fist image acquiring unit, configured to acquire a half fist image of the hand to be verified when the hand to be verified in the fist image is in a fist state;

A hand image acquisition unit, configured to acquire the to-be-verified hands in different clenched fist states between the clenched fist state and the half clenched fist state when the to-be-verified hand in the half-clenched fist image is in a half-clenched fist state hand image of the hand;

a similarity comparison unit, configured to detect the optical flow feature of the hand image by using the optical flow method, and compare the similarity between the optical flow feature of the hand image and the pre-stored optical flow feature;

a successful living body verification unit, used for successful living body verification when the similarity is greater than a third preset threshold;

A living body verification failure unit, configured to fail the living body verification when the similarity is less than or equal to the third preset threshold.

According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects:

The present invention provides a method and device for recognizing hand veins for double living verification. The method includes: acquiring the physiological parameters of the hand to be verified detected by the physiological parameter detection sensor; if the physiological parameter is greater than a first preset threshold, performing in vivo verification of motion characteristics of the hand to be verified to obtain a verification result; if the verification result indicates in vivo verification If successful, the palm vein image and the dorsal vein image of the hand to be verified are obtained; the pre-stored palm vein storage image and the palm vein storage image are obtained; the palm similarity of the palm vein image and the palm vein storage image, as well as the palm vein image and the palm vein image are calculated respectively. The dorsal hand veins store the hand back similarity of the image; if both the palm similarity and the hand back similarity are greater than the second preset threshold, the hand veins of the hand to be verified are successfully identified. The present invention adopts the combination of action feature and physiological parameter detection to perform double in vivo verification, extracts palm vein images and dorsal hand vein images respectively for vein identification, and improves reliability through double in vivo verification and dual vein identification of palm vein images and dorsal hand vein images. It is especially suitable for occasions with particularly high reliability requirements.

Description of drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the accompanying drawings required in the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some of the present invention. In the embodiments, for those of ordinary skill in the art, other drawings can also be obtained according to these drawings without creative labor.

1 is a flowchart of a method for identifying a vein in a hand according to an embodiment of the present invention;

FIG. 2 is a structural diagram of a hand vein identification device according to an embodiment of the present invention;

3 is a circuit connection diagram of a hand vein identification device according to an embodiment of the present invention;

Fig. 4 is the bottom view of the top collection area of the embodiment of the present invention;

5 is a structural diagram of a hand-placed panel area according to an embodiment of the present invention;

6 is a structural diagram of a protrusion according to an embodiment of the present invention;

FIG. 7 is a data processing flow chart of a blood oxygen heart rate sensor according to an embodiment of the present invention;

8 is a top view of a bottom collection area according to an embodiment of the present invention;

FIG. 9 is a flow chart of the use of the hand vein identification device according to the embodiment of the present invention;

10 is a flowchart of an identity registration process according to an embodiment of the present invention;

FIG. 11 is a flowchart of an action feature live body registration according to an embodiment of the present invention;

FIG. 12 is a flowchart of processing a stored picture according to an embodiment of the present invention;

13 is a flowchart of an identity verification process according to an embodiment of the present invention;

FIG. 14 is a flowchart of an action feature biometric verification according to an embodiment of the present invention;

FIG. 15 is a flowchart of processing a hand image according to an embodiment of the present invention.

Symbol description: 1. Top acquisition area; 2. Hand placement panel area; 3. Bottom acquisition area; 4. Top camera; 5. Top light source; 6. Display screen; 7. Hand placement area; 8. Hand texture map; 9. Protrusion; 10. Physiological parameter detection sensor; 11. Label; 12. Bottom camera; 13. Bottom light source; 14. Main controller; 15. Power supply.

Detailed ways

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

The purpose of the present invention is to provide a method and device for hand vein identification with double living verification, which solves the problems of low identification reliability and easy cracking of the existing vein identification equipment.

In order to make the above objects, features and advantages of the present invention more clearly understood, the present invention will be described in further detail below with reference to the accompanying drawings and specific embodiments.

This embodiment provides a method for recognizing hand veins by double living verification. FIG. 1 is a flowchart of the method for recognizing hand veins according to an embodiment of the present invention. Referring to Figure 1, the hand vein identification method includes:

Step 101: Obtain the physiological parameters of the hand to be verified detected by the physiological parameter detection sensor.

Step 101 specifically includes:

Obtain the pulse oxygen value of the hand to be verified detected by the blood oxygen heart rate sensor.

Step 102, if the physiological parameter is greater than the first preset threshold, perform in vivo verification of the motion characteristics of the hand to be verified to obtain a verification result. The first preset threshold is 90.

Perform in vivo verification of motion characteristics of the hand to be verified, and obtain verification results, including:

Get the fist image of the hand to be verified.

If the hand to be verified in the clenched fist image is in a clenched fist state, a half-clenched fist image of the hand to be verified is acquired.

If the hand to be verified in the half-clenched fist image is in a half-clenched fist state, the hand images of the to-be-verified hand in different fist-clenched states between the fist-clenched state and the half-clenched fist state are acquired.

