Living body detection method and device and corresponding electronic equipment

Abstract

A method and device for detecting a living body and a corresponding electronic device are disclosed. The method for detecting a living body comprises: determining at least three key points of a detection object in an image; determining at least one plane in a space according to the at least three key points; determining a normal vector of each plane in the at least one plane; and determining whether the detection object is a living body according to the normal vector of each plane. By using the method and device for detecting a living body and a corresponding electronic device according to the embodiments of the present disclosure, live body detection can be performed more accurately.

Description

Technical Field

[0001] The present disclosure generally relates to the technical field of pattern recognition, and in particular to a living body detection method and device and corresponding electronic equipment. Background Art

[0002] With the development of biometric technology, face recognition technology has become more stable and mature. However, illegal users can deceive the face recognition system through photos, videos, face masks, etc., and the accuracy of face detection and recognition will also be reduced in the case of poor face posture and lighting. Therefore, the accuracy of face liveness detection is very important for application scenarios with high security requirements such as access control and login. Summary of the invention

[0003] According to one aspect of the present disclosure, a living body detection method is provided, which includes: determining at least three key points of a detection object in an image; determining at least one plane in space based on the at least three key points; determining a normal vector of each plane in the at least one plane; and determining whether the detection object is a living body based on the normal vector of each plane.

[0004] According to another aspect of the present disclosure, a living body detection device is provided, which includes: a key point positioning unit, configured to determine at least three key points of a detection object in an image; a normal vector determination unit, configured to determine at least one plane in the space according to the at least three key points and determine the normal vector of each plane in the at least one plane; and a classification unit, configured to determine whether the detection object is a living body according to the normal vector of each plane.

[0005] According to another aspect of the present disclosure, an electronic device is provided, comprising a processor and a memory for storing instructions executable by the processor. The processor may be configured to read the instructions from the memory and execute the instructions to implement the above-mentioned living body detection method.

[0006] According to another aspect of the present disclosure, a computer-readable storage medium is provided, on which program instructions are stored. When the program instructions are executed by a computer, the above-mentioned living body detection method is performed.

[0007] By using the living body detection method and device and corresponding electronic equipment according to the embodiments of the present disclosure, living body detection can be performed more accurately. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] The above and other purposes, features and advantages of the present disclosure will become more apparent by describing the embodiments of the present disclosure in more detail in conjunction with the accompanying drawings. The accompanying drawings are used to provide a further understanding of the embodiments of the present disclosure and constitute a part of the specification. Together with the embodiments of the present disclosure, they are used to explain the present disclosure and do not constitute a limitation of the present disclosure. In the drawings, the same reference numerals generally represent the same components or steps.

[0009] Figure 1 It is an example of a living body detection method according to an embodiment of the present disclosure.

[0010] Figure 2 is an example of a living body detection device according to an embodiment of the present disclosure.

[0011] Figure 3 is an example of a living body detection device according to an embodiment of the present disclosure.

[0012] Figure 4 is an example of an electronic device according to an embodiment of the present disclosure. DETAILED DESCRIPTION

[0013] Below, the exemplary embodiments according to the present disclosure will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present disclosure, rather than all the embodiments of the present disclosure, and it should be understood that the present disclosure is not limited to the exemplary embodiments described here.

[0014] Overview

[0015] As mentioned above, illegal users can deceive the face recognition system through photos, videos, face masks, etc. This can effectively counter such attacks and improve the distinction between live and non-live features. It is expected to provide enhanced depth information, so that more fine-grained features can be extracted to improve the accuracy of liveness detection.

[0016] Exemplary Methods

[0017] Figure 1 An example of a living body detection method according to an embodiment of the present disclosure is shown, which may be applied to an electronic device and may include steps S110 , S120 , S130 , and S140 .

[0018] like Figure 1 As shown, in step S110, at least three key points of the detection object in the image may be determined.

[0019] According to different embodiments, the image in step S110 may be a single-channel or multi-channel image including any appropriate information obtained by any appropriate means.

[0020] For example, the image in step S110 may be an image captured by an appropriate binocular image acquisition device, such as one of a left image and a right image captured by a binocular camera for a detection object; it may also be one of two or more images that can be used for stereo vision matching obtained by any appropriate means; it may also be a single image captured by a monocular camera that works in conjunction with a depth image acquisition device such as an infrared depth camera, a laser radar, etc. In addition, each pixel in the image in step S110 itself may also include depth information, for example, a channel representing the depth value of each pixel may be further included on the basis of the color channel.

