WO2016197298A1

WO2016197298A1 - Living body detection method, living body detection system and computer program product

Info

Publication number: WO2016197298A1
Application number: PCT/CN2015/080964
Authority: WO
Inventors: 曹志敏; 贾开
Original assignee: 北京旷视科技有限公司; 北京小孔科技有限公司
Priority date: 2015-06-08
Filing date: 2015-06-08
Publication date: 2016-12-15
Also published as: CN105612533A; CN105612533B

Abstract

The present disclosure relates to a living body detection method, a living body detection system and a computer program product, which can realize living human body detection. The living body detection method comprises: acquiring video data collected by a video data collection apparatus; based on the video data, determining an object to be detected; acquiring a signal to be detected corresponding to the object to be detected; and determining whether the signal to be detected is a physiological signal of a living body, wherein the signal to be detected is a skin elasticity signal corresponding to the object to be detected.

Description

Living body detection method, living body detection system, and computer program product

Technical field

The present disclosure relates to the field of living body detection, and more particularly, to a living body detecting method, a living body detecting system, and a computer program product capable of realizing human body living body detection.

Background technique

At present, face recognition systems are increasingly used in security, finance and other fields that require authentication, such as bank remote account opening, access control systems, and remote transaction operation verification. In these high security level applications, in addition to ensuring that the face similarity of the verifier matches the database data stored in the database, the first person to be verified is a legal living organism. That is to say, the face recognition system needs to be able to prevent an attacker from using a photo, a 3D face model or a mask to attack.

The method of solving the above problem is usually called living body detection, and the purpose is to judge whether the acquired biometrics are from a living, on-the-spot, real person. There is currently no proven proven biometric solution in the technology products on the market. The existing biometric detection technology relies on special hardware devices (such as infrared cameras, depth cameras) or can only prevent simple static photo attacks.

Summary of the invention

The present disclosure has been made in view of the above problems. The present disclosure provides a living body detecting method, a living body detecting system, and a computer program product based on a common monocular camera, which can effectively prevent photos by detecting skin elastic characteristics in a video image sequence of a subject for living body detection. , 3D face models and mask attacks.

According to an embodiment of the present disclosure, a living body detecting method includes: acquiring video data collected via a video data collecting device; determining an object to be detected based on the video data; and acquiring a to-be-detected object Detecting a signal; and determining whether the signal to be detected is a living physiological signal, wherein the signal to be detected is a skin elasticity signal corresponding to the object to be detected.

Further, the living body detecting method according to an embodiment of the present disclosure, wherein determining the object to be detected based on the video data includes determining a face image therein as the object to be detected based on the video data, and determining the At least one key area in the face image.

Further, the living body detecting method according to an embodiment of the present disclosure, wherein the determining at least one key region in the face image includes determining a key point in the face image based on the video data, based on the key The point divides the face image into the at least one key area.

Further, the living body detecting method according to an embodiment of the present disclosure, wherein the acquiring the signal to be detected corresponding to the object to be detected includes: acquiring a pre-action region before and after a predetermined time point corresponding to the at least one key region An image and a post-action area image, the predetermined time point being a point in time at which the object to be detected performs a predetermined action.

Furthermore, the living body detecting method according to an embodiment of the present disclosure, wherein the acquiring the signal to be detected corresponding to the object to be detected further comprises: normalizing the pre-action area image and the post-action area image into A grayscale image having a predetermined size, and the normalized pre-action region image and the normalized post-action region image are superimposed as the to-be-detected signal.

Further, the living body detecting method according to an embodiment of the present disclosure, wherein the acquiring the signal to be detected corresponding to the object to be detected further comprises: setting the post-action area image and the predetermined range around the post-action area image The relevant area image is normalized to a grayscale image having the predetermined size as the signal to be detected.

Further, the living body detecting method according to an embodiment of the present disclosure, wherein determining whether the signal to be detected is a living body physiological signal comprises: comparing the signal to be detected with a preset living condition, and matching the signal to be detected in the signal to be detected When the living condition is set, the signal to be detected is determined to be a living physiological signal, wherein the preset living condition is a skin elasticity signal corresponding to the living body acquired based on the preset preset video data.

Furthermore, the living body detecting method according to an embodiment of the present disclosure further includes: starting the detection timing while acquiring the video data collected via the video data collecting device; and not when the detection timing reaches the preset time threshold In a case where it is determined whether the signal to be detected is a living physiological signal, it is determined that the signal to be detected is not a living physiological signal.