Use the optical flow method to detect the optical flow feature of the hand image, and compare the similarity between the optical flow feature of the hand image and the pre-stored optical flow feature, including:

The ith frame of the hand image is used as the first comparison image, and the region of interest of the first comparison image is extracted to obtain the first region of interest.

The i+1 th frame of the hand image is used as the second comparison image, and the region of interest of the second comparison image is extracted to obtain the second region of interest.

The optical flow method is used to detect the positions in the first region of interest and the second region of interest where the fluctuation range is greater than the fourth preset threshold, to obtain the optical flow feature. This step is based on the principle that the optical flow method can detect the position of the moving object.

In this embodiment, by collecting images of the process of each person clenching a fist to a half-clenching fist, the optical flow method can detect the changes of the joints on the back of the hand during the process, and the joint changes of each person are different, so as to achieve the function of living body recognition.

Let i=i+1, return to the step "use the i-th frame of hand image as the first comparison image, and extract the region of interest of the first comparison image to obtain the first region of interest", and obtain the optical flow of all hand images feature.

Get pre-stored optical flow features.

All optical flow features are compared with the corresponding pre-stored optical flow features to obtain the similarity.

If the similarity is greater than the third preset threshold, the in vivo verification is successful.

If the similarity is less than or equal to the third preset threshold, the in vivo verification fails.

Step 103, if the verification result indicates that the in vivo verification is successful, acquire the palm vein image and the dorsal vein image of the hand to be verified.

Step 104: Acquire pre-stored palm vein storage images and dorsal palm vein storage images.

Step 105: Calculate the palm similarity between the palm vein image and the palm vein stored image, and the palm similarity between the palm vein image and the palm vein stored image, respectively.

Step 106 , if both the palm similarity and the back similarity are greater than the second preset threshold, the hand veins of the hand to be verified are successfully identified.

The optical flow method can detect the instantaneous speed of the pixel motion of the space moving object on the observation imaging surface. When the object is moving, the brightness pattern of its corresponding point on the image is also moving accordingly. The apparent motion of this image brightness pattern is optical flow. The study of optical flow is to use the temporal variation and correlation of the intensity data of pixels in an image sequence to determine the "motion" of the respective pixel positions. Optical flow expresses changes in the image and can therefore be used by the observer to determine the movement of objects. Typically, optical flow results from camera motion, object motion in the scene, or a combination of both.

The optical flow field is derived from the optical flow, which refers to the two-dimensional instantaneous velocity field formed by the projection of the three-dimensional velocity vector of the visible pixels in the scene on the imaging surface. When the motion field in space is transferred to the image, it is represented as an optical flow field, and the optical flow field reflects the gray-scale variation trend of each point on the image. The optical flow field contains the motion information of the observed object and information about the rich three-dimensional structure of the scene.

The optical flow method detects moving objects, and its basic idea is to assign a velocity vector to each pixel in the image, thereby forming the motion field of the image. There is a one-to-one correspondence between the points on the image and the points on the three-dimensional object at a specific motion moment, and the image is dynamically analyzed according to the velocity vector characteristics of each pixel. If there is no moving target in the image, the optical flow vector changes continuously in the entire image area, and when there is relative motion between the object and the image background, the velocity vector formed by the moving object must be different from the velocity of the neighborhood background. vector, so that the position of the moving object can be detected.

This embodiment also provides a dual-living verification device for hand vein recognition. FIG. 2 is a structural diagram of the device for recognizing hand veins according to an embodiment of the present invention; and FIG. 2 and 3, the hand vein identification device includes: a top collection area 1, a hand placement panel area 2 and a bottom collection area 3; the top collection area 1 and the hand placement panel area 2 are connected to the bottom collection area 3.

The top collection area 1 is located on the top of the hand vein identification device. The top collection area 1 is used to collect the dorsal hand vein images and hand movements of the hand to be verified, and to display the hand vein recognition results of the bottom collection area 3.

The hand placement panel area 2 is located below the top collection area 1, and the hand placement panel area 2 is used for placing the hand to be verified and collecting the physiological parameters of the hand to be verified.

The bottom collection area 3 is located at the bottom of the hand vein identification device, and is located below the hand placement panel area 2, and the bottom collection area 3 is used to obtain the living body verification information, and process the living body verification information to obtain the hand vein identification result; The verification information includes dorsal hand vein images, physiological parameters, and palm vein images of the hand to be verified. The bottom collection area 3 processes the living body verification information through the hand vein identification system, which can be implemented by the main controller 14 .

FIG. 4 is a bottom view of the top collection area of the embodiment of the present invention. Referring to FIG. 4 , the top acquisition area 1 specifically includes: a top camera 4 , a top light source 5 and a display screen 6 .

The top camera 4 is used to collect the dorsal vein image of the hand to be verified; the output end of the top camera 4 is connected to the bottom collection area 3 . The top camera 4 is also used to collect the hand movements of the hand to be verified, and transmit the collected image data to the hand vein recognition system in the bottom collection area.