[0021] According to different embodiments, the detection object may include any type of object of interest such as a human body, certain parts of a human body, an animal, etc. For example, the detection object may be a human face or a part of a human body including a human face area.

[0022] According to different embodiments, the key points of the detection object may be at least three landmarks in any form that can locate or characterize key parts or key areas of the detection object on the image.

[0023] For example, when the detection object is a human face, the key points may be at least three marks that can locate the positions of key areas or key parts of the human face such as eyebrows, eyes, nose, mouth, facial contour, etc. in a given image.

[0024] All key points or one or more shape vectors formed by these points can characterize or represent the geometric shape of the detection object or the geometric shape of one or more key parts or key areas of the detection object.

[0025] Then, in step S120, at least one plane in the space may be determined according to the determined at least three key points.

[0026] According to different embodiments, one of the at least three key points determined or a point other than the at least three key points determined may be used as the origin, and a three-dimensional coordinate system may be established based on the relative positional relationship between each key point and the origin. In such a three-dimensional coordinate space, mathematically, every three points may determine a plane, thereby determining at least one plane. More details will be described later.

[0027] Then, in step S130 , the normal vector of each plane may be determined.

[0028] Mathematically, a vector represented by a straight line perpendicular to a plane is a normal vector of the plane, and each plane has countless normal vectors. In one embodiment, a normal vector for each plane can be determined for subsequent processing.

[0029] Then, in step S140, it can be determined whether the detected object is a living body according to the normal vector of each plane.

[0030] According to different embodiments, the normal vector of each plane can be provided as feature data to a liveness classifier constructed based on any appropriate type of model with any appropriate structure, such as a support vector machine, a convolutional neural network, a logistic regression, etc. The present disclosure is not limited to a specific type and / or structure of a classifier or model for liveness detection using a plane normal vector.

[0031] In the method according to the embodiment of the present disclosure, at least three key points of the detection object are first determined, and then at least one plane in the space is determined based on each key point, and then the normal vector of each plane is provided as input feature data to the liveness detection model. The plane normal vector feature can integrate the structural features of the detection object, so the method according to the embodiment of the present disclosure can significantly improve the accuracy of liveness recognition.

[0032] The following describes more details of the method according to the embodiment of the present disclosure by taking human face liveness detection as an example. However, it should be understood that the object that can be detected by the method according to the embodiment of the present disclosure is not limited to human face.

[0033] According to different embodiments, in step S110, any appropriate method and / or model such as a convolutional neural network can be used to perform face detection on the input image, and then any appropriate method and / or model can be used to locate / mark the key points of the detected partial face image; or a method and / or model that can directly mark or locate the key points of the face from the input image can be used to determine the key points of the face.

[0034] For example, in step S110, a key point detection method / model such as an active appearance model, a constrained local model, a cascaded shape regression, or a deep alignment network may be used to determine the key points of the detection object in the input image.

[0035] According to different embodiments, the number of the at least three key points determined may be any appropriate number such as 3, 5, 14, 34, 68, etc. For example, 51 internal key points related to the facial features may be determined, and 17 contour key points related to the outer contour of the face may be determined, thereby determining 68 key points.

[0036] It should be understood that the present disclosure is not limited to the methods and / or models used for face detection and key point localization / labeling, nor is it limited to the number of key points determined.

[0037] The key points determined by step S110 may have a relative position relationship on the plane of the input image. For example, when one of the at least three key points determined or a point other than the at least three key points determined is used as the coordinate origin on the image plane, each key point may have a corresponding two-dimensional coordinate on the image plane.

[0038] Then, the method according to the embodiment of the present disclosure may proceed to step S120 .

[0039] According to one embodiment, in step S120, the depth value of each key point of the at least three key points can be determined, and then the coordinates of each key point in the space can be determined based on the depth value of each key point and the coordinates of each key point in the image, and then one of the at least one planes can be determined based on the coordinates of every three key points in the space.

[0040] In this embodiment, the three-dimensional coordinates of each key point in space are determined according to the depth value of each key point and the two-dimensional coordinates of each key point in the image plane, and then at least one plane in the space is determined, so that the plane normal vector can reflect the depth features while fusing the structural features, thereby providing more and finer-grained feature information for more accurate living body recognition.