According to another embodiment of the present disclosure, a living body detection system is provided, comprising: a processor; a memory; and computer program instructions stored in the memory, when the computer program instructions are executed by the processor The following steps: acquiring video data collected by the video data collecting device; determining an object to be detected based on the video data; acquiring a signal to be detected corresponding to the object to be detected; and determining whether the signal to be detected is a living physiological signal The signal to be detected is a skin elasticity signal corresponding to the object to be detected.

Further, a living body detecting system according to another embodiment of the present disclosure, wherein the computer Determining, based on the video data, the determining of the object to be detected based on the video data, determining a face image therein as the object to be detected, and determining the face image At least one key area in the middle.

Further, a living body detecting system according to another embodiment of the present disclosure, wherein the step of determining at least one key region in the face image when the computer program instructions are executed by the processor includes: based on the video Data, determining a key point in the face image, and dividing the face image into the at least one key area based on the key point.

Further, a living body detecting system according to another embodiment of the present disclosure, wherein the step of acquiring the signal to be detected corresponding to the object to be detected when the computer program instruction is executed by the processor includes: obtaining a correspondence And a pre-action area image and a post-action area image before and after the predetermined time point of the at least one key area, where the predetermined time point is a time point at which the object to be detected performs a predetermined action.

Further, a living body detecting system according to another embodiment of the present disclosure, wherein the step of acquiring the signal to be detected corresponding to the object to be detected when the computer program instruction is executed by the processor further comprises: The pre-action area image and the post-action area image are normalized to a grayscale image having a predetermined size, and the normalized pre-action area image and the normalized post-action area image are superimposed as The signal to be detected.

Further, a living body detecting system according to another embodiment of the present disclosure, wherein the step of acquiring the signal to be detected corresponding to the object to be detected when the computer program instruction is executed by the processor further comprises: The post-action area image and the relevant area image of the predetermined range around the post-action area image are normalized to the gray image having the predetermined size as the signal to be detected.

Furthermore, a living body detecting system according to another embodiment of the present disclosure, wherein the step of determining whether the signal to be detected is a living body physiological signal when the computer program instruction is executed by the processor comprises: comparing the to-be-detected The signal and the preset living condition are determined, when the to-be-detected signal matches the preset living condition, determining that the to-be-detected signal is a living physiological signal, wherein the preset living condition is obtained based on pre-acquired preset video data. Corresponds to the skin elasticity signal of the living body.

Furthermore, a living body detection system according to another embodiment of the present disclosure further includes a detection timer, wherein when the computer program instructions are executed by the processor: acquiring the video data acquired via the video data collection device At the same time, the detection timer is started to execute the detection timing; In the case that the detection timing does not determine whether the signal to be detected is a living physiological signal when the detection timing reaches the preset time threshold, it is determined that the signal to be detected is not a living physiological signal.

According to still another embodiment of the present disclosure, a computer program product is provided, comprising a computer readable storage medium on which computer program instructions are stored, the computer program instructions being executed while being executed by a computer The following steps: acquiring video data collected by the video data collecting device; determining an object to be detected based on the video data; acquiring a signal to be detected corresponding to the object to be detected; and determining whether the signal to be detected is a living physiological signal The signal to be detected is a skin elasticity signal corresponding to the object to be detected.

It is to be understood that both the foregoing general description

DRAWINGS

The above as well as other objects, features and advantages of the present invention will become more apparent from the embodiments of the invention. The drawings are intended to provide a further understanding of the embodiments of the invention, In the figures, the same reference numerals generally refer to the same parts or steps.

FIG. 1 is a flow chart illustrating a living body detecting method according to an embodiment of the present invention.

2 is a functional block diagram illustrating a living body detection system in accordance with an embodiment of the present invention.

FIG. 3 is a flow chart illustrating a first example of acquiring a signal to be detected in a living body detecting method according to an embodiment of the present invention.

FIG. 4 is a second exemplary flowchart illustrating a method of acquiring a signal to be detected in a living body detecting method according to an embodiment of the present invention.

FIG. 5 is a flowchart further illustrating living body detection based on a signal to be detected in a living body detecting method according to an embodiment of the present invention.

FIG. 6 is a schematic block diagram illustrating a living body detecting system according to an embodiment of the present invention.

detailed description

In order to make the objects, the technical solutions and the advantages of the present invention more apparent, the exemplary embodiments according to the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiment is only one of the present invention. The invention is not limited by the exemplary embodiments described herein. It is to be understood that the invention is not limited to the exemplary embodiments described herein. All other embodiments obtained by those skilled in the art without departing from the scope of the invention are intended to be included within the scope of the present invention.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings.

