Face recognition method and device based on digital speckle interferometry and multi-camera detection

Abstract

The embodiment of the application provides a face recognition method and device based on digital speckle interference and multi-camera detection, which can be used in the field of artificial intelligence technology, and the method comprises the following steps: obtaining an interference pattern; performing living body detection on a to-be-detected user according to the interference pattern to obtain a living body detection result, wherein the detection result is obtained through a digital speckle interference technology and / or a digital shearing speckle interference technology; if the living body detection result is true, collecting a user image of the to-be-detected user; performing face recognition on the user image through a multi-camera-based target detection algorithm to obtain a recognition result; the digital speckle interference technology, the digital shearing speckle interference technology and the multi-camera face recognition technology are combined, the living body detection and the face recognition are performed through multiple indexes, the data source is rich, the accuracy of the living body detection and the face recognition on the measured user can be improved, and therefore the security of the face security verification is improved.

Description

Technical Field

[0001] This invention relates to the field of computer technology, particularly to the field of artificial intelligence technology, and especially to a face recognition method and apparatus based on digital speckle interferometry and multi-camera detection. Background Technology

[0002] Currently, facial recognition is widely used in fields such as robot vision, aerial surveying, reverse engineering, military applications, medical imaging, and industrial inspection. Related technologies typically employ either digital speckle interferometry combined with structured light projection or digital shear speckle interferometry combined with structured light projection for liveness detection and facial recognition. However, digital speckle interferometry has low accuracy in detecting the rate of change of the measured surface, while digital shear speckle interferometry has low accuracy in detecting the amount of change of the measured surface. Facial changes are complex; the degree of change and rate of change varies across different environments and surface parts. If only the rate of change or only the amount of change of the measured surface is detected, the data source is limited, making it difficult to distinguish between liveness and model, resulting in low accuracy in liveness detection and posing a risk to facial security verification. Summary of the Invention

[0003] One objective of this invention is to provide a face recognition method based on digital speckle interferometry and multi-camera detection. This method combines digital speckle interferometry, digital sheared speckle interferometry, and multi-camera face recognition technologies. It utilizes multiple indicators for liveness detection and face recognition, providing rich data sources and improving the accuracy of liveness detection and face recognition for the tested user, thereby enhancing the security of face verification. Another objective of this invention is to provide a face recognition device based on digital speckle interferometry and multi-camera detection. A further objective of this invention is to provide a computer-readable medium. A final objective of this invention is to provide a computer device.

[0004] To achieve the above objectives, this invention discloses a face recognition method based on digital speckle interferometry and multi-camera detection, comprising:

[0005] Obtain the interference pattern;

[0006] Based on the interferogram, liveness detection is performed on the user to be detected, and the liveness detection results are obtained through digital speckle interferometry and / or digital shear speckle interferometry.

[0007] If the liveness detection result is true, collect the user image of the user to be detected;

[0008] A multi-camera-based target detection algorithm is used to perform face recognition on user images, and the recognition results are obtained.

[0009] Optionally, the interferogram may include a first interferogram and a second interferogram;

[0010] Obtaining the interferogram includes:

[0011] The first interference pattern is obtained by projecting onto the user under test and the phase-shifting mirror using a speckle illumination system;

[0012] The second interference pattern is obtained by projecting a speckle illumination system onto the user to be tested.

[0013] Optionally, based on the interferogram, liveness detection is performed on the user to be detected to obtain liveness detection results, including:

[0014] The first detection result is obtained by performing liveness detection on the first interference pattern using digital speckle interferometry.

[0015] The second interferogram is subjected to liveness detection using digital shearing speckle interferometry to obtain the second detection result;

[0016] If at least one of the first and second test results is true, the liveness test result is determined to be true.

[0017] If both the first and second test results are false, the liveness test result is determined to be false.

[0018] Optionally, the method also includes:

[0019] If the liveness detection result is false, the preset guidance action instruction will be sent to the operation terminal, and the operation terminal will return the guidance action to be detected.

[0020] If the guided action to be detected is consistent with the standard guided action corresponding to the guided action instruction, the liveness detection result will be updated to true;

[0021] If the guided action to be detected is inconsistent with the standard guided action corresponding to the guided action instruction, the liveness detection result is considered false.

[0022] Optionally, a liveness detection is performed on the first interference pattern using digital speckle interferometry to obtain a first detection result, including:

[0023] The phase difference is obtained by extracting the phase from the first interferogram using time-phase shifting technology.

[0024] Based on the phase difference and wavelength, an optical path difference pattern is generated;

[0025] Feature extraction is performed on the optical path difference pattern to generate optical path difference features;

[0026] Based on the optical path difference feature, the first detection result is generated using a pre-defined classification model.

[0027] Optionally, a second detection result is obtained by performing liveness detection on the second interference pattern using digital shearing speckle interferometry, including:

[0028] The second interference pattern is sheared by a pre-set shearing device to generate a first sheared speckle interference pattern and a second sheared speckle interference pattern.

[0029] Subtract the first shear speckle interferogram from the second shear speckle interferogram to generate the corresponding phase change pattern;

[0030] Feature extraction is performed on the phase change pattern to generate phase change features;

[0031] Based on the phase change characteristics, a second detection result is generated using a pre-defined classification model.

[0032] This invention also discloses a face recognition device based on digital speckle interferometry and multi-camera detection, comprising:

[0033] Acquisition unit, used to acquire interference patterns;

[0034] The liveness detection unit is used to perform liveness detection on the user to be detected based on the interference pattern and obtain the liveness detection result. The detection result is obtained through digital speckle interferometry and / or digital sheared speckle interferometry.