[0041] According to different embodiments, the depth value of each key point may be determined in various appropriate ways according to the circumstances.

[0042] For example, as described above, in step S110, while obtaining a conventional image of the detection object through a camera, a corresponding depth image can be obtained using a depth image acquisition device such as an infrared depth camera, a laser radar, etc., and then the pixels in the conventional image and the depth image can be associated. In this way, for each key point of the detection object determined based on the conventional image, the depth value of each key point can be determined according to the association relationship between the pixels in the conventional image and the depth image.

[0043] In another embodiment, determining the depth value of each of the at least three key points may include: determining the disparity value of each key point based on the coordinate difference of each of the at least three key points in the image and another image, wherein the image and the other image are from a binocular image acquisition device; and determining the depth value of each key point based on the disparity value of each key point.

[0044] For example, the image in step S110 may be an image captured by an appropriate binocular image acquisition device (e.g., a binocular camera), and accordingly, the disparity value of each key point may be determined according to the coordinate difference of each key point in the image and another image among the at least three key points, wherein the image and the another image are from the binocular image acquisition device, for example, they may be one and the other of the left image and the right image captured by the binocular camera for the detection object, respectively. Then, the depth value of each key point may be determined according to the disparity value of each key point.

[0045] In order to facilitate obtaining the coordinate difference of each key point in the image and the other image, in one embodiment, the method according to the embodiment of the present disclosure may further include calibrating the binocular image acquisition device so that each point on the same photographed object has the same vertical coordinate value in the image and the other image. Then, the image and the other image of the detection object may be captured using the calibrated binocular image acquisition device.

[0046] Then, for example, key points at the center of rigid regions such as the center of the nose, the center of the eyebrows, etc. can be set as the coordinate origin, and the two-dimensional coordinates of each key point on the image plane of image IMG1 and / or IMG2 can be determined based on the relative position relationship between the key points.

[0047] If the coordinates of a key point P on images IMG1 and IMG2 are (x 1 ,y 1 ) and (x 2 ,y 2 ), since the binocular image acquisition device has been corrected by, for example, epipolar correction, each point on the same photographed object has the same ordinate value in IMG1 and IMG2, that is, y 1 =y 2 Then, the parallax value V of P can be determined according to the following equation (1): PAX .

[0048] V PAX =x 1 -x 2 (1)

[0049] Then, for example, the following equation (2) can be used to calculate the parallax value V of P: PAX Determine the depth value of P:

[0050] V DEP =B*F / V PAX (2)

[0051] Among them, B may be the baseline length of the binocular image acquisition device, and F may be the focal length of the binocular image acquisition device.

[0052] Then, for example, for the key point P mentioned above, its three-dimensional coordinates in space can be determined as (x 1 ,y 1 ,V DEP ).

[0053] In this embodiment, the disparity value of each key point is determined by using the coordinate difference of the key point in the two images from the binocular image acquisition device. This can avoid complicated alignment and coordination operations when obtaining the corresponding depth image using a depth image acquisition device such as an infrared depth camera, a lidar, etc., and can also avoid the use of a complicated stereo matching network, thereby improving the processing efficiency of living body detection and reducing hardware costs and the complexity of software control.

[0054] In another embodiment, correction may be performed so that each point on the same photographed object has the same abscissa value in IMG1 and IMG2, and the difference between the ordinate values ​​is used as the coordinate difference of the point on IMG1 and IMG2.

[0055] According to different embodiments, the depth value of each point (or key point) may be an absolute depth or a relative depth value.

[0056] For example, in step S120, the initial depth value of each key point may be determined according to the disparity value of each key point, and then the depth value of each key point may be determined according to the difference between the initial depth value of each key point and the average value of the initial depth values ​​of all key points.

[0057] For example, assuming that the key point P is determined in step S110 1 , P 2 ,……,P n (where n can be an integer greater than or equal to 3), and for example, through the above equation (2), the initial depth value of each key point is V 1 、V 2 ,……,V n , then for any key point P i (1≤i≤n), its final depth value D i It can be determined according to the following equation (3):

[0058] D i =V i -(Σ 1≤j≤n V j ) / n (3)

[0059] By using relative depth values, the influence of different distances from the camera lens to the detected object on the detection result can be removed, thereby effectively enhancing the robustness of living body detection at different distances.

