Methods, devices, equipment, and storage media for liveness detection based on retinal images

Abstract

The present invention discloses a method, device, equipment and storage medium for liveness detection based on retinal mirroring. The present invention obtains the corresponding eye retinal video of a user when watching a detection animation video; selects a target video containing the mirroring content of the detection animation video from the eye retinal video; extracts a set of frames to be detected from the target video according to a detection start frame in the target video and a positioning frame of the detection animation video; obtains a detection reference frame corresponding to each frame to be detected in the set of frames to be detected, and the detection reference frame is extracted from the detection animation video based on the positioning frame; performs liveness detection on the user according to the frames to be detected and the detection reference frame. Compared with the traditional liveness detection method for detecting user facial movements, the present invention adopts the method of recording and parsing the retinal mirroring content of the user's eye to carry out liveness detection, avoids being deceived by malicious software, and solves the problem of how to improve the accuracy of liveness detection.

Description

Technical Field

[0001] The present invention relates to the technical field of liveness detection, and in particular to a liveness detection device, equipment and storage medium based on retinal mirroring. Background Art

[0002] With the popularization of smart mobile terminals and the rise of wireless network technology, a large number of financial institutions have also shifted their personal authentication services from offline identification to online identity identification, completing online identity verification and contract signing through smart mobile terminals. Due to the popularity of this online service, the work efficiency of financial institutions has also been enhanced.

[0003] At present, online identity authentication is mainly carried out through traditional "specified action" liveness detection, such as "opening the mouth", "blinking", "shaking the head left and right", etc. Compared with the binocular camera face recognition, infrared face recognition, TOF face recognition and other solutions used in the offline authentication process, the above actions can be replaced by software-synthesized videos. Therefore, online identity authentication also has certain risks, which has buried hidden dangers for transactions and contracts of various financial institutions. Therefore, how to improve the accuracy of liveness detection has become an urgent problem to be solved.

[0004] The above contents are only used to assist in understanding the technical solution of the present invention and do not constitute an admission that the above contents are prior art. Summary of the invention

[0005] The main purpose of the present invention is to provide a method, device, equipment and storage medium for liveness detection based on retinal images, aiming to solve the technical problem that the existing technology cannot improve the accuracy of liveness detection.

[0006] To achieve the above object, the present invention provides a living body detection method based on retinal mirror image, the method comprising the following steps:

[0007] Obtain the eye retina video corresponding to the user when watching the detection animation video;

[0008] Selecting a target video containing mirror image content of the detection animation video from the eye retina video;

[0009] Extracting a frame set to be detected from the target video according to a detection start frame in the target video and a positioning frame of the detection animation video;

[0010] Acquire a detection reference frame corresponding to each frame to be detected in the set of frames to be detected, wherein the detection reference frame is extracted from the detection animation video based on the positioning frame;

[0011] Performing liveness detection on the user according to the frame to be detected and the detection reference frame.

[0012] Optionally, the step of obtaining the eye retina video corresponding to the user watching the detection animation video includes:

[0013] Get the retinal reflection video of the user when watching the video;

[0014] Performing clarity screening on the retinal reflection video to obtain an optimal retinal reflection video;

[0015] A corresponding eye retina video is obtained according to the optimal retinal reflection video.

[0016] Optionally, before the step of obtaining the eye retina video corresponding to the user watching the detection animation video, the method further includes:

[0017] Generate a detection pattern on a preset coordinate system, taking the center point of the detection pattern as a starting point;

[0018] Generate a curve according to the center point of the detection pattern;

[0019] Generate a function graph of distance and time by intercepting trajectory points on the curve;

[0020] A detection animation video including the detection pattern is generated according to the function image.

[0021] Optionally, the step of extracting a frame set to be detected from the target video according to the detection start frame in the target video and the positioning frame of the detection animation video comprises:

[0022] Obtaining the first playback moment of the detection start frame in the target video;

[0023] Obtaining a positioning frame of the detection animation video and a second playback time of the positioning frame;

[0024] A frame set to be detected is extracted from the target video according to the time difference between the first playback moment and the second playback moment.