The top light source 5 is arranged corresponding to the top camera 4, and the top light source 5 is used to illuminate the back of the hand to be verified. The top light source 5 is connected to the power source of the bottom collection area, and the power source of the bottom collection area supplies power to the top light source. The top light source 5 is a near-infrared lamp group, and the near-infrared lamp group includes 6 near-infrared lamps with a diameter of 850 nanometers (nm).

The display screen 6 is connected to the output end of the bottom collection area, and the display screen 6 is used to display the hand vein recognition result in the bottom collection area. The display screen 6 adopts a touch screen, and is also used for human-computer interaction, and the human-computer interaction includes user operation prompts.

FIG. 5 is a structural diagram of a hand-placed panel area according to an embodiment of the present invention. Referring to FIG. 5 , the hand placement panel area 2 specifically includes: a hand placement area 7 , a hand pattern diagram 8 , a protrusion 9 and a physiological parameter detection sensor 10 .

The hand placement area 7 is located just below the top camera 4, and the hand placement area 7 is used to place the hand to be authenticated. The hand placement area 7 is a glass panel, preferably a transparent acrylic plate, which is also used to support hand placement.

The hand texture picture 8 is engraved in the center of the hand placement area 7 . FIG. 6 is a structural diagram of a protrusion according to an embodiment of the present invention. Referring to FIG. 6 , the protrusion 9 is located at the first joint of the middle finger of the hand pattern of FIG. 8 . The hand texture map 8 and the protrusions 9 are used to display the placement position of the hand to be verified. The height of the protrusions 9 is 1 millimeter (mm). When using the palm face down, the first joint of the middle finger of the hand is placed directly above the bulge 9, and the middle finger is placed in the middle finger area of the hand texture map; the use of the bulge and the hand texture map can ensure the hand to be verified. The positions of the parts are relatively fixed each time.

The physiological parameter detection sensor 10 is located under the protrusion 9, and the physiological parameter detection sensor 10 is used to detect the physiological parameter. The physiological parameter detection sensor needs to be in close contact with the finger, so it is placed under the hand placement area. The physiological parameter detection sensor 10 is connected to the bottom collection area 3, and the physiological parameter detection sensor transmits the collected physiological parameters to the hand vein identification system in the bottom collection area. The physiological parameter detection sensor 10 is specifically located below the middle finger region of the hand pattern in Fig. 8 near the protrusion 9, and determines whether the hand to be verified is a living body by detecting the physiological parameters of the middle finger. The physiological parameter detection sensor 10 adopts the MAX30100 blood oxygen heart rate sensor, which can detect the physiological parameters such as the blood oxygen heart rate of the human body. The physiological parameter detection sensor transmits the detected physiological parameters to the hand vein recognition system, and the physiological parameter is the pulse of the hand to be verified. blood oxygen value.

The MAX30100 is a sensor that integrates pulse oximeter and heart rate detection. The blood oxygen heart rate sensor integrates a red LED (Light Emitting Diode, light-emitting diode), an infrared LED, optical devices, photoelectric sensors, and a belt environment. The light-suppressed, low-noise electronic circuit is the smallest, low-power pulse oximetry sensor in the industry.

The detection of the blood oxygen heart rate sensor is based on photoplethysmography. Photoplethysmography is mainly based on the light absorption characteristics of human body tissue to detect changes in the light intensity of reflected light and/or transmitted light through photoelectric sensors to determine blood volume changes. The blood oxygen heart rate sensor irradiates light with a known wavelength to human body tissues. Through the Lambert-Beer law, it can be known that the absorption of light by human skin, muscles and other non-blood tissues is unchanged, but the expansion of the human heart will cause blood volume and The pressure changes periodically, and the heart beat is compared to the change of light intensity. Therefore, the light intensity detected by the photoelectric sensor can know the heart beat, that is, the pulse. At the same time, the size of the vascular volume depends on the size of the ejection volume. When the heart contracts, the ejection volume increases, and the vascular volume increases; on the contrary, when the heart relaxes, the ejection volume decreases, so the vascular volume decreases, and the vascular volume decreases. Increase means that the absorption of light by blood increases, and the light intensity received by the receiving end decreases due to the constant light intensity of the incident light source; conversely, the decrease in blood vessel volume means that the absorption of light by blood decreases, and the light intensity received by the receiving end decreases. The pulse wave is obtained by photoelectric conversion.

FIG. 7 is a data processing flow chart of a blood oxygen heart rate sensor according to an embodiment of the present invention. Referring to Figure 7, the working principle of the blood oxygen heart rate sensor for measuring blood oxygen saturation pulse is to drive the LED built in the photoelectric sensor through the LED driver inside the photoelectric sensor to alternately irradiate red light and infrared light according to the preset timing, and then photoelectric The sensor collects the reflected light signal, and the photoelectric sensor converts the reflected light signal into an analog electrical signal. The analog electrical signal output by the photoelectric sensor is amplified and filtered, and then converted into a digital signal by analog-to-digital conversion. Finally, the digital signal is stored in the FIFO (First Input First Output, first-in, first-out) buffer. The MAX30100 sensor can communicate through serial ports. Pulse oximetry can be obtained by sending a digital signal to the main controller via the I ² C bus. After receiving the pulse blood oxygen value, the main controller judges whether the pulse blood oxygen value is greater than 90; if the pulse blood oxygen value is greater than 90, it will pass the in vivo verification of the physiological parameters; verify.