FIG. 1 is a flow chart illustrating a living body detecting method according to an embodiment of the present invention. As shown in FIG. 1, a living body detecting method according to an embodiment of the present invention includes the following steps.

In step S101, video data acquired via the video capture device is acquired. In one embodiment of the present invention, the video capture device is a camera capable of acquiring video data of a subject (such as a front or rear camera of a smart phone, a camera of an access control system, etc.). Acquiring video data collected via the video capture device includes, but is not limited to, receiving video data transmitted from the video capture device via a wired or wireless method after the video capture device configured by the physical location is separately configured to acquire video data. Alternatively, the video capture device may be physically co-located with other modules or components in the biometric detection system or even within the same housing, and other modules or components in the biometric detection system are received from the video capture device via the internal bus. Video data.

In one embodiment of the invention, the video data acquired via the video capture device may be a video of a continuous predetermined period of time (eg, 3 seconds). A face that is a subject of living body detection needs to be able to clearly appear in the video. In a preferred embodiment of the present invention, in the video of the predetermined period of time, it is necessary to record an image of a specific region before and after the specific object is completed according to the indication. The specific action may be, for example, pressing the skin of the two jaws with a finger, or inhaling the two jaws. Thereafter, the processing proceeds to step S102.

In step S102, an object to be detected is determined based on the video data. In one embodiment of the invention, a pre-trained face detector, such as Adaboost Cascade, may be used to obtain the location of the face in the video image in the video data. Specifically, for example, a large number of (for example, N=10000) face images are collected in advance, and a series of key points such as the corners of the face, the corners of the mouth, the nose, and the highest point of the cheekbone are manually marked in each image, using a machine learning algorithm (such as Deep learning, or regression techniques based on local features), training to obtain face detectors (such as Adaboost Cascade). Using the trained face detector, you can output the face position and key point coordinates based on the input image. Further, after obtaining the face position and the key point coordinates thereon, at least one key area in the face area is determined according to the key point. For example, the face area can be divided into a series of triangular elements, and the images of the triangular elements located in the chin, the tibia, the two sides, etc. are taken as key areas. image. Thereafter, the processing proceeds to step S103.

In step S103, a signal to be detected corresponding to the object to be detected is acquired. In one embodiment of the present invention, after the image of the key region before and after the specific motion is completed is recorded according to the indication, the image of the key region before and after the captured specific motion is taken as the signal to be detected. In another embodiment of the present invention, after recording the image of the key area before and after the specific object is completed according to the indication, the captured image of the key area after the specific action and the predetermined range around the key area after the specific action are recorded. The relevant area image is used as a signal to be detected. If the object to be detected is a living body, the signal to be detected will include a characteristic signal reflecting the skin elasticity of the living body. Hereinafter, how to process and acquire a signal to be detected will be described in further detail with reference to a flowchart. Thereafter, the processing proceeds to step S104.

In step S104, it is determined whether the signal to be detected is a living physiological signal. In one embodiment of the invention, the signal to be detected acquired in step S103 is sent to the trained classifier. If the classifier determines that the signal to be detected is a living physiological signal, it outputs 1; otherwise, it outputs 0. The process of training the classifier can be done offline. For example, an image of a front and rear frame in which a living person performs a prescribed action is collected in advance, and an attack image using a photograph, a video playback, a paper mask, and a 3D model to perform a prescribed action is collected, and the former is taken as a positive sample and the latter as a negative sample. The classifier is then trained using statistical learning methods such as deep learning and support vector machines.

According to the living body detecting method according to the embodiment of the present invention, the skin elasticity characteristic in the video image sequence of the subject is detected to perform living body detection, thereby effectively preventing photos, 3D face models, and mask attacks.

Hereinafter, a living body detecting system that executes the above-described living body detecting method will be further described with reference to FIG.

2 is a functional block diagram illustrating a living body detection system in accordance with an embodiment of the present invention. As shown in FIG. 2, the living body detecting system 20 according to the embodiment of the present invention includes a video data acquiring module 21, an object to be detected determining module 22, a signal to be detected acquiring module 23, and a living body detecting module 24. The video data acquiring module 21, the object to be detected determining module 22, the to-be-detected signal acquiring module 23, and the living body detecting module 24 may be, for example, by hardware (for example, a camera, a server, a dedicated computer or a CPU, a GPU, and various application specific integrated circuits). Etc.), software, firmware, and any feasible combination of them.