[0035] The acquisition unit is used to acquire the user image of the user to be detected if the liveness detection result is true.

[0036] The face recognition unit is used to perform face recognition on user images using a multi-camera-based target detection algorithm to obtain the recognition result.

[0037] The present invention also discloses a computer-readable medium having a computer program stored thereon, which, when executed by a processor, implements the method described above.

[0038] The present invention also discloses a computer device, including a memory and a processor, wherein the memory is used to store information including program instructions, and the processor is used to control the execution of the program instructions, wherein the processor executes the program to implement the method described above.

[0039] The present invention also discloses a computer program product, including a computer program / instruction, which, when executed by a processor, implements the method described above.

[0040] This invention acquires an interference pattern; based on the interference pattern, it performs liveness detection on the user to be detected, obtaining a liveness detection result obtained through digital speckle interferometry and / or digital sheared speckle interferometry; if the liveness detection result is true, it acquires a user image of the user to be detected; and through a multi-camera-based target detection algorithm, it performs face recognition on the user image to obtain a recognition result. By combining digital speckle interferometry, digital sheared speckle interferometry, and multi-camera face recognition technology, it performs liveness detection and face recognition through multiple indicators, with rich data sources, which can improve the accuracy of liveness detection and face recognition for the tested user, thereby improving the security of face security verification. Attached Figure Description

[0041] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0042] Figure 1 This is a schematic diagram of the structure of a face recognition system based on digital speckle interferometry and multi-camera detection, provided in an embodiment of the present invention.

[0043] Figure 2 A flowchart illustrating a face recognition method based on digital speckle interferometry and multi-camera detection, provided in an embodiment of the present invention;

[0044] Figure 3 A flowchart illustrating another face recognition method based on digital speckle interferometry and multi-camera detection provided in this embodiment of the invention;

[0045] Figure 4 This is a schematic diagram of the structure of a speckle interference module provided in an embodiment of the present invention;

[0046] Figure 5 This is a schematic diagram of the structure of a shear speckle interference module provided in an embodiment of the present invention;

[0047] Figure 6 A schematic diagram of a geometric model of digital shear speckle interference provided in an embodiment of the present invention;

[0048] Figure 7 This is a schematic diagram of the structure of a face recognition device based on digital speckle interferometry and multi-camera detection provided in an embodiment of the present invention;

[0049] Figure 8 This is a schematic diagram of the structure of a computer device provided in an embodiment of the present invention. Detailed Implementation

[0050] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0051] It should be noted that the face recognition method and device based on digital speckle interferometry and multi-camera detection disclosed in this application can be used in the field of artificial intelligence technology, or in any field other than artificial intelligence technology. The application field of the face recognition method and device based on digital speckle interferometry and multi-camera detection disclosed in this application is not limited.

[0052] To facilitate understanding of the technical solution provided in this application, the relevant content of the technical solution will be explained below. Digital speckle interferometry is a non-contact, full-field real-time measurement technology. Due to its advantages such as strong versatility, high measurement accuracy, wide frequency range, and ease of measurement, it has experienced rapid development in recent years. It is suitable for non-destructive testing in various scenarios and can perform various tests on displacement, strain, surface defects, and cracks. Its basic principle is to use laser speckle as a carrier of information on changes in the measured surface, utilizing the changes on the surface of the measured object after it is irradiated (due to the design of face detection, the laser speckle in this patent can be replaced by infrared speckle).

[0053] Digital shearing speckle interferometry utilizes the speckle pattern formed in space when coherent light is shone onto a non-smooth object surface to detect surface displacement and deformation. It features full-field measurement, a simple optical path, convenient adjustment, and low environmental requirements.

[0054] The real-time subtraction method is a method for obtaining phase difference distribution in the field of shear speckle interferometry. Shear speckle interference fringes can be directly obtained by subtraction operation. It has fast measurement speed, good fringe contrast, and strong anti-interference ability. It is often used for qualitative analysis of phase difference. However, this method can only reflect the entire number of cycles, so the resolution is relatively low. This can be compensated for by interpolation.

[0055] Time-phase shifting is one of the main methods for introducing phase in digital speckle interferometry. By driving a phase-shifting mirror with piezoelectric ceramic, a camera can acquire multiple speckle interferometric images containing a fixed phase difference, from which the phase distribution related to the deformation can be calculated. Due to its simple structure and high accuracy, time-phase shifting is widely used. Time-phase shifting techniques include three-step, four-step, and five-step phase-shifting methods.

[0056] Binocular vision technology is an important form of machine vision. It is based on the principle of parallax and uses imaging equipment to acquire two images of the object under test from different positions. By calculating the positional deviation of corresponding points in the images, the three-dimensional geometric information of the object can be obtained.

[0057] While a human face appears motionless in a quasi-static state without any expression changes, the surface actually exhibits localized dynamic changes due to the pulsation of subcutaneous arteries. These changes are difficult for the naked eye and ordinary cameras to capture, especially in areas near the neck, temples, and forehead. Digital speckle interferometry (DPI) and digital shear speckle interferometry (DSI) are both microscopic measurement techniques that calculate the amount or rate of change on the surface of the measured object by measuring phase changes. Their accuracy reaches the nanometer level, with a measurement range from tens of nanometers to hundreds of micrometers. This precision and range can capture subtle facial changes or tremors that are imperceptible to the naked eye and ordinary cameras. Furthermore, these multi-point localized changes are difficult to replicate using models, allowing this technology to distinguish whether the measured target is a real face. This invention combines DPI, DSI, and multi-camera face recognition technologies to achieve three-dimensional facial information detection and liveness detection.