[0025] Optionally, before the step of extracting a frame set to be detected from the target video according to the detection start frame in the target video and the positioning frame of the detection animation video, the step further includes:

[0026] Detecting whether there is a starting detection frame in the target video according to the detection pattern identifier corresponding to the detection animation video;

[0027] When there is a start detection frame, a step of extracting a frame set to be detected from the target video according to the detection start frame in the target video and the positioning frame of the detection animation video is performed.

[0028] Optionally, the step of performing liveness detection on the user according to the frame to be detected and the detection reference frame includes:

[0029] Obtaining a center point distance between a center point of the detection pattern mark in each to-be-detected frame and a center point of the detection pattern mark in a corresponding detection reference frame;

[0030] Perform liveness detection on the user according to the center point distance.

[0031] Optionally, the step of obtaining the center point distance between the center point of the detection pattern mark in each to-be-detected frame and the center point of the detection pattern mark in the corresponding detection reference frame includes:

[0032] Obtaining the position of the center point of the detection pattern identification in each frame to be detected on the preset coordinate system to obtain the first center point coordinates;

[0033] Acquire the position of the center point of the detection pattern mark in the detection reference frame corresponding to each of the frames to be detected on the preset coordinate system to obtain the second center point coordinates;

[0034] The distance between the center points is determined according to the first center point coordinates and the second center point coordinates.

[0035] In addition, to achieve the above-mentioned purpose, the present invention further proposes a living body detection device based on retinal image, and the living body detection device based on retinal image comprises:

[0036] An acquisition module, used to acquire the eye retina video corresponding to the user when watching the detection animation video;

[0037] A selection module, used for selecting a target video containing the mirror image content of the detection animation video from the eye retinal video;

[0038] An extraction module, used for extracting a set of frames to be detected from the target video according to a detection start frame in the target video and a positioning frame of the detection animation video;

[0039] A generation module, used for obtaining a detection reference frame corresponding to each frame to be detected in the set of frames to be detected, wherein the detection reference frame is extracted from the detection animation video based on the positioning frame;

[0040] A judgment module is used to perform liveness detection on the user according to the frame to be detected and the detection reference frame.

[0041] In addition, to achieve the above-mentioned purpose, the present invention also proposes a retinal image-based liveness detection device, which includes: a memory, a processor, and a retinal image-based liveness detection program stored in the memory and executable on the processor, wherein the retinal image-based liveness detection program is configured to implement the steps of the retinal image-based liveness detection method as described above.

[0042] In addition, to achieve the above-mentioned purpose, the present invention also proposes a storage medium, on which a retinal image-based liveness detection program is stored. When the retinal image-based liveness detection program is executed by a processor, the steps of the retinal image-based liveness detection method described above are implemented.

[0043] The present invention obtains the corresponding eye retinal video when the user watches the detection animation video; selects the target video containing the mirror image content of the detection animation video from the eye retinal video; extracts the frame set to be detected from the target video according to the detection start frame in the target video and the positioning frame of the detection animation video; obtains the detection reference frame corresponding to each frame to be detected in the frame set to be detected, and the detection reference frame is extracted from the detection animation video based on the positioning frame; performs liveness detection on the user according to the frame to be detected and the detection reference frame. Compared with the traditional liveness detection method for detecting the user's facial movements, the present invention adopts the method of recording and parsing the user's eye retinal mirror image content to carry out liveness detection, avoids being deceived by malicious software, and solves the problem of how to improve the accuracy of liveness detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 It is a structural schematic diagram of a liveness detection device based on retinal imaging in a hardware operating environment involved in an embodiment of the present invention;

[0045] Figure 2 It is a flow chart of the first embodiment of the living body detection method based on retinal image of the present invention;

[0046] Figure 3 A schematic diagram of a preset coordinate system in the first embodiment of the living body detection method based on retinal imaging of the present invention;

[0047] Figure 4 It is a schematic diagram of a detection pattern in the first embodiment of the living body detection method based on retinal image of the present invention;