The hand placement panel area 2 also includes a label 11 located at the upper left of the hand placement area 7 . The label 11 is a prompt label, and the label 11 is used to indicate the use steps and precautions of the hand vein identification device, which is convenient for the first-time user and the user with less use to operate.

8 is a top view of the bottom collection area according to the embodiment of the present invention. Referring to FIG. 8 , the bottom collection area 3 specifically includes: a bottom camera 12 , a bottom light source 13 , a hand vein recognition system and a power supply 15 .

The bottom camera 12 is used to collect the palm vein image of the hand to be verified; the output end of the bottom camera is connected to the input end of the hand vein identification system, and specifically transmits the collected image to the system through a universal serial bus (Universal Serial Bus, USB). Hand vein recognition system, the hand vein recognition system processes the image. The top camera 4 and bottom camera 12 should correspond to the center of the hand texture map 8 . The frame rate of the top camera 4 and the bottom camera 12 is greater than 30.

The bottom light source 13 is disposed corresponding to the bottom camera 12, and the bottom light source 13 is used to illuminate the palm of the hand to be verified. The bottom light source 13 is a near-infrared lamp group, and the near-infrared lamp group includes 6 near-infrared lamps with a diameter of 850 nanometers (nm).

According to the absorption principle of near-infrared light by hemoglobin, the top collection area and the bottom collection area collect the vein images of the palm and the back of the hand respectively under the irradiation of the near-infrared lamp. In this embodiment, the vein images of the back of the hand and the palm need to be identified and verified at the same time, which improves the security of vein biometric identification.

The hand vein identification system is also connected with the top camera 4 and the physiological parameter detection sensor 10 respectively, and the input end of the hand vein identification system is also connected with the output end of the top camera 4 and the physiological parameter detection sensor 10 respectively; In order to obtain the living body verification information, and process the living body verification information, the hand vein identification result is obtained. The hand vein recognition system is also used for information interaction with the touch screen, such as: accepting instructions from the touch screen, and displaying the hand vein recognition result on the touch screen. The hand vein recognition system obtains the result of hand vein recognition through living detection. The living detection includes two parts: the living detection of physiological parameters and the living detection of action characteristics. Through the combination of these two living detection methods, it is almost impossible to forge the prosthesis; The processing of the living body verification information includes the processing and recognition of the vein images of the palm and the back of the hand, and the processing of the living body detection.

The power source 15 is respectively connected with the top light source 5, the bottom light source 13, the physiological parameter detection sensor 10 and the hand vein identification system, and the power source 15 is used to supply power for the top light source 5, the bottom light source 13, the physiological parameter detection sensor 10 and the hand vein identification system , that is, supply voltage.

The hand vein recognition system specifically includes:

The physiological parameter acquisition module is used for acquiring the physiological parameters of the hand to be verified detected by the physiological parameter detection sensor.

The physiological parameter acquisition module specifically includes:

The physiological parameter acquisition unit is used for acquiring the pulse blood oxygen value of the hand to be verified detected by the blood oxygen heart rate sensor.

The action feature in vivo verification module is used for performing action feature in vivo verification on the hand to be verified when the physiological parameter is greater than the first preset threshold to obtain the verification result.

The action feature liveness verification module specifically includes:

The fist image acquisition unit is used to acquire the fist image of the hand to be verified.

The half clenched fist image acquisition unit is configured to acquire the half clenched fist image of the to-be-verified hand when the to-be-verified hand in the clenched fist image is in a clenched fist state.

The hand image acquisition unit is configured to acquire hand images of the to-be-verified hand in different fist-clenched states between the fist-clenched state and the half-clenched fist state when the to-be-verified hand in the half-clenched fist image is in a half-clenched fist state.

The similarity comparison unit is used to detect the optical flow feature of the hand image by using the optical flow method, and compare the similarity between the optical flow feature of the hand image and the pre-stored optical flow feature.

The similarity comparison unit specifically includes:

The first region of interest extraction subunit is used for taking the i-th frame of the hand image as the first comparison image, and extracting the region of interest of the first comparison image to obtain the first region of interest.

The second region of interest extraction subunit is used for taking the i+1th frame of the hand image as the second comparison image, and extracting the region of interest of the second comparison image to obtain the second region of interest.

The optical flow feature detection subunit is configured to use the optical flow method to detect the positions in the first region of interest and the second region of interest whose fluctuation range is greater than a fourth preset threshold to obtain the optical flow feature.

Returning to the subunit, for setting i=i+1, returning to the first region of interest extraction subunit to obtain the optical flow features of all hand images.