[0058] Figure 1 This is a schematic diagram of a face recognition system based on digital speckle interferometry and multi-camera detection, provided in an embodiment of the present invention. Figure 1 As shown, the system includes an operation terminal 100, a control system 200, a speckle irradiation system 300, a phase shift mirror 400, a user to be tested 500, and a shearing device 600.

[0059] The operating terminal 100 is communicatively connected to the control system 200. The control system 200 is communicatively connected to the speckle irradiation system 300, the phase shift mirror 400, and the shearing device 600, respectively. The speckle irradiation system 300 is communicatively connected to the phase shift mirror 400. The user to be tested 500 is communicatively connected to the shearing device 600.

[0060] The control system 200 includes a first image acquisition unit 201, a second image acquisition unit 202, a piezoelectric control system 203, and an auxiliary light control system 204. The first image acquisition unit 201 is communicatively connected to the shearing device 600, the second image acquisition unit 202 is communicatively connected to the user to be tested 500 and the phase shift mirror 400, respectively, the piezoelectric control system 203 is communicatively connected to the phase shift mirror 400, and the auxiliary light control system 204 is communicatively connected to the speckle irradiation system 300.

[0061] The user sends a face recognition command to the control system 200 via the operation terminal 100. The auxiliary light control system 204 in the control system 200 controls the speckle illumination system 300 to project infrared laser speckle. The infrared laser speckle is split into two beams by a beam splitter. One beam is expanded and projected onto the surface of the user 500 to be detected. After diffuse reflection from the surface of the user 500, it enters the second image acquisition unit 202. The other beam illuminates the phase shift mirror 400. The piezoelectric control system 203 drives the phase shift mirror to move via piezoelectric ceramics. The phase shift mirror 400 directly reflects the other beam into the second image acquisition unit 202, so that the two beams of light form an interference pattern on the target surface of the second image acquisition unit 202. The interference pattern is filtered, unwrapped, feature extracted, and classified to obtain the first detection result. The light after diffuse reflection from the surface of the user 500 to be detected also enters the shearing device 600 for shearing processing. The sheared speckle interference pattern enters the first image acquisition unit 201 for processing to obtain the second detection result.

[0062] In the technical solution provided by this invention, an interference pattern is obtained; based on the interference pattern, liveness detection is performed on the user to be detected to obtain a liveness detection result, which is obtained through digital speckle interferometry and / or digital shear speckle interferometry; if the liveness detection result is true, a user image of the user to be detected is acquired; a multi-camera-based target detection algorithm is used to perform face recognition on the user image to obtain a recognition result. By combining digital speckle interferometry, digital shear speckle interferometry, and multi-camera face recognition technology, liveness detection and face recognition are performed through multiple indicators. The data source is rich, which can improve the accuracy of liveness detection and face recognition of the user to be detected, thereby improving the security of face security verification.

[0063] It is worth noting that, Figure 1 The face recognition scenario shown, based on digital speckle interferometry and multi-camera detection, is also applicable to... Figure 2 or Figure 3 The face recognition method based on digital speckle interferometry and multi-camera detection will not be elaborated here.

[0064] The following example uses a face recognition device based on digital speckle interferometry and multi-camera detection as the execution subject to illustrate the implementation process of the face recognition method based on digital speckle interferometry and multi-camera detection provided in this embodiment of the invention. It is understood that the execution subject of the face recognition method based on digital speckle interferometry and multi-camera detection provided in this embodiment of the invention includes, but is not limited to, a face recognition device based on digital speckle interferometry and multi-camera detection.

[0065] Figure 2 A flowchart of a face recognition method based on digital speckle interferometry and multi-camera detection provided in an embodiment of the present invention is shown below. Figure 2As shown, the method includes:

[0066] Step 101: Obtain the interference pattern.

[0067] Step 102: Based on the interference pattern, perform liveness detection on the user to be tested to obtain the liveness detection results. The detection results are obtained through digital speckle interferometry and / or digital shear speckle interferometry.

[0068] Step 103: If the liveness detection result is true, acquire the user image of the user to be detected.

[0069] Step 104: Perform face recognition on the user image using a multi-camera-based target detection algorithm to obtain the recognition result.

[0070] In the technical solution provided by this invention, an interference pattern is obtained; based on the interference pattern, liveness detection is performed on the user to be detected to obtain a liveness detection result, which is obtained through digital speckle interferometry and / or digital shear speckle interferometry; if the liveness detection result is true, a user image of the user to be detected is acquired; a multi-camera-based target detection algorithm is used to perform face recognition on the user image to obtain a recognition result. By combining digital speckle interferometry, digital shear speckle interferometry, and multi-camera face recognition technology, liveness detection and face recognition are performed through multiple indicators. The data source is rich, which can improve the accuracy of liveness detection and face recognition of the user to be detected, thereby improving the security of face security verification.

[0071] Figure 3 A flowchart of another face recognition method based on digital speckle interferometry and multi-camera detection provided in this embodiment of the invention is shown below. Figure 3 As shown, the method includes:

[0072] Step 201: Obtain the interference pattern.

[0073] In this embodiment of the invention, each step is performed by a face recognition device based on digital speckle interferometry and multi-camera detection.

[0074] In this embodiment of the invention, the interference pattern includes a first interference pattern and a second interference pattern. The first interference pattern and the second interference pattern are obtained through two projection channels, respectively.

[0075] In this embodiment of the invention, step 201 specifically includes:

[0076] Step 2011: Project the first interference pattern onto the user to be tested and the phase shift mirror using the speckle illumination system.