[0048] Figure 5 A schematic diagram of a curve generated in the first embodiment of the living body detection method based on retinal imaging of the present invention;

[0049] Figure 6 This is a schematic diagram of a function image of distance and time in the first embodiment of the living body detection method based on retinal imaging of the present invention;

[0050] Figure 7 It is a schematic diagram of the trajectory distance-time coordinate system in the first embodiment of the living body detection method based on retinal image of the present invention;

[0051] Figure 8 This is a schematic diagram of an example screen of a video in the first embodiment of the living body detection method based on retinal imaging of the present invention;

[0052] Fig. 9 It is a schematic diagram of the eye retina video in the first embodiment of the living body detection method based on retinal image of the present invention;

[0053] Fig.10 It is a schematic diagram of the optimal retinal reflection video in the first embodiment of the living body detection method based on retinal mirror image of the present invention;

[0054] Fig.11 This is a schematic diagram of the target video in the first embodiment of the living body detection method based on retinal mirror image of the present invention;

[0055] Fig.12 It is a schematic diagram of positioning frame and detection frame in the first embodiment of the living body detection method based on retinal image of the present invention;

[0056] Fig.13 It is a flow chart of the second embodiment of the living body detection method based on retinal image of the present invention;

[0057] Fig.14 This is a structural block diagram of the first embodiment of the living body detection device based on retinal imaging of the present invention.

[0058] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0059] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0060] Reference Figure 1 , Figure 1 The figure is a schematic diagram of the structure of a liveness detection device based on retinal imaging in the hardware operating environment involved in the embodiment of the present invention.

[0061] like Figure 1As shown, the liveness detection device may include: a processor 1001, such as a central processing unit (CPU), a communication bus 1002, a user interface 1003, a network interface 1004, and a memory 1005. Among them, the communication bus 1002 is used to realize the connection and communication between these components. The user interface 1003 may include a display screen (Display), an input unit such as a keyboard (Keyboard), and the optional user interface 1003 may also include a standard wired interface and a wireless interface. The network interface 1004 may optionally include a standard wired interface and a wireless interface (such as a wireless fidelity (WIreless-FIdelity, WI-FI) interface). The memory 1005 may be a high-speed random access memory (Random Access Memory, RAM) memory, or a stable non-volatile memory (Non-Volatile Memory, NVM), such as a disk memory. The memory 1005 may also be a storage device independent of the aforementioned processor 1001.

[0062] Those skilled in the art will understand that Figure 1 The structure shown in the figure does not constitute a limitation on the living body detection device based on retinal imaging, and may include more or less components than shown in the figure, or combine certain components, or arrange the components differently.

[0063] like Figure 1 As shown, the memory 1005 as a storage medium may include an operating system, a data storage module, a network communication module, a user interface module, and a liveness detection program based on retinal imaging.

[0064] exist Figure 1 In the liveness detection device based on retinal mirroring shown, the network interface 1004 is mainly used for data communication with the network server; the user interface 1003 is mainly used for data interaction with the user; the processor 1001 and the memory 1005 in the liveness detection device based on retinal mirroring of the present invention can be set in the liveness detection device based on retinal mirroring, and the liveness detection device based on retinal mirroring calls the liveness detection program based on retinal mirroring stored in the memory 1005 through the processor 1001, and executes the liveness detection method based on retinal mirroring provided in an embodiment of the present invention.

[0065] The embodiment of the present invention provides a method for detecting living body based on retinal image. Figure 2 , Figure 2 It is a flowchart of the first embodiment of the living body detection method based on retinal imaging of the present invention.

[0066] In this embodiment, the living body detection method includes the following steps:

[0067] Step S10: Obtain the eye retina video corresponding to the user when watching the detection animation video.

[0068] It should be noted that the execution subject of the method of this embodiment can be a smart phone or a device with the same function. This embodiment and the following embodiments are described using a smart phone as an example.

[0069] It can be understood that the detection animation refers to an animated video generated by a liveness detection device based on retinal imaging. In order to prevent malicious software from falsifying authentication, the generation of the detection animation is random. At the same time, in order to ensure that detection can be performed based on the accuracy of the relative position of the detection frame and the positioning frame, in the detection animation, the detection pattern moves at a uniform speed along a random trajectory.