Specifically, the video data acquiring module 21 is configured to acquire video data. In one embodiment of the present invention, the video data acquisition module 21 may be a video capture device including an RGB camera capable of acquiring video data of a subject. In another embodiment of the invention, the view The frequency data acquisition module 21 may include a video capture device of a depth camera (depth camera) capable of acquiring depth information of a subject. The video data acquiring module 21 can be physically separated from the object to be detected determining module 22, the signal to be detected module 23, and the living body detecting module 24, or physically located at the same location or even inside the same casing. In a case where the video data acquiring module 21 and the object to be detected determining module 22, the to-be-detected signal acquiring module 23, and the living body detecting module 24 are physically separated from each other, the video data acquiring module 21 further performs the wired or wireless manner. The depth video data acquired by the video capture device is sent to a subsequent module. In the case where the video data acquiring module 21 and the object to be detected determining module 22, the to-be-detected signal acquiring module 23, and the living body detecting module 24 are physically located at the same position or even inside the same casing, the video data acquiring module The depth video data acquired by the video capture device is sent to the subsequent module via the internal bus. The video data may be RGB color video data or RGBD video data including depth information. Before the video data is transmitted via wired or wireless means or via the internal bus, its predetermined format can be encoded and compressed into video data packets to reduce the amount of traffic and bandwidth required for transmission.

Further, as described above, in one embodiment of the present invention, the frequency data acquired via the video data acquiring module 21 as the video capturing device may be a video of a continuous predetermined period of time (for example, 3 seconds). A face that is a subject of living body detection needs to be able to clearly appear in the video. In a preferred embodiment of the present invention, in the video of the predetermined period of time, it is necessary to record an image of a specific region before and after the specific object is completed according to the indication. The specific action may be, for example, pressing the skin of the two jaws with a finger, or inhaling the two jaws.

The object to be detected determining module 22 is configured to determine an object to be detected based on the video data collected by the video data acquiring module 21 . As described above, the object to be detected determination module 22 can use a pre-trained face detector (such as Adaboost Cascade) to acquire the position of the face in the video image in the video image. Using the trained face detector, you can output the face position and key point coordinates based on the input image. Further, after obtaining the face position and the key point coordinates thereon, at least one key area in the face area is determined according to the key point. For example, the face area may be divided into a series of triangular slices, and images of triangular pieces located in areas such as chin, tibia, and two cymbals may be used as key area images.

The to-be-detected signal acquisition module 23 is configured to acquire a to-be-detected signal corresponding to the object to be detected determined by the object to be detected determining module 22. Specifically, in an embodiment of the present invention, after the image of the key area before and after the specific action is completed according to the indication is recorded, the object is captured. The image of the key area before and after the specific action is obtained as a signal to be detected. In another embodiment of the present invention, after recording the image of the key area before and after the specific object is completed according to the indication, the captured image of the key area after the specific action and the predetermined range around the key area after the specific action are recorded. The relevant area image is used as a signal to be detected. If the object to be detected is a living body, the signal to be detected will include a characteristic signal reflecting the skin elasticity of the living body.

The biometric detection module 24 is configured to perform biometric detection on the to-be-detected signal extracted by the to-be-detected signal acquisition module 23 to determine whether the to-be-detected signal is a living physiological signal. In one embodiment of the invention, the living body detection module 24 is a trained classifier. If the classifier determines that the signal to be detected is a living physiological signal, it outputs 1; otherwise, it outputs 0. The process of training the classifier can be done offline. For example, an image of a front and rear frame in which a living person performs a prescribed action is collected in advance, and an attack image using a photograph, a video playback, a paper mask, and a 3D model to perform a prescribed action is collected, and the former is taken as a positive sample and the latter as a negative sample. The classifier is then trained using statistical learning methods such as deep learning and support vector machines.

Hereinafter, the flow of each specific step of the living body detecting method according to the embodiment of the present invention executed by each module of the living body detecting system according to the embodiment of the present invention will be described in detail with reference to the accompanying drawings.

FIG. 3 is a flow chart illustrating a first example of acquiring a signal to be detected in a living body detecting method according to an embodiment of the present invention. As shown in FIG. 3, a first example of acquiring a signal to be detected in a living body detecting method according to an embodiment of the present invention includes the following steps.