[0077] In this embodiment of the invention, the auxiliary light control system controls the speckle irradiation system to project infrared laser speckle. The infrared laser speckle is split into two beams by a beam splitter. One beam is expanded and projected onto the surface of the user to be tested. After diffuse reflection from the surface of the user to be tested, a first reflected light is obtained. The other beam is irradiated onto a phase shift mirror. The piezoelectric control system 203 drives the phase shift mirror to move through piezoelectric ceramics, so that the phase shift mirror reflects the other beam to obtain a second reflected light. The first reflected light and the second reflected light form an interference pattern on the target surface of the second image acquisition device, namely: the first interference pattern.

[0078] Step 2012: Project the image onto the user to be tested using a speckle illumination system to obtain a second interference pattern.

[0079] In this embodiment of the invention, the auxiliary light control system controls the speckle irradiation system to project infrared laser speckle onto the user to be detected. After diffuse reflection from the surface of the user to be detected, the first reflected light is obtained, namely, the second interference pattern.

[0080] It is worth noting that the speckle irradiation system can project not only infrared laser speckle, but also dot matrix, stripe, pattern, etc., and the embodiments of the present invention do not limit this.

[0081] Step 202: Perform liveness detection on the first interference pattern using digital speckle interferometry to obtain the first detection result.

[0082] Figure 4 This is a schematic diagram of the structure of a speckle interference module provided in an embodiment of the present invention, as shown below. Figure 4 As shown, the speckle interferometry module includes: an infrared projector 41, a beam splitter 42, a beam expander 43, an object under test 44, a phase shifter 45, an imaging lens 46, a camera 47, and a first plane mirror 48. The phase shifter 45 includes a second plane mirror 451 and a piezoelectric ceramic (PZT) element 452. This digital speckle interferometry module employs an off-surface measurement optical path. The infrared projector 41 projects light, which is split into two beams by the beam splitter 42. One beam is projected onto the surface of the object under test 44 by the beam expander 43, and after diffuse reflection from the object, it is converged by the imaging lens 46 and projected onto the target surface of the camera 47. The other beam is reflected by the first plane mirror 48 and the phase shifter 45 and then onto the target surface of the camera 47. Finally, the two beams form an interference image on the target surface of the camera 47.

[0083] In this embodiment of the invention, step 202 specifically includes:

[0084] Step 2021: Using time phase shifting technology, the phase of the first interference pattern is extracted to obtain the phase difference.

[0085] In this embodiment of the invention, when two beams of light interfere, the phase of the object under test satisfies the following relationship:

[0086]

[0087] Where λ is the wavelength of the light source, n is the refractive index of the medium, L is the geometric path difference between the two beams, β is a phase constant (phase is proportional to deformation), and φ is the phase of the object being measured. Generally, n = 1. Differentiating the above formula yields the following formula:

[0088]

[0089] Here, the geometric path difference δL is a spatial variation. To solve for the phase, we use time-phase shifting techniques. Commonly used time-phase shifting techniques include, but are not limited to, the three-step, four-step, and five-step phase shifting methods. Taking the four-step phase shifting method as an example, to solve for multiple unknowns in the phase difference, multiple related equations need to be introduced to solve for the unknown φ(x,y). Time-phase shifting techniques obtain an image containing a preset phase by introducing a known phase, and then establish a system of equations to solve for the phase and phase difference distribution. The phase is introduced by driving a phase-shifting mirror with piezoelectric ceramics. The four-step phase shifting method requires the acquisition of four images, and the phase difference introduced between every two adjacent images is π / 2. The equations are as follows.

[0090]

[0091] Where I0(x,y) is the background light intensity, u(x,y) is the modulation degree, φ(x,y) is the random phase, and I a1 (x,y), I a2 (x,y), I a3 (x,y), I a4 (x, y) represent the light intensity distribution of the speckle interference image recorded at each phase shift step. The phase of the image can be calculated using the above formula, as follows:

[0092]

[0093] The phase difference Δ can be obtained by subtracting the phases before and after deformation, and then the change δL of the measured object surface can be obtained.

[0094] It is worth noting that when the angle between the incident light and the optical axis is small, the obtained phase difference distribution mainly reflects the change away from the surface, as expressed below:

[0095]

[0096] Where Δ is the phase difference; λ is the wavelength of the light source; δw is the component of the deformation in the z-axis direction, i.e., the change from the surface; δu is the component of the deformation in the surface, which is related to the incident direction of the illumination light; θ is the angle between the optical axis and the optical axis when the optical path is built.

[0097] It is worth noting that, in this embodiment of the invention, the object to be tested refers to the user to be tested.

[0098] Step 2022: Generate an optical path difference pattern based on the phase difference and wavelength length.

[0099] Specifically, each phase difference in the phase difference distribution is multiplied by the wavelength of the corresponding light source to obtain the optical path difference pattern.

[0100] Step 2023: Extract features from the optical path difference pattern to generate optical path difference features.

[0101] Specifically, the optical path difference pattern is extracted using a feature extraction algorithm to obtain the optical path difference features.

[0102] It is worth noting that feature extraction is an existing technology, and will not be described in detail in this embodiment of the invention.

[0103] Step 2024: Generate the first detection result based on the optical path difference feature using the preset classification model.

[0104] In this embodiment of the invention, the classification model is pre-trained, and the training process of the classification model is the same as that of a conventional model, which will not be described in detail in this embodiment of the invention.

[0105] Specifically, the optical path difference feature is input into the classification model to detect whether the user to be detected is a live face, and the first detection result is output. The first detection result is either true or false. If the first detection result is true, it indicates that the user to be detected is a live face; if the first detection result is false, it indicates that the user to be detected is not a live face and may be a face mask.

[0106] Step 203: Perform liveness detection on the second interference pattern using digital shear speckle interferometry to obtain the second detection result.