[0070] Furthermore, in order to improve the safety during liveness detection, before step S10 it also includes: generating a detection pattern on a preset coordinate system, with the center point of the detection pattern as the starting point; generating a curve based on the center point of the detection pattern; generating a function graph of distance and time by intercepting trajectory points on the curve; and generating a detection animation video containing the detection pattern based on the function image.

[0071] It should be noted that if Figure 3 As shown, the preset coordinate system is defined as a plane rectangular coordinate system, and a rectangular pattern of specified length and width is used as the plane rectangular coordinate system detection area marking pattern, where the horizontal side is the X-axis, the vertical side is the Y-axis, and the intersection of the two sides is used as the origin O, with 1 pixel as a unit length.

[0072] It is understandable that if Figure 4 As shown, the detection pattern is to draw a circle of appropriate size in the rectangle, with its center being A. This circle is used as the detection feature mark pattern and is made tangent to the X and Y axes. The position at this time is the starting position.

[0073] In the specific implementation, Figure 5 As shown, generating a curve means that within the rectangular range, on the premise that circle A does not exceed the rectangular boundary, the center of circle A is taken as the starting point, and a continuous and smooth motion trajectory is randomly drawn, and the end point of the trajectory is recorded as point B. The intermediate motion trajectory is the curve.

[0074] In the specific implementation, Figure 6As shown in the figure, the function graph of distance and time is generated by intercepting the trajectory points on the curve. Assuming that the expected process time of liveness detection is i seconds, i-1 points are randomly selected on the trajectory in a relatively uniform manner, and are recorded as t1 to t(i-1). For example, if the expected process time of liveness detection is 5 seconds, 4 points are selected on the trajectory, namely points t1 to t4. The trajectory length between each point t1 to t(i-1) and point A and the trajectory length between points AB are calculated, and the following is established: Figure 7 The trajectory distance length-time coordinate system shown is in units of 1s, and all points are connected. The lines between the points are adjusted so that the lines between the points become a coherent and smooth curve (for example, applying a cubic Bezier curve). This curve is the distance-time variation curve of circle A moving on trajectory AB with center A. The variation curve formed after the coordinate system in the previous step is adjusted. According to the motion variation curve generated in the previous step, a detection animation video with a width of m pixels, a height of n pixels, a duration of i seconds, and a frame rate of f frames / second is generated, with the rectangle in step 1 as the background, and circle A moving on trajectory AB with center A. The video example screen is shown as follows. Figure 8 shown.

[0075] Furthermore, in order to improve the accuracy of liveness detection, step S10 also includes: obtaining a retinal reflection video of the user when watching a video; screening the clarity of the retinal reflection video to obtain an optimal retinal reflection video; and obtaining a corresponding eye retinal video based on the optimal retinal reflection video.

[0076] It is understandable that if Fig. 9 As shown, the corresponding eye retinal video refers to the video of the retinal area of ​​the human eye recorded by the smart phone during the shooting process, and the eye retinal video includes the left eye retinal video and the right eye retinal video.

[0077] In a specific implementation, the clarity screening of the retinal reflection video refers to selecting the starting frame of the detection animation in the left eye retinal video, and selecting the starting frame of the detection animation in the right eye retinal video, comparing the clarity of the starting frame in the left eye retinal video with the clarity of the starting frame in the right eye retinal video, and selecting the side with the highest clarity as the optimal retinal reflection video.

[0078] In the specific implementation, Fig.10 As shown, obtaining the corresponding eye retinal video according to the optimal retinal reflection video means stretching, twisting, flipping, etc. the retinal reflection video image so that the position of the detection feature marking pattern in the adjusted detection start frame relative to the detection area marking pattern matches the position of the detection feature marking pattern in the positioning frame relative to the detection area marking pattern.

[0079] Step S20: selecting a target video containing the mirror image content of the detection animation video from the eye retinal video.