The pre-stored optical flow feature acquisition subunit is used to obtain the pre-stored optical flow features.

The similarity calculation subunit is used to compare all optical flow features with the corresponding pre-stored optical flow features to obtain the similarity.

The living body verification success unit is used to confirm the living body verification is successful when the similarity is greater than the third preset threshold.

The living body verification failure unit is configured to fail the living body verification when the similarity is less than or equal to the third preset threshold.

The vein image acquisition module is used to acquire the palm vein image and the dorsal vein image of the hand to be verified when the verification result indicates that the in vivo verification is successful.

The vein storage image acquisition module is used for acquiring pre-stored palm vein storage images and dorsal palm vein storage images.

The similarity calculation module is used for calculating the palm similarity between the palm vein image and the palm vein storage image, and the palm similarity between the palm vein image and the palm vein storage image, respectively.

The hand vein identification module is used for successfully identifying the hand veins of the hand to be verified when both the similarity of the palm and the similarity of the back of the hand are greater than the second preset threshold.

The use process of the hand vein recognition device mainly includes: identity registration and identity verification; identity registration is to store the biometric information of the user, and the biometric information includes the biometrics of the user's hand; identity verification is to extract the biometrics of the hand to be verified and Matching with the pre-stored biometrics, the pre-stored biometrics are the biometric information of the user stored during identity registration; the usage process of identity registration and identity verification is basically the same. FIG. 9 is a flow chart of the use of the hand vein recognition device according to the embodiment of the present invention. Referring to FIG. 9 , the use process of the hand vein recognition device is: touch screen function selection, the touch screen function selection includes identity registration and identity verification, and the registration function is selected through the touch screen Perform identity registration; select the verification function through the touch screen to authenticate.

10 is a flowchart of an identity registration process according to an embodiment of the present invention. Referring to FIG. 10 , the identity registration process includes: placing the hand to be registered in a designated area as required; the requirements are: the back of the hand is up, the palm is down, and the middle finger is placed In the middle finger region of the palm image of the hand texture map, and the first joint of the middle finger is placed on the bulge.

Obtain the physiological parameters detected by the physiological parameter detection sensor, and verify the liveness through the physiological parameters.

It is judged whether the verification of the physiological parameter to verify the liveness is passed; if the verification is passed, the liveness is verified by the action feature; otherwise, the liveness verification fails.

Action feature verification liveness: Obtain the optical flow feature of the hand to be registered when performing the registered action.

Stores optical flow features.

The hand images are collected by the top camera (top camera) to obtain the vein map of the back of the hand; the hand images are collected by the bottom camera (bottom camera) to obtain the palm vein map. The hand image captured by the top camera is the hand image including the fingers and the back of the hand, and the dorsal vein map is the image that only includes the back of the hand. The hand image collected by the bottom camera is the hand image including the fingers and palm, and the palm vein map is the image that only includes the palm part.

The dorsal hand vein image is segmented from the dorsal hand vein map. The dorsal hand vein image is an image containing part of the veins on the back of the hand.

The palm vein image is segmented from the palm vein map. The palm vein image is an image containing veins in part of the palm.

Calculate the image feature values of the dorsal hand vein image and the palm vein image respectively, store them, and return to the step "respectively collect the hand image through the top camera (top camera) to obtain the dorsal hand vein map; collect the hand image through the bottom camera (bottom camera), Obtain palm vein map", the cycle is executed three times, and the image feature values obtained by the three calculations are stored.

FIG. 11 is a flowchart of action feature liveness registration according to an embodiment of the present invention. Referring to FIG. 11 , the action feature verification liveness includes:

The display screen prompts "please make a fist", the top camera collects the hand picture, and judges whether it is a fist state through the hand image; if it is a fist state, the display screen prompts "please make a half fist", and starts to store the collected hand pictures; Otherwise, go back to the step "the top camera captures the hand picture, and judges whether it is in the fist state through the hand image".

The top camera collects the hand picture, and judges whether it is in a half fist state through the hand image; if it is a half fist state, stop storing the hand picture; otherwise, go back to step "The top camera collects the hand picture and judges by the hand image. Is it a half-clawed fist?"