[0107] The main internal structure of the sheared speckle interferometer module is a shearing device. This invention uses the classic Michelson shearing structure as an example. Other popular shearing devices are also included within the scope of this invention, such as Mach-Zehnder structures, optical wedge structures, and coated glass structures, which are not limited here. The Michelson shearing structure includes a beam splitter and two plane mirrors.

[0108] Figure 5 This is a schematic diagram of the structure of a shear speckle interference module provided in an embodiment of the present invention, as shown below. Figure 5 As shown, the shear speckle interferometer module includes: an infrared projector 51, an object under test 52, a beam splitter 53, a third plane mirror 54, a fourth plane mirror 55, an imaging lens 56, and a camera 57.

[0109] Infrared projector 51 projects light onto the object 52. The object light diffusely reflected from the surface of the object 52 passes through a shearing device to form an interfering image, which is then imaged onto the target surface of camera 47 by imaging lens 56. The third plane mirror 54 has a small angle with the optical axis; this angle is called the shearing angle, and the third plane mirror 54 is called the shearing mirror. The object light is split into two beams by beam splitter 53, and then reflected by shearing mirror (third plane mirror 54) and fourth plane mirror 55, respectively. However, the beam reflected by shearing mirror (third plane mirror 54) will be deflected by a certain angle, while the beam reflected by fourth plane mirror 55 is parallel to the optical axis. Finally, the two reflected beams converge again through beam splitter 53 and, after passing through imaging lens 56, form two misaligned interference images on the target surface of camera 57.

[0110] Step 2031: The second interference pattern is sheared using a pre-set shearing device to generate a first shear speckle interference pattern and a second shear speckle interference pattern.

[0111] In this embodiment of the invention, when two beams of light interfere, the phase of the measured surface satisfies the following relationship:

[0112]

[0113] Where λ represents the wavelength of the light source, n represents the refractive index of the medium, and L represents the geometric path difference between the two beams of light.

[0114] β represents the phase constant, and the phase is directly proportional to the deformation. Differentiating the above formula yields the following formula:

[0115]

[0116] Where Δ=δφ represents the change in phase, and δλ, δn, and δL represent the variables of light source wavelength, medium refractive index, and geometric path difference, respectively. Generally, n=1, therefore the formula can be obtained:

[0117]

[0118] Wherein, the geometric path difference δL is a spatial variation quantity, which can be converted into components in the x, y, and z directions. Therefore, the above formula can be transformed into the following:

[0119]

[0120] Where δu, δv, and δw are the displacement vector components in three directions, and A, B, and C are displacement sensitivity factors. The formula can also be written as:

[0121]

[0122] Where δx is the shear amount in the x-direction. Since the shear amount is relatively small, the difference terms δu / δx, δv / δx, and δw / δx in the formula can be expressed as differentials, thus yielding the following formula:

[0123]

[0124] In the formula, and Let u, v, and w represent the partial derivatives of the displacement components u, v, and w in the x-direction, respectively. When the y-axis is the shear direction, the corresponding formula is:

[0125]

[0126] In the formula, δy refers to the shear amount in the y-direction. and These represent the spatial gradients of the displacement components in the y-direction, thus establishing the relationship between the spatial displacement gradient and the phase change.

[0127] The relationship between phase difference and displacement spatial gradient can be seen from the above formula, where A, B and C are displacement sensitivity factors and are determined by the geometric relationship of the optical system arrangement. Figure 6 This is a schematic diagram of a geometric model of digital shear speckle interference provided in an embodiment of the present invention, as shown below. Figure 6 As shown, the laser light originates from the light source S(x) s ,y s ,z s The light emitted from point P1(x,y,z) and reflected from point P1(x,y,z) finally strikes the target surface of the camera at point C(x). c ,y c ,z c Image is formed at point P1(x,y,z) and point P2(x+δx,y,z) when the target under test deforms (as shown by the dashed line). Points P1(x,y,z) and P2(x+δx,y,z) move to points P1'(x+u,y+v,z+w) and P2'(x+δx+u+δu,y+v+δv,z+w+δw), respectively, with corresponding displacements of (u,v,w) and (u+δu,v+δv,w+δw). The relative changes in P1 and P2 are (δu,δv,δw).

[0128] When point P1 moves to point P1', the change in optical path δL1 when passing through point P1 (P1') is shown in the following formula:

[0129]

[0130] according to Figure 6 The geometric model diagram and formula derivation shown below reveal the relationship between the phase difference of the first shear speckle interferogram and the spatial displacement gradient, as illustrated in the following formula:

[0131]

[0132] Where θ is the angle between the optical axis and the z-axis of the illumination light when the optical path is constructed; Δ x Δ represents the phase difference when the shear direction is in the x-direction. y Let be the phase difference when the shearing direction is in the y-direction.

[0133] The relationship between the phase difference of the second shear speckle interferogram and the spatial displacement gradient is shown in the following formula:

[0134]

[0135] Step 2032: Subtract the first shear speckle interferogram and the second shear speckle interferogram to generate the corresponding phase change pattern.

[0136] In this embodiment of the invention, a real-time subtraction method is used to perform subtraction operations on the first shear speckle interferogram and the second shear speckle interferogram to obtain the corresponding phase change pattern.