In step S301, based on the video data, the face image therein is determined as the object to be detected. As described above, in one embodiment of the present invention, a pre-trained face detector such as Adaboost Cascade is used to acquire the position of a face in a video image in video data. Thereafter, the processing proceeds to step S302.

In step S302, key points in the face image are determined. In one embodiment of the invention, the key points include, but are not limited to, the corners of the face, the corners of the mouth, the nose, the highest point of the tibia, and the like. Thereafter, the processing proceeds to step S303.

In step S303, the face image is divided into at least one key area based on the key points. In an embodiment of the present invention, based on the key points determined in step S302, such as the corners of the eyes, the corners of the mouth, the nose, the highest point of the tibia, etc., the face area will be divided into a series of triangular pieces, which will be located at the chin, 颧An image of a triangular piece of a bone, two ridges, and the like is used as a key area image. Thereafter, the processing proceeds to step S304.

In step S304, acquiring, before and after the predetermined time point corresponding to the at least one key area Pre-action area image and post-action area image. In an embodiment of the invention, the predetermined time point is a point in time at which the object to be detected performs a predetermined action. The specific action may be, for example, pressing the skin of the two jaws with a finger, or inhaling the two jaws. Thereafter, the processing proceeds to step S305.

In step S305, the pre-action area image and the post-action area image are normalized into a grayscale image having a predetermined size. Specifically, the pre-action area image and the post-action area image are normalized to a 40x40 grayscale image. Thereafter, the processing proceeds to step S306.

In step S306, the normalized pre-action area image and the normalized post-action area image are superimposed as a signal to be detected. Specifically, the grayscale images normalized to the size of 40x40 are then stacked together to obtain a 40x40x2 dual-channel image signal (tensor signal).

Through the processing of steps S301 to S306, the obtained signal to be detected will be provided to the trained convolutional neural network, and a series of designed convolutional layer, pooling layer and full-link layer will finally obtain a two-category result, and whether the output is output. Is a living judgment probability (probability value between 0-1). This is due to the difference in material between live skin and photos, video playback, paper masks and 3D models. The live skin image changes smoothly before and after the action, which is photo, video playback, paper mask and 3D model. I can't simulate it.

The method of acquiring a signal to be detected in the living body detecting method according to the embodiment of the present invention is not limited thereto. FIG. 4 is a second exemplary flowchart illustrating a method of acquiring a signal to be detected in a living body detecting method according to an embodiment of the present invention. As shown in FIG. 4, a second example of acquiring a signal to be detected in a living body detecting method according to an embodiment of the present invention includes the following steps.

Steps S401 to S404 in Fig. 4 are the same as steps S301 to S304 shown in Fig. 3, and a repetitive description thereof will be omitted herein.

After acquiring the pre-action area image and the post-action area image before and after the predetermined time point corresponding to the at least one key area in step S404, the processing proceeds to step S405.

In step S405, the post-action area image and the relevant area image of the predetermined range around the post-action area image are normalized into a grayscale image having a predetermined size as a signal to be detected. Unlike the first example shown in FIG. 3, in the second example shown in FIG. 4, the signal to be detected does not include the pre-action area image, but includes the post-action area image and the predetermined range from the post-action area image to the surrounding area. Related area image.

The signal to be detected thus obtained is also provided to the trained convolutional neural network, and finally obtains a two-category result through a series of designed convolutional layer, pooling layer and full-link layer, and whether the output is The probability of a living body (probability value between 0-1). This is due to the outward expansion of the corresponding skin area after the user performs an action of suffocating. For real human skin, from the two jaws to the lower jaw, the skin will gradually move from the bulge to the lower jaw drum, and the whole process is smooth. For general photos, video playback, etc., it is naturally impossible to achieve the aeration effect. For a simple mask made of printing paper, when it is covered on the face to make a bulging action, since the paper is hard, various edges, lines, and the like appear locally, which is also very different from real living skin.

FIG. 5 is a flowchart further illustrating living body detection based on a signal to be detected in a living body detecting method according to an embodiment of the present invention. As shown in FIG. 5, the living body detection based on the signal to be detected according to an embodiment of the present invention includes the following steps.

In step S501, video data acquired via the video capture device is acquired. As described above with reference to FIG. 1, the video data acquired via the video capture device may be a video of a continuous predetermined period of time (eg, 3 seconds). A face that is a subject of living body detection needs to be able to clearly appear in the video. In a preferred embodiment of the present invention, in the video of the predetermined period of time, it is necessary to record an image of a specific region before and after the specific object is completed according to the indication. The specific action may be, for example, pressing the skin of the two jaws with a finger, or inhaling the two jaws. Thereafter, the process proceeds to step S502.