Use the optical flow method to process the stored hand picture to obtain the optical flow feature; refer to Figure 12 for details, extract the hand picture stored in the second frame, and extract the region of interest in the hand picture stored in the second frame, this implementation In the example, the region of interest is between the first joint of the finger and the wrist, that is, the back of the hand; the extracted hand picture is compared with the previous frame of pictures, and the optical flow of the region of interest of the previous frame of pictures needs to be extracted before the comparison. feature, the previous frame picture is the hand picture stored in the previous frame of the currently extracted hand picture; judge whether the extracted hand picture is the hand picture stored in the last frame; if it is the hand picture stored in the last frame, then Draw the optical flow angle and modulus value of the hand image stored in the first frame, and use the optical flow angle and modulus value as the optical flow feature; if it is not the hand image stored in the last frame, extract the hand stored in the next frame. picture, and return to the step "Compare the extracted hand picture with the previous frame of picture, and extract the optical flow features of the region of interest of the previous frame of picture before comparing." The optical flow angle and modulus value of the hand image stored in the first frame are compared to the speed u in the x direction (horizontal direction) and the velocity u in the y direction (vertical direction) of each pixel in the hand image stored in the two adjacent frames. The distribution of the velocity v in the vertical direction), depicting the velocity curve of each point from the first frame to the last frame, the cut angle between the starting point and the ending point of the velocity curve is the optical flow angle, and the distance between the starting point and the ending point is the modulus value, and the optical flow angle and modulus are the eigenvalues of the optical flow feature of the starting point. If the cut angle and distance are greater than the preset threshold, the starting point is determined as a feature point, and the eigenvalues of the starting point are stored.

Stores optical flow features.

13 is a flowchart of an identity verification process according to an embodiment of the present invention. Referring to FIG. 13 , the identity verification process includes: placing the hand to be verified in a designated area as required; the requirements are: the back of the hand is up, the palm is down, and the middle finger is placed In the middle finger region of the palm image of the hand texture map, and the first joint of the middle finger is placed on the bulge.

Obtain the physiological parameters detected by the physiological parameter detection sensor, and verify the liveness through the physiological parameters.

It is judged whether the verification of the physiological parameter to verify the liveness is passed; if the verification is passed, the liveness is verified by the action feature; otherwise, the liveness verification fails.

Judging whether the verification of action feature verification of liveness is passed; if it passes the verification, the top camera (top camera) is used to collect the hand image to obtain the vein map of the back of the hand, and the hand image is collected through the bottom camera (bottom camera) to obtain the palm vein map. ; otherwise, the liveness verification fails.

The dorsal hand vein image is segmented from the dorsal hand vein map.

The palm vein image is segmented from the palm vein map.

Calculate the similarity between the dorsal hand vein image and the pre-stored dorsal vein image, and the similarity between the palm vein image and the pre-stored palm vein image, judge whether the hand vein identification verification is passed according to the similarity, and display the hand vein on the touch screen. Identify verification results. The pre-stored dorsal hand vein storage image and the pre-stored palm vein storage image are the dorsal hand vein image and the palm vein image stored at the time of identity registration.

The action feature verification of liveness includes: the touch screen prompts to make a fist, the top camera detects the fist image, starts to detect the half fist motion, and ends when the half fist motion is detected. Touch screen prompt: Please make a fist. After the top camera detects the fist motion, the touch screen prompts: Please make a half fist. If the fist motion is not detected within 2s, it will prompt that the liveness verification fails. Extract all the pictures between the two actions from the fist-clenching action to the half-clenching fist action, from the beginning of the fist-clenching action to the end of the half-clenching fist action, compare the pictures of two adjacent frames, and track the position with a large change in the picture by the optical flow method , and save the position with larger variation as the optical flow feature, and the saved optical flow feature is: the position of the point with larger variation on the back of the hand and the direction information of the movement of the point. Repeat the comparison between the pictures of two adjacent frames, compare the similarity between the extracted optical flow feature and the pre-stored optical flow feature pair, and judge the similarity according to the third preset threshold. The basis of action feature verification of liveness is: each person's joint movement from clenched fist to half clenched fist is specific, and the optical flow characteristics of the entire clenched fist to half clenched fist are used as the judgment basis. There are few optical flow features here, and the third preset threshold is low, which is only used as a preliminary identity screening.

See Figure 14 for details on the flow of action feature liveness verification:

The display screen prompts "please make a fist", the top camera collects the hand picture, and judges whether it is a fist state through the hand image; if it is a fist state, the display screen prompts "please make a half fist", and starts to store the collected hand pictures; Otherwise, go back to the step "the top camera captures the hand picture, and judges whether it is in the fist state through the hand image".

The top camera collects the hand picture, and judges whether it is in a half fist state through the hand image; if it is a half fist state, stop storing the hand picture; otherwise, go back to step "The top camera collects the hand picture and judges by the hand image. Is it a half-clawed fist?"

Use the optical flow method to process the stored hand pictures to obtain the optical flow characteristics, and compare the optical flow characteristics with the optical flow characteristics stored during the identity registration. If the similarity is greater than the third preset threshold, it means living body verification Passed; otherwise, it indicates that the liveness verification failed.

In this embodiment, the state of making a fist is determined as if the first joint of the index finger, middle finger and ring finger of the hand is identified as protruding, and the state of making a fist can be determined. The state of the half-clenched fist is judged as the state of the second joint protruding during the process from clenching the fist to the state of the palm being flat. This state is the half-clenched fist state. Therefore, by identifying the protrusion of the second joint of the index finger, middle finger and ring finger, you can It is judged to be a half fist state.