[0137] The working principle is as follows: the distribution of interference light intensity I1(x,y) can be expressed as follows:

[0138] I1(x,y)=I0(x,y)[1+u(x,y)cosφ(x,y)]

[0139] Where I0(x,y) represents the background light intensity, u(x,y) is the modulation degree, and φ(x,y) is the random phase. The deformed interferogram can be represented as:

[0140] I1(x,y)=I0(x,y)[1+u(x,y)cos[φx,y)+Δ(x,y)]]

[0141] Where Δ(x,y) represents the phase change distribution caused by the user being detected. Subtracting the two shear speckle interferograms before and after deformation yields the following formula:

[0142]

[0143] In the above equation, sin(φ(x,y)+Δ(x,y) / 2) includes the random signal φ(x,y), which is a high-frequency term and appears as random speckle in the speckle plot. sin[Δ(x,y) / 2], which changes relatively slowly, is a low-frequency term and appears as stripes in the speckle plot. When Δ(x,y)=2nπ, n=0,1,2,3…, it appears as dark stripes. When Δ(x,y)=(2n+1)π, n=0,1,2,3…, it appears as bright stripes, |I d x,y)| represents the phase change pattern.

[0144] Step 2033: Extract features from the phase change pattern to generate phase change features.

[0145] Specifically, the phase change pattern is extracted using a feature extraction algorithm to obtain phase change features.

[0146] It is worth noting that feature extraction is an existing technology, and will not be described in detail in this embodiment of the invention.

[0147] Step 2034: Generate a second detection result based on the phase change characteristics using a preset classification model.

[0148] In this embodiment of the invention, the classification model is pre-trained, and the training process of the classification model is the same as that of a conventional model, which will not be described in detail in this embodiment of the invention.

[0149] Specifically, the phase change features are input into the classification model to detect whether the user to be detected is a live face, and a second detection result is output. The second detection result is either true or false. If the second detection result is true, it indicates that the user to be detected is a live face; if the second detection result is false, it indicates that the user to be detected is not a live face and may be a face mask.

[0150] It's worth noting that the image difference displayed by the computer is a negative value, and the absolute value of the subtraction result needs to be taken before displaying it. The real-time subtraction method determines the distribution of the phase difference based on the order of the stripes. Dark stripes change over the entire number of cycles, while bright stripes change over the entire cycle plus half a cycle. Values ​​between bright and dark stripes can be estimated using interpolation. Since the focus of this invention is on measuring facial changes, feature analysis and classification algorithms can be used to determine whether the user being tested is a live person or has a real face. Furthermore, the real-time subtraction method offers fast measurement speed, good stripe contrast, and strong anti-interference capabilities, making it suitable for rapid measurement in unstable environments.

[0151] Step 204: Determine whether at least one of the first and second detection results is true. If yes, proceed to step 205; otherwise, proceed to step 207.

[0152] Specifically, if at least one of the first and second detection results is true, the liveness detection result is determined to be true, that is, the user to be detected is a live face, and step 205 is continued; if both the first and second detection results are false, the liveness detection result is determined to be false, that is, the user to be detected is not a live face, and step 207 is continued.

[0153] Step 205: Collect user images of the user to be detected.

[0154] In this embodiment of the invention, user images of the user to be detected are acquired through a multi-camera system.

[0155] Step 206: Perform face recognition on the user image using a multi-camera-based target detection algorithm to obtain the recognition result, and the process ends.

[0156] Specifically, the user image is preprocessed to obtain a preprocessed user image; features are extracted from the preprocessed user image to obtain facial features; the facial features are input into a multi-camera-based target detection algorithm for face recognition, that is, feature matching is performed with the stored facial images to obtain the recognition result, and the process ends.

[0157] Among them, the target detection algorithm based on multi-cameras adopts the principle of binocular vision detection. The principle of binocular vision detection is an existing technology, and will not be described in detail in this embodiment of the invention.

[0158] Image preprocessing includes, but is not limited to, grayscale transformation processing, geometric correction processing, image enhancement processing, and image filtering processing, in order to ensure image quality and thus ensure the accuracy of recognition results.

[0159] Step 207: Send the preset guidance action instruction to the operation terminal, and accept the guidance action to be detected returned by the operation terminal.

[0160] In this embodiment of the invention, to reduce the false positive rate, a fault-tolerant method is further proposed. If the user to be detected is not a live face, a preset guidance action instruction is sent to the operation terminal so that the user to be detected can perform the guidance action to be detected according to the guidance action instruction; the operation terminal collects the guidance action to be detected and returns it.

[0161] It is worth noting that the guidance action instructions can be set according to actual needs, and the embodiments of the present invention do not limit this.

[0162] Step 208: Determine whether the guide action to be detected is consistent with the standard guide action corresponding to the guide action instruction. If yes, proceed to step 205; otherwise, the process ends.

[0163] In this embodiment of the invention, if the guidance action to be detected is consistent with the standard guidance action corresponding to the guidance action indication, it indicates that the user to be detected is a live face, the liveness detection result is updated to true, and step 205 is continued; if the guidance action to be detected is inconsistent with the standard guidance action corresponding to the guidance action indication, it indicates that the user to be detected is not a live face, the liveness detection result is kept as false, and the process ends.

[0164] Furthermore, if the liveness detection result is false, the liveness detection result will be sent to the operation terminal to notify the user that the face recognition failed.

[0165] This invention effectively avoids the loopholes in face recognition methods that rely on models or photos to roughly detect faces, reduces reliance on users, and shortens verification time while saving costs.

[0166] This invention applies digital speckle interferometry to the field of face recognition, broadening the application scope of this technology and providing a new solution for liveness detection in face recognition.

[0167] It is worth noting that the acquisition, storage, use, and processing of data in the technical solution of this application all comply with relevant laws and regulations. The user information in the embodiments of this application was obtained through legal and compliant means, and the acquisition, storage, use, and processing of user information have been authorized and agreed upon by the client.