In step S502, the detection timing is started. In a preferred embodiment of the present invention, steps S501 and S502 are performed simultaneously, that is, while the video data is initially collected via the video capture device to perform the live detection, the timer is turned on to perform the detection timing. Thereafter, the processing proceeds to step S503.

In step S503, an object to be detected is determined based on the video data. As described above, the pre-trained face detector can be used to acquire the position of the face in the video data as the object to be detected in the video image. For example, a large number of face images are pre-captured, and a series of key points such as the corners of the face, the corners of the mouth, the nose, and the highest point of the cheekbone are manually marked in each image, using machine learning algorithms (such as deep learning, or based on local features). Regression algorithm), training to obtain a face detector. Using the trained face detector, you can output the face position and key point coordinates based on the input image. Thereafter, the processing proceeds to step S504.

In step S504, a signal to be detected corresponding to the object to be detected is acquired. Specifically, after obtaining the face position and the key point coordinates thereon, at least one key area in the face area is determined according to the key point. For example, the face area may be divided into a series of triangular slices, and images of triangular pieces located in areas such as chin, tibia, and two cymbals may be used as key area images. In an embodiment of the present invention, an image of a key area before and after a specific action is completed according to the indication is recorded. After that, the image of the key area before and after the specific action captured is taken as the signal to be detected. The specific action may be, for example, pressing the skin of the two jaws with a finger, or inhaling the two jaws. More specifically, the pre-action area image and the post-action area image are normalized to a 40x40 grayscale image. The grayscale images normalized to 40x40 are superimposed on the image before and after the motion to obtain a 40x40x2 dual-channel image signal (tensor signal), which is used as the signal to be detected. Alternatively, in another embodiment of the present invention, the post-action area image and the predetermined range of related area images around the post-action area image are normalized to a grayscale image having a predetermined size as a signal to be detected. In still another embodiment of the present invention, and superimposing the grayscale images normalized to 40x40 before and after the action image, the obtained two-channel image signal, and the post-action area image and the post-action area image are scheduled around. A grayscale image having a predetermined size normalized to the relevant area image of the range is used as a signal to be detected. Thereafter, the processing proceeds to step S505.

In step S505, it is determined whether the signal to be detected matches a preset living condition. Specifically, the determination is performed in a trained classifier, such as a convolutional neural network. In order to obtain the preset living conditions, an image of the front and rear frames of the living body performing the prescribed action may be collected in advance, and an attack image using a photo, a video playback, a paper mask, and a 3D model to perform a prescribed action may be collected, and the former is taken as a positive sample. The latter is used as a negative sample, and then the classifier is trained using statistical learning methods such as deep learning and support vector machines. Due to the difference in material between living skin and photos, video playback, paper mask and 3D model, the living skin image changes smoothly before and after the movement, and the smooth expansion contraction before and after the action is photo, video playback, Paper masks and 3D models cannot be imitated. Similarly, after the user performs the airing action, the corresponding skin area is outwardly expanded. For real human skin, from the two jaws to the lower jaw, the skin will gradually move from the bulge to the lower jaw drum, and the whole process is smooth. For general photos, video playback, etc., it is naturally impossible to achieve the aeration effect. For a simple mask made of printing paper, when it is covered on the face to make a bulging action, since the paper is hard, various edges, lines, and the like appear locally, which is also very different from real living skin. Specifically, in step S505, if it is determined that the signal to be detected matches the preset living body condition, the processing proceeds to step S507.

In step S507, the classifier outputs a result of determining that the signal to be detected is a living body physiological signal.

Conversely, if a negative result is obtained in step S505, that is, it is judged that the signal to be detected does not match the preset living body condition, the processing proceeds to step S506.

In step S506, it is determined whether the detection timing reaches a preset time threshold. If a negative result is obtained in step S506, that is, the detection timing has not reached the preset time threshold, the process returns Step S503, to continue the biometric detection based on the video data.

Conversely, if a positive result is obtained in step S506, that is, the detection timing reaches the preset time threshold, the processing proceeds to step S508.

In step S508, since the preset time threshold has been reached, and the signal to be detected matching the preset living condition is still not obtained, it is determined that the signal to be detected is not the living physiological signal, and the living body detecting process is ended. In this way, it is possible to prevent trespassers from constantly trying to perform live verification with photos, video playback, paper masks, and 3D models.