Fig. 15 is a flow chart of hand image processing according to an embodiment of the present invention. Referring to Fig. 15, the step "collect hand image through the top camera (top camera), obtain the vein map of the back of the hand, and collect the hand image through the bottom camera (bottom camera) at the same time. , obtain the palm vein map; segment the palm vein image from the palm vein map; segment the palm vein image from the palm vein map; The similarity of the palm vein stored images, judge whether the verification of hand vein identification is passed according to the similarity, and display the verification result of hand vein identification on the touch screen.” Specifically, it includes: the top camera and the bottom camera collect the hand image; the dorsal hand vein map and The palm vein image is segmented to obtain the dorsal hand vein image and the palm vein image; image normalization, Niblack binarization, median filtering, SIFT (Scale-invariant feature transform, scale-invariant feature transform) are performed on the dorsal hand vein image and the palm vein image in turn. variable feature transformation) feature extraction to obtain the back of the hand feature and the palm feature; perform feature matching with the pre-stored back of the hand feature and the palm feature respectively to obtain a matching score; judge whether the matching score is greater than the second preset threshold; if it matches If the score is greater than the second preset threshold, the hand vein identification verification is successful; if the matching score is less than or equal to the second preset threshold, the hand vein identification verification fails; the back of the hand feature and the palm feature are stored. The algorithm used in this part is relatively common, and the completion of this part is only to ensure the process integrity of the entire hand vein recognition device. The pre-stored features of the back of the hand are obtained from the stored images of the back of the hand veins stored during the identity registration, and the pre-stored palm features are obtained from the stored images of the palm veins stored during the identity registration.

The present invention firstly uses the blood oxygen heart rate sensor to detect whether there is real blood flow, and uses the blood oxygen heart rate sensor to detect the pulse blood oxygen value to determine whether the hand to be verified is a living body, which can decipher the forgery of the prosthesis; When doing fist-clenching and half-clenching movements, the beating of each person's joints on the back of the hand is different from fist-clenching to half-clenching. The optical flow method is used to analyze the vitality of the fist according to the path of the joint beating and the fist-clenching movement, and judge the waiting time. Verifying whether the hand is a living body improves the reliability of the living body verification; extracts the vein patterns on the palm and the back of the hand, and adopts the combination of the back of the hand and the palm to improve the safety and reliability.

The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments, and the same and similar parts between the various embodiments can be referred to each other. For the system disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant part can be referred to the description of the method.

In this paper, specific examples are used to illustrate the principles and implementations of the present invention. The descriptions of the above embodiments are only used to help understand the methods and core ideas of the present invention; meanwhile, for those skilled in the art, according to the present invention There will be changes in the specific implementation and application scope. In conclusion, the contents of this specification should not be construed as limiting the present invention.

Claims (10)