[0080] In the specific implementation, Fig.11 As shown, selecting a target video containing the mirror content of the detection animation video from the retinal video of the eye means finding the corresponding start frame and end frame in the retinal video of the eye according to the detection animation, and taking the video content between the start frame and the end frame as the target video.

[0081] Step S30: extracting a set of frames to be detected from the target video according to the detection start frame in the target video and the positioning frame of the detection animation video

[0082] It should be noted that a frame is the smallest unit of a single image in an image animation, which is equivalent to each frame on a film. A frame is a still picture, and continuous frames form an animation, such as a TV image. We usually say the number of frames, which is the number of frames of the picture transmitted in 1 second. It can also be understood as the number of times the graphics processor can refresh per second, usually expressed in fps (Frames Per Second). Each frame is a still image, and displaying frames quickly and continuously creates the illusion of movement. A high frame rate can produce smoother and more realistic animations. The more frames per second (fps), the smoother the displayed action will be.

[0083] It can be understood that the detection start frame refers to the first frame with the clearest image containing the detection area marking pattern in the image formed by the retina of the display device found by the smartphone in the target video. The image of the detection pattern in the target video in the retinal reflection video when it is at the initial position.

[0084] In a specific implementation, the frame set to be detected is extracted from the target video by recording the playback duration position of the detection start frame in the video to be detected, and extracting the frame set to be detected in the target video according to the playback duration position and the positioning frame.

[0085] In the specific implementation, let k be the accuracy coefficient of the detection (k <= f), randomly select k frames from the detection animation video within every 1 second as the detection frame, and record the playback time and the coordinates of the center point of the detection feature mark pattern relative to the detection area mark pattern. At the same time, the starting frame of the video is extracted as the positioning frame. If k = 1, the generated positioning frame and detection frame are as follows Fig.12 as shown.

[0086] Furthermore, in order to improve the accuracy of living body detection, step S30 also includes: obtaining the first playback time of the detection start frame in the target video; obtaining the positioning frame of the detection animation video, and the second playback time of the positioning frame; extracting the frame set to be detected from the target video according to the time difference between the first playback time and the second playback time.

[0087] In a specific implementation, the first playback moment refers to a moment further determined by detecting a start frame and a positioning frame in a video by determining relative positions in the same coordinate system.

[0088] It can be understood that the second playback moment of the detection animation is the moment corresponding to the positioning frame in the detection animation video detected when the smart phone performs liveness detection in real time.

[0089] In a specific implementation, extracting the frame set to be detected from the target video according to the time difference between the first playback moment and the second playback moment is performed by intercepting the target video content between the corresponding moments.

[0090] Furthermore, in order to reduce the call of resources during the living body detection process, before step S40 it also includes: detecting whether there is a starting detection frame in the target video according to the detection pattern identifier corresponding to the detection animation video; when there is a starting detection frame, executing the step of extracting a set of frames to be detected from the target video according to the detection starting frame in the target video and the positioning frame of the detection animation video.

[0091] In a specific implementation, detecting whether there is a start detection frame in the target video according to the detection pattern identifier corresponding to the detection animation video is determined by performing clarity analysis on the target video.

[0092] Step S40: obtaining a detection reference frame corresponding to each frame to be detected in the set of frames to be detected, wherein the detection reference frame is extracted from the detection animation video based on the positioning frame.

[0093] It should be noted that the extraction of the detection reference frame from the detection animation video based on the positioning frame refers to capturing an image of the screen at a frame rate f in the target video containing the detection animation.

[0094] Step S50: performing liveness detection on the user according to the frame to be detected and the detection reference frame.

[0095] In a specific implementation, the smaller the relative position of the detection animation between the frame to be detected and the detection reference frame is, the higher the matching degree between the corneal mirror image and the display image is, which also means that the liveness detection is more accurate.

[0096] In a specific implementation, performing liveness detection on the user based on the frame to be detected and the detection reference frame means matching the frame to be detected and the detection reference frame based on the degree of similarity between the frames. When the degree of similarity between the two frames reaches a preset percentage, the liveness detection can be deemed successful.