[0060] Then, in step S130 , a normal vector of each plane may be selected for subsequent processing.

[0061] For example, if the equation of a plane determined in step S120 is ax+by+cz=d, then in step S130, a normal vector (a, b, c) may be determined for subsequent processing.

[0062] Then, in step S140, the normal vector of each plane determined in step S130 may be provided to a classifier such as a support vector machine, a neural network, or a logistic regression, and whether the detected object is a living body may be determined based on the classification result of the classifier.

[0063] In another embodiment, in step S140, it is also possible to determine whether the detected object is a living body based on the normal vector of each plane and the depth value of each key point and / or information such as a directional gradient histogram generated based on the image.

[0064] For example, the feature vector including the normal vector of each plane, the depth value of each key point, and the directional gradient histogram generated according to the image can be provided to a classifier such as a support vector machine, a neural network, a logistic regression, etc., and whether the detected object is a living body is determined according to the classification result of the classifier. By providing more quantity and more dimensions of feature information, the accuracy of liveness detection can be further enhanced.

[0065] As described above, in the liveness detection method according to the embodiment of the present disclosure, a plane of key points can be constructed, and the feature information of the plane normal vector can be extracted, thereby being able to more richly express the spatial position relationship of the detection object, thereby being able to enhance the distinctiveness of the depth information feature. In addition, a spatial coordinate system can also be established based on relative depth, so that good robustness can be obtained under different distances. In addition, the liveness detection method according to the embodiment of the present disclosure can be independent of complex stereo matching algorithms and / or additional depth information acquisition devices, but can only use key point positioning technology to restore depth information, thereby being able to significantly improve the calculation speed and real-time performance, and reduce the implementation cost.

[0066] Exemplary Devices

[0067] Figure 2An example of a liveness detection device according to an embodiment of the present disclosure is shown. The device can implement the liveness detection method according to an embodiment of the present disclosure and can include a key point positioning unit 210, a normal vector determination unit 220 and a classification unit 230.

[0068] The key point positioning unit 210 may be configured to determine at least three key points of the detection object in the image, that is, configured to implement step S110 of the living body detection method according to the embodiment of the present disclosure.

[0069] According to different embodiments, the key point positioning unit 210 may include a general-purpose processor such as a central processing unit and a graphics processing unit, or may be a dedicated processor developed based on a field programmable gate array, etc. For example, in the case of performing object detection and / or key point labeling based on a convolutional neural network, the key point positioning unit 210 may also include a multiplication and addition unit array, an adder array, a distortion operator and other elements for accelerating operations such as convolution, pooling, point-by-point addition, activation, etc., as well as a static random access memory for caching data, etc.

[0070] The normal vector determination unit 220 can be configured to determine at least one plane in the space according to the at least three key points and determine the normal vector of each plane in the at least one plane, that is, configured to implement steps S120 and S130 of the living body detection method according to an embodiment of the present disclosure.

[0071] According to different embodiments, the normal vector determination unit 220 may also include a general-purpose processor such as a central processing unit and a graphics processor, or may be a dedicated processor developed based on a field programmable gate array, etc.

[0072] In one embodiment, Figure 3 As shown, the normal vector determination unit 220 may include a depth value calculation module 2210 and a plane determination module 2220, wherein the depth value calculation module 2210 may be configured to determine the depth value of each key point, and the plane determination module 2220 may be configured to determine the coordinates of each key point in the space based on the depth value of each key point and the coordinates in the image and to determine one of the at least one planes based on the coordinates of every three key points in the space.

[0073] The classification unit 230 may be configured to determine whether the detection object is a living body according to the normal vector of each plane, that is, configured to implement step S140 of the living body detection method according to the embodiment of the present disclosure.

[0074] Similar to the key point positioning unit 210, according to different embodiments, the classification unit 230 may also include a general-purpose processor such as a central processing unit and a graphics processing unit, or may be a dedicated processor developed based on a field programmable gate array, etc. For example, in the case of performing liveness detection based on a convolutional neural network, the classification unit 230 may also include elements such as a multiplication and addition unit array, an adder array, a distortion operator, etc. for accelerating operations such as convolution, pooling, point-by-point addition, activation, etc., and a static random access memory for caching data, etc.

[0075] In one embodiment, the above-mentioned units and / or modules may reuse one or more computing acceleration components. In another embodiment, the functions of the above-mentioned units and / or modules may be implemented by one or more general or dedicated processors such as a central processing unit, a graphics processing unit, a field programmable gate array, etc.