FIG. 6 is a schematic block diagram illustrating a living body detecting system according to an embodiment of the present invention. As shown in FIG. 6, a living body detection system 6 according to an embodiment of the present invention includes a processor 61, a memory 62, and computer program instructions 63 stored in the memory 62.

The computer program instructions 63 may implement the functions of the respective functional modules of the living body detection system according to an embodiment of the present invention when the processor 61 is in operation, and/or may perform various steps of the living body detection method according to an embodiment of the present invention.

Specifically, when the computer program instruction 63 is executed by the processor 61, the following steps are performed: acquiring video data collected via a video data collecting device; determining an object to be detected based on the video data; acquiring corresponding to the waiting Detecting a signal to be detected of the object; and determining whether the signal to be detected is a living physiological signal, wherein the signal to be detected is a skin elasticity signal corresponding to the object to be detected.

Further, when the computer program instructions 63 are executed by the processor 61, the step of determining an object to be detected based on the video data comprises: determining a face image therein as the object to be detected based on the video data And determining at least one key area in the face image.

Furthermore, the step of determining the at least one key region in the face image when the computer program instructions 63 are executed by the processor 61 comprises determining a key point in the face image based on the video data And dividing the face image into the at least one key area based on the key point.

Further, the step of acquiring the signal to be detected corresponding to the object to be detected when the computer program instruction 63 is executed by the processor 61 comprises: acquiring a predetermined time point corresponding to the at least one key region The pre-action area image and the post-action area image before and after, the predetermined time point is a time point at which the object to be detected performs a predetermined action.

Further, the obtaining is performed when the computer program instructions 63 are run by the processor 61 The step of the signal to be detected corresponding to the object to be detected further includes: normalizing the pre-action area image and the post-action area image into a grayscale image having a predetermined size, and normalizing the The pre-action area image and the normalized post-action area image are overlapped as the to-be-detected signal.

Furthermore, the step of acquiring the signal to be detected corresponding to the object to be detected when the computer program instruction 63 is executed by the processor 61 further comprises: displaying the post-action area image and the post-action area The image of the relevant region of the predetermined range around the image is normalized to a grayscale image having the predetermined size as the signal to be detected.

Furthermore, the step of determining whether the signal to be detected is a living physiological signal when the computer program instruction 63 is executed by the processor 61 comprises: comparing the signal to be detected with a preset living condition, in the to-be-detected When the signal matches the preset living condition, the signal to be detected is determined to be a living physiological signal, wherein the preset living condition is a skin elasticity signal corresponding to the living body acquired based on the preset preset video data.

Each module in the living body detecting system according to an embodiment of the present invention may be implemented by a computer program stored in a memory stored in a processor in a living body detecting system according to an embodiment of the present invention, or may be in a computer according to an embodiment of the present invention The computer instructions stored in the computer readable storage medium of the program product are implemented by the computer when executed.

The computer readable storage medium can be any combination of one or more computer readable storage media, for example, a computer readable storage medium includes computer readable program code for randomly generating a sequence of action instructions, and another computer can The read storage medium contains computer readable program code for performing face activity recognition.

The computer readable storage medium may include, for example, a memory card of a smart phone, a storage component of a tablet, a hard disk of a personal computer, a random access memory (RAM), a read only memory (ROM), an erasable programmable read only memory. (EPROM), Portable Compact Disk Read Only Memory (CD-ROM), USB memory, or any combination of the above storage media.

The exemplary embodiments of the invention are described above in detail and are not intended to be limiting. It will be understood by those skilled in the art that various modifications, combinations or sub-combinations of the embodiments may be made without departing from the spirit and scope of the invention.

Claims

A living body detection method includes:

Obtaining video data collected via a video data collection device;

Determining an object to be detected based on the video data;

Acquiring a signal to be detected corresponding to the object to be detected;

Determining whether the signal to be detected is a living physiological signal,

The signal to be detected is a skin elasticity signal corresponding to the object to be detected.
The living body detecting method according to claim 1, wherein determining the object to be detected based on the video data comprises:

Based on the video data, a face image therein is determined as the object to be detected, and at least one key region in the face image is determined.
The living body detecting method according to claim 2, wherein the determining at least one key region in the face image comprises:

Determining a key point in the face image based on the video data, and dividing the face image into the at least one key area based on the key point.
The method of detecting a living body according to claim 2, wherein the acquiring the signal to be detected corresponding to the object to be detected comprises:

Obtaining a pre-action area image and a post-action area image corresponding to the at least one key area before and after a predetermined time point, the predetermined time point being a time point at which the object to be detected performs a predetermined action.
The method of detecting a living body according to claim 4, wherein the acquiring the signal to be detected corresponding to the object to be detected further comprises:

Normalizing the pre-action area image and the post-action area image into a grayscale image having a predetermined size, and superimposing the normalized pre-action area image and the normalized post-action area image As the signal to be detected.
The method of detecting a living body according to claim 4, wherein the acquiring the signal to be detected corresponding to the object to be detected further comprises:

The post-action area image and the relevant area image of the predetermined range around the post-action area image are normalized into the gray image having the predetermined size as the to-be-detected signal.
The living body detecting method according to claim 1, wherein the signal to be detected is determined The physiological signals for the living include:

Comparing the to-be-detected signal with a preset living condition, determining that the to-be-detected signal is a living physiological signal when the to-be-detected signal matches the preset living condition, wherein the preset living condition is based on pre-acquisition The skin elasticity signal corresponding to the living body acquired by the preset video data.
The living body detecting method according to claim 7, further comprising:

Initiating the detection timing while acquiring the video data collected via the video data collection device;

When it is not determined whether the signal to be detected is a living physiological signal when the detection timing reaches a preset time threshold, it is determined that the signal to be detected is not a living physiological signal.
A living body detection system includes:

processor;

Memory; and

Computer program instructions stored in the memory perform the following steps when the computer program instructions are executed by the processor:

Obtaining video data collected via a video data collection device;

Determining an object to be detected based on the video data;

Acquiring a signal to be detected corresponding to the object to be detected;

Determining whether the signal to be detected is a living physiological signal,

The signal to be detected is a skin elasticity signal corresponding to the object to be detected.
The living body detection system according to claim 9, wherein the step of determining an object to be detected based on the video data when the computer program instructions are executed by the processor comprises:

Based on the video data, a face image therein is determined as the object to be detected, and at least one key region in the face image is determined.
The living body detection system according to claim 10, wherein the step of determining at least one critical region in the face image when the computer program instructions are executed by the processor comprises:

Determining a key point in the face image based on the video data, and dividing the face image into the at least one key area based on the key point.
The living body detecting system according to claim 10, wherein the step of acquiring the signal to be detected corresponding to the object to be detected when the computer program instruction is executed by the processor comprises:

Obtaining a pre-action area image and a post-action area image corresponding to the at least one key area before and after a predetermined time point, the predetermined time point being a time point at which the object to be detected performs a predetermined action.
The living body detecting system according to claim 12, wherein the step of acquiring the signal to be detected corresponding to the object to be detected when the computer program instruction is executed by the processor further comprises:

Normalizing the pre-action area image and the post-action area image into a grayscale image having a predetermined size, and superimposing the normalized pre-action area image and the normalized post-action area image As the signal to be detected.
The living body detecting system according to claim 12, wherein the step of acquiring the signal to be detected corresponding to the object to be detected when the computer program instruction is executed by the processor further comprises:

The post-action area image and the relevant area image of the predetermined range around the post-action area image are normalized into the gray image having the predetermined size as the to-be-detected signal.
The living body detecting system according to claim 9, wherein the step of determining whether the signal to be detected is a living body physiological signal when the computer program instruction is executed by the processor comprises:

Comparing the to-be-detected signal with a preset living condition, determining that the to-be-detected signal is a living physiological signal when the to-be-detected signal matches the preset living condition, wherein the preset living condition is based on pre-acquisition The skin elasticity signal corresponding to the living body acquired by the preset video data.
The living body detection system of claim 15 further comprising a detection timer, wherein when said computer program instructions are executed by said processor:

And starting the detection timer to perform detection timing while acquiring the video data collected via the video data collection device;

When it is not determined whether the signal to be detected is a living physiological signal when the detection timing reaches a preset time threshold, it is determined that the signal to be detected is not a living physiological signal.
A computer program product comprising a computer readable storage medium having stored thereon computer program instructions, the computer program instructions, when executed by a computer, perform the following steps:

Obtaining video data collected via a video data collection device;

Determining an object to be detected based on the video data;

Acquiring a signal to be detected corresponding to the object to be detected;

Determining whether the signal to be detected is a living physiological signal,

The signal to be detected is a skin elasticity signal corresponding to the object to be detected.