[0097] This embodiment obtains the retinal video of the eye corresponding to the user when watching the detection animation video; selects the target video containing the mirror content of the detection animation video from the retinal video of the eye; extracts the frame set to be detected from the target video according to the detection start frame in the target video and the positioning frame of the detection animation video; obtains the detection reference frame corresponding to each frame to be detected in the frame set to be detected, and the detection reference frame is extracted from the detection animation video based on the above positioning frame; performs liveness detection on the user according to the frame to be detected and the detection reference frame. Compared with the traditional liveness detection method of detecting the user's facial movements, the present invention records and analyzes the mirror content of the user's eye retina to carry out liveness detection, avoids being deceived by malicious software, and solves the problem of improving the accuracy of liveness detection.

[0098] refer to Fig.13 , Fig.13 FIG. 4 is a flow chart of a second embodiment of a living body detection method according to the present invention.

[0099] Based on the above first embodiment, in this embodiment, step S50 includes:

[0100] S501: Obtaining a center point distance between a center point of the detection pattern mark in each to-be-detected frame and a center point of the detection pattern mark in a corresponding detection reference frame;

[0101] In a specific implementation, the center point distance between the center point of the detection pattern identifier in each frame to be detected and the center point of the detection pattern identifier in the corresponding detection reference frame is obtained by obtaining the position of the center point of the detection pattern identifier in each frame to be detected on the preset coordinate system to obtain the first center point coordinate; obtaining the position of the center point of the detection pattern identifier in the detection reference frame corresponding to each frame to be detected on the preset coordinate system to obtain the second center point coordinate; and determining the distance between the center points based on the first center point coordinate and the second center point coordinate.

[0102] In a specific implementation, the matching degree is calculated by the distance between the center points of the detection image contained in the detection animation contained in the detection frame and the detection reference frame. The detection detection patterns of the detection frame and the detection reference frame are placed in the same coordinate system generated when the detection video is generated, so that the detection frame and the detection reference frame can be calculated by a smart phone.

[0103] In a specific implementation, the coordinates of the frame to be detected and the detection reference frame in the same coordinate system are determined, the position coordinates of the frame to be detected and the position coordinates of the detection reference frame are obtained, and the distance between the center points is determined by calculation.

[0104] S502: Performing liveness detection on the user according to the center point distance,

[0105] In a specific implementation, performing liveness detection on the user according to the center point distance refers to setting a preset threshold, and determining that the liveness detection is successful when the center point distance is less than the preset threshold.

[0106] This embodiment obtains the center point distance between the center point of the detection pattern identifier in each frame to be detected and the center point of the detection pattern identifier in the corresponding detection reference frame; performs liveness detection on the user according to the center point distance, thereby realizing the determination of the image matching degree between the frame to be detected and the detection frame, avoiding the traditional method of comparing the images between two images, reducing the system calling too many resources in liveness detection, and solving the problem of how to quickly perform liveness detection.

[0107] In addition, an embodiment of the present invention further proposes a storage medium, on which a liveness detection program based on retinal image is stored. When the liveness detection program based on retinal image is executed by a processor, the steps of the liveness detection method based on retinal image as described above are implemented.

[0108] Reference Fig.14 , Fig.14 This is a structural block diagram of the first embodiment of the living body detection device based on retinal imaging of the present invention.

[0109] like Fig.14 As shown, the living body detection device based on retinal image proposed in the embodiment of the present invention includes:

[0110] Acquisition module 301: Acquisition of eye retinal video corresponding to the user when watching the detection animation video;

[0111] Selection module 302: selecting a target video containing the mirror image content of the detection animation video from the eye retinal video;

[0112] Extraction module 303: extracting a frame set to be detected from the target video according to a detection start frame in the target video and a positioning frame of the detection animation video;

[0113] Generating module 304: acquiring a detection reference frame corresponding to each frame to be detected in the frame set to be detected, wherein the detection reference frame is extracted from the detection animation video based on the positioning frame;

[0114] Determination module 305: Performing liveness detection on the user according to the frame to be detected and the detection reference frame.