[0168] The technical solution of the face recognition method based on digital speckle interferometry and multi-camera detection provided in this invention involves: acquiring an interference pattern; performing liveness detection on the user to be detected based on the interference pattern to obtain a liveness detection result, which is obtained through digital speckle interferometry and / or digital sheared speckle interferometry; if the liveness detection result is true, acquiring a user image of the user to be detected; and performing face recognition on the user image using a multi-camera target detection algorithm to obtain a recognition result. This method combines digital speckle interferometry, digital sheared speckle interferometry, and multi-camera face recognition technologies, using multiple indicators for liveness detection and face recognition. The rich data source improves the accuracy of liveness detection and face recognition for the tested user, thereby enhancing the security of face verification.

[0169] Figure 7 This is a schematic diagram of a face recognition device based on digital speckle interferometry and multi-camera detection, provided in an embodiment of the present invention. This device is used to execute the aforementioned face recognition method based on digital speckle interferometry and multi-camera detection, such as... Figure 7 As shown, the device includes: an acquisition unit 11, a liveness detection unit 12, a data acquisition unit 13, and a face recognition unit 14.

[0170] The acquisition unit 11 is used to acquire the interference pattern.

[0171] The liveness detection unit 12 is used to perform liveness detection on the user to be detected based on the interference pattern, and obtain the liveness detection result. The detection result is obtained by digital speckle interferometry and / or digital shear speckle interferometry.

[0172] The acquisition unit 13 is used to acquire the user image of the user to be detected if the liveness detection result is true.

[0173] The face recognition unit 14 is used to perform face recognition on user images using a multi-camera-based target detection algorithm to obtain recognition results.

[0174] In this embodiment of the invention, the interference pattern includes a first interference pattern and a second interference pattern; the acquisition unit 11 is specifically used to project the first interference pattern onto the user to be tested and the phase shift mirror respectively through the speckle illumination system; and to project the second interference pattern onto the user to be tested through the speckle illumination system.

[0175] In this embodiment of the invention, the liveness detection unit 12 is specifically used to perform liveness detection on the first interference pattern using digital speckle interferometry to obtain a first detection result; and to perform liveness detection on the second interference pattern using digital shearing speckle interferometry to obtain a second detection result; if at least one of the first and second detection results is true, the liveness detection result is determined to be true; if both the first and second detection results are false, the liveness detection result is determined to be false.

[0176] In this embodiment of the invention, the device further includes a sending unit 15, an updating unit 16, and a holding unit 17.

[0177] The sending unit 15 is used to send a preset guidance action instruction to the operation terminal if the liveness detection result is false, and to receive the guidance action to be detected returned by the operation terminal.

[0178] The update unit 16 is used to update the liveness detection result to true if the guidance action to be detected is consistent with the standard guidance action corresponding to the guidance action instruction.

[0179] The holding unit 17 is used to keep the liveness detection result as false if the guided action to be detected is inconsistent with the standard guided action corresponding to the guided action instruction.

[0180] In this embodiment of the invention, the liveness detection unit 12 is specifically used to extract the phase of the first interference pattern using time phase shifting technology to obtain the phase difference; generate an optical path difference pattern based on the phase difference and wavelength length; extract features from the optical path difference pattern to generate optical path difference features; and generate a first detection result based on the optical path difference features using a preset classification model.

[0181] In this embodiment of the invention, the liveness detection unit 12 is specifically used to cut the second interference pattern using a pre-set shearing device to generate a first shear speckle interference pattern and a second shear speckle interference pattern; subtract the first shear speckle interference pattern and the second shear speckle interference pattern to generate a corresponding phase change pattern; extract features from the phase change pattern to generate phase change features; and generate a second detection result based on the phase change features using a preset classification model.

[0182] In this embodiment of the invention, an interference pattern is acquired; based on the interference pattern, liveness detection is performed on the user to be detected to obtain a liveness detection result, which is obtained through digital speckle interferometry and / or digital shear speckle interferometry; if the liveness detection result is true, a user image of the user to be detected is acquired; a multi-camera-based target detection algorithm is used to perform face recognition on the user image to obtain a recognition result. By combining digital speckle interferometry, digital shear speckle interferometry, and multi-camera face recognition technology, liveness detection and face recognition are performed through multiple indicators, with rich data sources, which can improve the accuracy of liveness detection and face recognition of the user being tested, thereby improving the security of face security verification.

[0183] The systems, devices, modules, or units described in the above embodiments can be implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer device, specifically, a computer device can be, for example, a personal computer, a laptop computer, a cellular phone, a camera phone, a smartphone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or any combination of these devices.

[0184] This invention provides a computer device, including a memory and a processor. The memory is used to store information including program instructions, and the processor is used to control the execution of the program instructions. When the program instructions are loaded and executed by the processor, they implement the steps of the above-described embodiment of the face recognition method based on digital speckle interferometry and multi-camera detection. For a detailed description, please refer to the above-described embodiment of the face recognition method based on digital speckle interferometry and multi-camera detection.

[0185] The following is for reference. Figure 8 It shows a schematic diagram of the structure of a computer device 600 suitable for implementing the embodiments of this application.

[0186] like Figure 8 As shown, the computer device 600 includes a central processing unit (CPU) 601, which can perform various appropriate tasks and processes based on programs stored in read-only memory (ROM) 602 or programs loaded from storage section 608 into random access memory (RAM) 603. The RAM 603 also stores various programs and data required for the operation of the computer device 600. The CPU 601, ROM 602, and RAM 603 are interconnected via a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.