1. a hand vein identification method for double living verification, is characterized in that, comprises: Obtain the physiological parameters of the hand to be verified detected by the physiological parameter detection sensor; If the physiological parameter is greater than the first preset threshold, performing in vivo verification of the motion characteristics of the hand to be verified to obtain a verification result; If the verification result indicates that the in vivo verification is successful, acquiring the palm vein image and the dorsal hand vein image of the hand to be verified; Obtain pre-stored palm vein storage images and dorsal vein storage images; respectively calculating the palm similarity between the palm vein image and the palm vein stored image, and the palm similarity between the palm vein image and the palm vein stored image; If both the similarity of the palm and the similarity of the back of the hand are greater than the second preset threshold, the hand vein identification of the hand to be verified is successful. 2. The method for identifying hand veins for dual in vivo verification according to claim 1, wherein the acquiring the physiological parameters of the hand to be verified detected by the physiological parameter detection sensor specifically comprises: Acquire the pulse blood oxygen value of the hand to be verified detected by the blood oxygen heart rate sensor. 3. The method for identifying hand veins for dual in vivo verification according to claim 1, wherein the in vivo verification of the motion characteristics of the hand to be verified is performed to obtain a verification result, specifically comprising: obtaining the fist image of the hand to be verified; If the to-be-verified hand in the clenched fist image is in a clenched fist state, acquiring a half-clenched fist image of the to-be-verified hand; If the to-be-verified hand in the half-clenched fist image is in a half-clenched fist state, acquiring hand images of the to-be-verified hand in different clenched fist states between the clenched fist state and the half-clenched fist state; Use the optical flow method to detect the optical flow feature of the hand image, and compare the similarity between the optical flow feature of the hand image and the pre-stored optical flow feature; If the similarity is greater than the third preset threshold, the in vivo verification is successful; If the similarity is less than or equal to the third preset threshold, the living body verification fails. 4 . The method for recognizing hand veins for dual living verification according to claim 3 , wherein the optical flow feature of the hand image is detected by using an optical flow method, and the optical flow of the hand image is compared. 5 . The similarity between the feature and the pre-stored optical flow feature, including: Taking the hand image of the ith frame as the first comparison image, and extracting the region of interest of the first comparison image to obtain the first region of interest; Taking the hand image of the i+1th frame as the second comparison image, and extracting the region of interest of the second comparison image to obtain the second region of interest; Using the optical flow method to detect the positions in the first region of interest and the second region of interest where the fluctuation range is greater than a fourth preset threshold, to obtain optical flow characteristics; Let i=i+1, return to the step "use the hand image of the i-th frame as the first comparison image, and extract the region of interest of the first comparison image to obtain the first region of interest", and obtain all the Optical flow features of hand images; Get pre-stored optical flow features; All the optical flow features are compared with corresponding pre-stored optical flow features respectively to obtain the similarity. 5. A hand vein identification device for double living verification, characterized in that it comprises: a top collection area, a hand placement panel area, and a bottom collection area; the top collection area and the hand placement panel area are all connected to the bottom Collection area connection; The top collection area is located at the top of the hand vein identification device, and the top collection area is used to collect the dorsal hand vein image of the hand to be verified; The hand placement panel area is located below the top collection area, and the hand placement panel area is used to collect the physiological parameters of the hand to be verified; The bottom collection area is located at the bottom of the hand vein identification device and below the hand placement panel area, and the bottom collection area is used to obtain living body verification information, and process the living body verification information to obtain The hand vein identification result; the living verification information includes the dorsal hand vein image, the physiological parameter and the palm vein image of the hand to be verified. 6. The hand vein identification device for dual living verification according to claim 5, wherein the top collection area specifically comprises: a top camera, a top light source and a display screen; The top camera is used to collect the dorsal vein image of the hand to be verified; the output end of the top camera is connected to the bottom acquisition area; The top light source is arranged corresponding to the top camera, and the top light source is used to illuminate the back of the hand to be verified; The display screen is connected to the output end of the bottom collection area, and the display screen is used to display the hand vein recognition result of the bottom collection area. 7 . The hand vein identification device for dual living verification according to claim 6 , wherein the hand placement panel area specifically comprises: a hand placement area, a hand texture map, a protrusion and a physiological parameter detection sensor; 8 . The hand placement area is located just below the top camera, and the hand placement area is used to place the to-be-verified hand; The hand texture map is engraved in the center of the hand placement area, the protrusion is located at the first joint of the middle finger of the hand texture map, and the hand texture map and the protrusion are used to display the to-be-verified the placement of the hand; The physiological parameter detection sensor is located under the protrusion, and the physiological parameter detection sensor is used for detecting the physiological parameter. 8. The hand vein identification device for dual living verification according to claim 7, wherein the bottom collection area specifically comprises: a bottom camera, a bottom light source, a hand vein identification system and a power supply; the bottom camera is used to collect the palm vein image of the hand to be verified; the output end of the bottom camera is connected to the input end of the hand vein identification system; The bottom light source is arranged corresponding to the bottom camera, and the bottom light source is used to illuminate the palm of the hand to be verified; The hand vein identification system is respectively connected with the top camera and the physiological parameter detection sensor; the hand vein identification system is used to obtain the living body verification information, process the living body verification information, and obtain the hand vein identification system. External vein identification results; The power source is respectively connected with the top light source, the bottom light source, the physiological parameter detection sensor and the hand vein identification system, and the power source is used for the top light source, the bottom light source, the physiological parameter The parameter detection sensor and the hand vein recognition system are powered. 9. The hand vein identification device for dual living verification according to claim 8, wherein the hand vein identification system specifically comprises: a physiological parameter acquisition module, used for acquiring the physiological parameters of the hand to be verified detected by the physiological parameter detection sensor; an action feature in vivo verification module, configured to perform action feature in vivo verification on the to-be-verified hand when the physiological parameter is greater than a first preset threshold to obtain a verification result; a vein image acquisition module, configured to acquire the palm vein image and the dorsal vein image of the hand to be verified when the verification result indicates that the in vivo verification is successful; The vein storage image acquisition module is used to acquire the pre-stored palm vein storage image and dorsal palm vein storage image; a similarity calculation module, configured to calculate the palm similarity between the palm vein image and the palm vein stored image, and the palm similarity between the palm vein image and the palm vein stored image, respectively; A hand vein identification module, configured to successfully identify the hand veins of the to-be-verified hand when both the palm similarity and the hand back similarity are greater than a second preset threshold. 10. The hand vein identification device for dual living verification according to claim 9, wherein the action feature living verification module specifically comprises: a clenched fist image obtaining unit, used for obtaining the clenched fist image of the hand to be verified; a half fist image acquiring unit, configured to acquire a half fist image of the hand to be verified when the hand to be verified in the fist image is in a fist state; A hand image acquisition unit, configured to acquire the to-be-verified hands in different clenched fist states between the clenched fist state and the half clenched fist state when the to-be-verified hand in the half-clenched fist image is in a half-clenched fist state hand image of the hand; a similarity comparison unit, configured to detect the optical flow feature of the hand image by using the optical flow method, and compare the similarity between the optical flow feature of the hand image and the pre-stored optical flow feature; a successful living body verification unit, used for successful living body verification when the similarity is greater than a third preset threshold; A living body verification failure unit, configured to fail the living body verification when the similarity is less than or equal to the third preset threshold.

Images