[0115] This embodiment obtains the retinal video of the eye corresponding to the user when watching the detection animation video; selects the target video containing the mirror content of the detection animation video from the retinal video of the eye; extracts the frame set to be detected from the target video according to the detection start frame in the target video and the positioning frame of the detection animation video; obtains the detection reference frame corresponding to each frame to be detected in the frame set to be detected, and the detection reference frame is extracted from the detection animation video based on the above positioning frame; performs liveness detection on the user according to the frame to be detected and the detection reference frame. Compared with the traditional liveness detection method of detecting the user's facial movements, the present invention records and analyzes the mirror content of the user's eye retina to carry out liveness detection, avoids being deceived by malicious software, and solves the problem of how to improve the accuracy of liveness detection.

[0116] In one embodiment, the acquisition module 301 is also used to acquire a retinal reflection video of a user when watching a video; perform clarity screening on the retinal reflection video to obtain an optimal retinal reflection video; and obtain a corresponding eye retinal video based on the optimal retinal reflection video.

[0117] In one embodiment, the acquisition module 301 is also used to generate a detection pattern on a preset coordinate system, with the center point of the detection pattern as the starting point; generate a curve based on the center point of the detection pattern; generate a function image of distance and time by intercepting trajectory points on the curve; and generate a detection animation video containing the detection pattern based on the function image.

[0118] In one embodiment, the extraction module 303 is also used to obtain the first playback time of the detection start frame in the target video; obtain the positioning frame of the detection animation video, and the second playback time of the positioning frame; and extract the frame set to be detected from the target video according to the time difference between the first playback time and the second playback time.

[0119] In one embodiment, the extraction module 303 is also used to detect whether there is a starting detection frame in the target video based on the detection pattern identifier corresponding to the detection animation video; when there is a starting detection frame, execute the step of extracting a set of frames to be detected from the target video based on the detection starting frame in the target video and the positioning frame of the detection animation video.

[0120] In one embodiment, the judgment module 305 is further used to obtain the center point distance between the center point of the detection pattern identifier in each frame to be detected and the center point of the detection pattern identifier in the corresponding detection reference frame; and perform liveness detection on the user according to the center point distance.

[0121] In one embodiment, the judgment module 305 is also used to obtain the position of the center point of the detection pattern identification in each frame to be detected on the preset coordinate system to obtain the first center point coordinates; obtain the position of the center point of the detection pattern identification in the detection reference frame corresponding to each frame to be detected on the preset coordinate system to obtain the second center point coordinates; determine the distance between the center points based on the first center point coordinates and the second center point coordinates.

[0122] Other embodiments or specific implementations of the living body detection device of the present invention can refer to the above-mentioned method embodiments, which will not be described in detail here.

[0123] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or system including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or system. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the existence of other identical elements in the process, method, article or system including the element.

[0124] The serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.

[0125] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course can also be implemented by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as a read-only memory / random access memory, a magnetic disk, or an optical disk), and includes a number of instructions for a terminal device (which can be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.) to execute the methods described in each embodiment of the present invention.

[0126] The above are only preferred embodiments of the present invention, and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A method for detecting living bodies based on retinal images, characterized in that: The method comprises: Obtain the eye retina video corresponding to the user when watching the detection animation video; The step of obtaining the eye retina video corresponding to the user watching the detection animation video includes: Get the retinal reflection video of the user when watching the video; Performing clarity screening on the retinal reflection video to obtain an optimal retinal reflection video; Obtaining a corresponding eye retina video according to the optimal retinal reflection video; Selecting a target video containing mirror image content of the detection animation video from the eye retina video; Extracting a frame set to be detected from the target video according to a detection start frame in the target video and a positioning frame of the detection animation video; The step of extracting a frame set to be detected from the target video according to the detection start frame in the target video and the positioning frame of the detection animation video comprises: Obtaining the first playback moment of the detection start frame in the target video; Obtaining a positioning frame of the detection animation video and a second playback time of the positioning frame; Extracting a frame set to be detected from the target video according to a time difference between the first playback moment and the second playback moment; Acquire a detection reference frame corresponding to each frame to be detected in the set of frames to be detected, wherein the detection reference frame is extracted from the detection animation video based on the positioning frame; Performing liveness detection on the user according to the frame to be detected and the detection reference frame; The step of performing liveness detection on the user according to the frame to be detected and the detection reference frame comprises: Obtaining the center point distance between the center point of the detection pattern mark in each to-be-detected frame and the center point of the detection pattern mark in the corresponding detection reference frame; Perform liveness detection on the user according to the center point distance.