[0076] According to different embodiments, the above-mentioned units and / or modules may be interconnected through various appropriate means such as a bus, a crossbar, a shared memory, etc.

[0077] It should be understood that Figure 2 or Figure 3 This is only an example of a device according to an embodiment of the present disclosure, and the present disclosure is not limited thereto. For example, in another example, the device according to an embodiment of the present disclosure may also include a memory for storing intermediate data or result data and / or one or more interfaces for receiving data or transmitting detection results to the outside.

[0078] Exemplary Electronic Devices

[0079] like Figure 4 As shown, an embodiment of the present disclosure may also be an electronic device, and the electronic device 400 may include one or more processors 410 and a memory 420 .

[0080] The processor 410 is a central processing unit (CPU) or other forms of processing units having data processing capabilities and / or instruction execution capabilities, and may control other components in the electronic device 400 to perform desired functions.

[0081] The memory 420 may include one or more computer program products, which may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory may include, for example, a random access memory (RAM) and / or a cache memory (cache), etc. The non-volatile memory may include, for example, a read-only memory (ROM), a hard disk, a flash memory, etc. One or more computer program instructions may be stored on the computer-readable storage medium, and the processor 410 may run the program instructions to implement the liveness detection method of each embodiment of the present application described above and / or other desired functions. Various contents such as input signals, signal components, noise components, etc. may also be stored in the computer-readable storage medium.

[0082] In one example, the electronic device 400 may further include: an input device 430 and an output device 440 , and these components are interconnected via a bus system and / or other forms of connection mechanisms (not shown).

[0083] The input device 430 may also include, for example, a keyboard, a mouse, etc.

[0084] The output device 440 can output various information to the outside, including the determined distance information, direction information, etc. The output device 440 can include, for example, a display, a speaker, a printer, a communication network and a remote output device connected thereto, and the like.

[0085] Of course, to simplify, Figure 4 Only some of the components related to the present application in the electronic device 400 are shown, and components such as a bus, an input / output interface, etc. are omitted. In addition, according to specific application situations, the electronic device may also include any other appropriate components.

[0086] Exemplary computer program products and computer-readable storage media

[0087] In addition to the above-mentioned methods and devices, an embodiment of the present disclosure may also be a computer program product, which includes computer program instructions, which, when executed by a processor, enable the processor to execute the steps of the method according to various embodiments of the present disclosure described in the above-mentioned "Exemplary Method" section of this specification.

[0088] The computer program product may be written in any combination of one or more programming languages ​​to write program code for performing the operations of the disclosed embodiments, and the programming language may include an object-oriented programming language, such as Java, C++, etc., and a conventional procedural programming language, such as "C" or a similar programming language. The program code may be executed entirely on the user computing device, partially on the user device, as a separate software package, partially on the user computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0089] In addition, an embodiment of the present disclosure may also be a computer-readable storage medium, such as a computer-readable non-temporary storage medium, on which program instructions are stored, and when the program instructions are executed by a processor, the processor executes the steps in the method according to various embodiments of the present disclosure described in the above “Exemplary Method” section of this specification.

[0090] Computer readable storage media can adopt any combination of one or more readable media. The readable medium can be a readable signal medium or a readable storage medium. The readable storage medium can include, for example, but is not limited to, a system, device or device of electricity, magnetism, light, electromagnetic, infrared, or semiconductor, or any combination of the above. More specific examples (non-exhaustive list) of readable storage media include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0091] The basic principles of the present disclosure are described above in conjunction with specific embodiments. However, it should be noted that the advantages, strengths, effects, etc. mentioned in the present disclosure are only examples and not limitations, and it cannot be considered that these advantages, strengths, effects, etc. are required by each embodiment of the present disclosure. In addition, the specific details disclosed above are only for the purpose of illustration and ease of understanding, and are not limitations. The above details do not limit the present disclosure to the necessity of adopting the above specific details to be implemented.