[0187] The following components are connected to I / O interface 605: an input section 606 including a keyboard, mouse, etc.; an output section 607 including a cathode ray tube (CRT), liquid crystal feedback (LCD), etc., and speakers, etc.; a storage section 608 including a hard disk, etc.; and a communication section 609 including a network interface card such as a LAN card, modem, etc. The communication section 609 performs communication processing via a network such as the Internet. A drive 610 is also connected to I / O interface 605 as needed. A removable medium 611, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on drive 610 as needed so that computer programs read from it can be installed in storage section 608 as needed.

[0188] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program tangibly embodied on a machine-readable medium, the computer program including program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication section 609, and / or installed from removable medium 611.

[0189] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0190] For ease of description, the above devices are described separately by function as various units. Of course, in implementing this application, the functions of each unit can be implemented in one or more software and / or hardware.

[0191] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0192] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0193] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0194] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0195] The acquisition, storage, use, and processing of data in this application all comply with the relevant provisions of national laws and regulations.

[0196] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0197] This application can be described in the general context of computer-executable instructions, such as program modules, that are executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform a specific task or implement a specific abstract data type. This application can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0198] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to interchangeably. Each embodiment focuses on describing the differences from other embodiments. In particular, the system embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments.

[0199] The above description is merely an embodiment of this application and is not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A face recognition method based on digital speckle interferometry and multi-camera detection, characterized in that, The method includes: Obtain the interference pattern; Based on the interference pattern, a liveness detection is performed on the user to be detected to obtain a liveness detection result, which is obtained through digital speckle interferometry and / or digital shear speckle interferometry. If the liveness detection result is true, collect the user image of the user to be detected; The user image is subjected to face recognition using a multi-camera-based target detection algorithm to obtain the recognition result; The interference pattern includes a first interference pattern and a second interference pattern; the acquisition of the interference pattern includes: projecting the first interference pattern onto the user to be tested and the phase shift mirror respectively through a speckle illumination system; and projecting the second interference pattern onto the user to be tested through the speckle illumination system; wherein the speckle illumination system projects infrared laser speckle, dot matrix, stripes or patterns; The step of performing liveness detection on the user to be detected based on the interference pattern to obtain a liveness detection result includes: performing liveness detection on the first interference pattern using the digital speckle interferometry to obtain a first detection result; performing liveness detection on the second interference pattern using the digital sheared speckle interferometry to obtain a second detection result; if at least one of the first and second detection results is true, the liveness detection result is determined to be true; if both the first and second detection results are false, the liveness detection result is determined to be false. The step of performing face recognition on the user image using a multi-camera-based target detection algorithm to obtain a recognition result includes: performing image preprocessing on the user image to obtain a preprocessed user image, wherein the image preprocessing includes grayscale transformation processing, geometric correction processing, image enhancement processing, and image filtering processing; extracting features from the preprocessed user image to obtain face features; and inputting the face features into the multi-camera-based target detection algorithm for feature matching with stored face images to obtain the recognition result.

2. The face recognition method based on digital speckle interferometry and multi-camera detection according to claim 1, characterized in that, The method further includes: If the liveness detection result is false, a preset guidance action instruction will be sent to the operation terminal, and the detection guidance action returned by the operation terminal will be accepted. If the detected guided action is consistent with the standard guided action corresponding to the guided action indication, the liveness detection result is updated to true. If the detected guided action is inconsistent with the standard guided action corresponding to the guided action indication, the liveness detection result remains false.

3. The face recognition method based on digital speckle interferometry and multi-camera detection according to claim 1, characterized in that, The step of performing liveness detection on the first interference pattern using the digital speckle interferometry to obtain a first detection result includes: The phase difference is obtained by extracting the phase from the first interference pattern using time-phase shifting technology; Based on the phase difference and wavelength, an optical path difference pattern is generated; Feature extraction is performed on the optical path difference pattern to generate optical path difference features; Based on the optical path difference feature, a first detection result is generated using a preset classification model.

4. The face recognition method based on digital speckle interferometry and multi-camera detection according to claim 1, characterized in that, The step of performing liveness detection on the second interference pattern using the digital shear speckle interferometry technique to obtain a second detection result includes: The second interference pattern is sheared using a pre-set shearing device to generate a first shear speckle interference pattern and a second shear speckle interference pattern. Subtract the first shear speckle interferogram from the second shear speckle interferogram to generate the corresponding phase change pattern; The phase change pattern is subjected to feature extraction to generate phase change features; A second detection result is generated based on the phase change characteristics using a preset classification model.

5. A face recognition device based on digital speckle interferometry and multi-camera detection, characterized in that, The apparatus is used to perform the face recognition method based on digital speckle interferometry and multi-camera detection as described in any one of claims 1 to 4; The device includes: Acquisition unit, used to acquire interference patterns; A liveness detection unit is used to perform liveness detection on the user to be detected based on the interference pattern, and obtain a liveness detection result, wherein the detection result is obtained by digital speckle interferometry and / or digital shear speckle interferometry. The acquisition unit is used to acquire the user image of the user to be detected if the liveness detection result is true. The face recognition unit is used to perform face recognition on the user image using a multi-camera-based target detection algorithm to obtain the recognition result.

6. A computer-readable medium having a computer program stored thereon, characterized in that, When executed by a processor, the program implements the face recognition method based on digital speckle interferometry and multi-camera detection as described in any one of claims 1 to 4.

7. A computer device comprising a memory and a processor, the memory for storing information including program instructions, and the processor for controlling the execution of the program instructions, characterized in that, When the program instructions are loaded and executed by the processor, they implement the face recognition method based on digital speckle interferometry and multi-camera detection as described in any one of claims 1 to 4.

8. A computer program product comprising a computer program / instructions, characterized in that, When the computer program / instructions are executed by the processor, they implement the face recognition method based on digital speckle interferometry and multi-camera detection as described in any one of claims 1 to 4.

Images