2. The method according to claim 1, characterized in that Before the step of obtaining the eye retina video corresponding to the user watching the detection animation video, the method further includes: Generate a detection pattern mark on a preset coordinate system, and take the center point of the detection pattern mark as a starting point; Generate a curve according to the center point of the detection pattern identification; Generate a function graph of distance and time by intercepting trajectory points on the curve; A detection animation video including the detection pattern is generated according to the function image.

3. The method according to claim 1, characterized in that Before the step of extracting a frame set to be detected from the target video according to the detection start frame in the target video and the positioning frame of the detection animation video, the step further includes: Detecting whether there is a starting detection frame in the target video according to the detection pattern identifier corresponding to the detection animation video; When there is a start detection frame, a step of extracting a frame set to be detected from the target video according to the detection start frame in the target video and the positioning frame of the detection animation video is performed.

4. The method according to claim 1, characterized in that The step of obtaining the center point distance between the center point of the detection pattern mark in each to-be-detected frame and the center point of the detection pattern mark in the corresponding detection reference frame includes: Obtaining the position of the center point of the detection pattern identification in each frame to be detected on a preset coordinate system to obtain the first center point coordinates; Acquire the position of the center point of the detection pattern mark in the detection reference frame corresponding to each of the frames to be detected on the preset coordinate system to obtain the second center point coordinates; The distance between the center points is determined according to the first center point coordinates and the second center point coordinates.

5. A living body detection device based on retinal image, characterized in that: The living body detection device based on retinal image includes: Acquisition module: acquiring the eye retina video corresponding to the user when watching the detection animation video; the acquisition of the eye retina video corresponding to the user when watching the detection animation video specifically includes: acquiring the retinal reflection video of the user when watching the video; performing clarity screening on the retinal reflection video to obtain the optimal retinal reflection video; and obtaining the corresponding eye retina video according to the optimal retinal reflection video; A selection module: selecting a target video containing the mirror image content of the detection animation video from the eye retinal video; Extraction module: extracting a frame set to be detected from the target video according to the detection start frame in the target video and the positioning frame of the detection animation video; the step of extracting the frame set to be detected from the target video according to the detection start frame in the target video and the positioning frame of the detection animation video comprises: obtaining a first playback time of the detection start frame in the target video; obtaining a positioning frame of the detection animation video and a second playback time of the positioning frame; extracting a frame set to be detected from the target video according to the time difference between the first playback time and the second playback time; A generation module: obtaining a detection reference frame corresponding to each frame to be detected in the frame set to be detected, wherein the detection reference frame is extracted from the detection animation video based on the positioning frame; Judgment module: performing liveness detection on the user according to the frame to be detected and the detection reference frame; performing liveness detection on the user according to the frame to be detected and the detection reference frame, including: obtaining the center point distance between the center point of the detection pattern identifier in each frame to be detected and the center point of the detection pattern identifier in the corresponding detection reference frame; performing liveness detection on the user according to the center point distance.

6. A living body detection device based on retinal image, characterized in that: The device includes: a memory, a processor, and a retinal image-based liveness detection program stored in the memory and executable on the processor, wherein the retinal image-based liveness detection program is configured to implement the steps of the retinal image-based liveness detection method as described in any one of claims 1 to 4.

7. A storage medium, characterized in that: The storage medium stores a liveness detection program based on retinal image, and when the liveness detection program based on retinal image is executed by the processor, the steps of the liveness detection method based on retinal image as described in any one of claims 1 to 4 are implemented.

Images