[0092] The block diagrams of the devices, apparatuses, equipment, and systems involved in this disclosure are only illustrative examples and are not intended to require or imply that they must be connected, arranged, and configured in the manner shown in the block diagrams. As will be appreciated by those skilled in the art, these devices, apparatuses, equipment, and systems can be connected, arranged, and configured in any manner. Words such as "including," "comprising," "having," and the like are open words, referring to "including but not limited to," and can be used interchangeably therewith. The words "or" and "and" used herein refer to the words "and / or," and can be used interchangeably therewith, unless the context clearly indicates otherwise. The word "such as" used herein refers to the phrase "such as but not limited to," and can be used interchangeably therewith.

[0093] It should also be noted that in the apparatus, device and method of the present disclosure, each component or each step can be decomposed and / or recombined. Such decomposition and / or recombination should be regarded as equivalent solutions of the present disclosure.

[0094] In this document, modifiers such as "first", "second", etc. without quantifiers are intended to distinguish different elements / components / circuits / modules / devices / steps, and are not used to emphasize the order, positional relationship, importance, priority, etc. In contrast, modifiers such as "first", "second", etc. with quantifiers can be used to emphasize the order, positional relationship, importance, priority, etc. of different elements / components / circuits / modules / devices / steps.

[0095] The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use the present disclosure. Various modifications to these aspects will be readily apparent to those skilled in the art, and the general principles defined herein may be applied to other aspects without departing from the scope of the present disclosure. Therefore, the present disclosure is not intended to be limited to the aspects shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

[0096] The above description has been given for the purpose of illustration and description. In addition, this description is not intended to limit the embodiments of the present disclosure to the forms disclosed herein. Although multiple example aspects and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, changes, additions and sub-combinations thereof.

Claims

1. A method for detecting a living body, comprising: Determine at least three key points of the detected object in the image; Determine at least one plane in the space according to the depth values ​​of each of the at least three key points; determining a normal vector for each of the at least one plane; as well as Determine whether the detected object is a living body according to the normal vector of each plane, Wherein, determining at least one plane in the space according to the depth values ​​of each of the at least three key points comprises: Determining a depth value for each of the at least three key points; Determine the coordinates of each key point in the space according to the depth value of each key point and the coordinates of each key point in the image; and Determine one of the at least one plane according to the coordinates of each three key points in the space, Wherein, determining the depth value of each key point of the at least three key points comprises: Determine a disparity value of each key point according to a coordinate difference of each key point of the at least three key points in the image and another image, wherein the image and the another image are from a binocular image acquisition device; and The depth value of each key point is determined based on the disparity value of each key point.

2. The method for detecting a living body according to claim 1, wherein: Determining the depth value of each key point based on the disparity value of each key point includes: Determining an initial depth value for each key point based on the disparity value of each key point; and The depth value of each key point is determined based on the difference between the initial depth value of each key point and the average of the initial depth values ​​of all key points.

3. The liveness detection method according to claim 1, further comprising: Calibrate the binocular image acquisition device so that each point on the same photographed object has the same vertical coordinate value in the image and the other image; as well as The image and the other image of the detection object are captured by using the binocular image acquisition device.

4. The method for detecting a living body according to any one of claims 1 to 3, wherein: Determining whether the detected object is a living body according to the normal vector of each plane includes: Whether the detected object is a living body is determined according to at least one of a depth value of each key point and a directional gradient histogram generated according to the image and a normal vector of each plane.

5. A living body detection device, comprising: A key point locating unit, configured to determine at least three key points of a detection object in an image; a normal vector determining unit, configured to determine at least one plane in the space and a normal vector of each plane in the at least one plane according to the depth values ​​of each of the at least three key points; as well as a classification unit, configured to determine whether the detected object is a living body according to the normal vector of each plane, Wherein, the normal vector determination unit comprises: A depth value calculation module, configured to determine a depth value of each key point; and a plane determination module, configured to determine the coordinates of each key point in the space according to the depth value of each key point and the coordinates in the image, and to determine one of the at least one plane according to the coordinates of every three key points in the space, wherein the depth value calculation module determines the disparity value of each key point according to the coordinate difference of each key point of the at least three key points in the image and another image, and the image and the another image are from a binocular image acquisition device; and The depth value of each key point is determined based on the disparity value of each key point.

6. An electronic device comprising: processor; a memory for storing instructions executable by the processor; The processor is configured to read the instruction from the memory and execute the instruction to implement the living body detection method according to any one of claims 1 to 4. 7 . A computer-readable storage medium having program instructions stored thereon, wherein the program instructions, when executed by a computer, perform the living body detection method according to any one of claims 1 to 4.

Images