WO2019200572A1 - 身份鉴权方法、身份鉴权装置、和电子设备 - Google Patents

身份鉴权方法、身份鉴权装置、和电子设备 Download PDF

Info

Publication number
WO2019200572A1
WO2019200572A1 PCT/CN2018/083615 CN2018083615W WO2019200572A1 WO 2019200572 A1 WO2019200572 A1 WO 2019200572A1 CN 2018083615 W CN2018083615 W CN 2018083615W WO 2019200572 A1 WO2019200572 A1 WO 2019200572A1
Authority
WO
WIPO (PCT)
Prior art keywords
tested
identity authentication
image information
infrared
dimensional image
Prior art date
Application number
PCT/CN2018/083615
Other languages
English (en)
French (fr)
Inventor
田浦延
Original Assignee
深圳阜时科技有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 深圳阜时科技有限公司 filed Critical 深圳阜时科技有限公司
Priority to CN201880000342.XA priority Critical patent/CN108566777A/zh
Priority to PCT/CN2018/083615 priority patent/WO2019200572A1/zh
Publication of WO2019200572A1 publication Critical patent/WO2019200572A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/16—Human faces, e.g. facial parts, sketches or expressions
    • G06V40/172—Classification, e.g. identification
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/16—Human faces, e.g. facial parts, sketches or expressions
    • G06V40/161—Detection; Localisation; Normalisation
    • G06V40/166—Detection; Localisation; Normalisation using acquisition arrangements
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/16—Human faces, e.g. facial parts, sketches or expressions
    • G06V40/168—Feature extraction; Face representation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/40—Spoof detection, e.g. liveness detection
    • G06V40/45—Detection of the body part being alive

Definitions

  • the application relates to an identity authentication method, an identity authentication device, and an electronic device.
  • fingerprint recognition technology For example, fingerprint recognition technology, iris recognition technology, and the like.
  • fingerprint recognition technology and iris recognition technology have their own limitations.
  • fingerprint recognition technology can not perform long-distance sensing, and iris recognition technology has slower sensing response speed.
  • the embodiments of the present application aim to at least solve one of the technical problems existing in the prior art. To this end, the embodiments of the present application need to provide an identity authentication method, an identity authentication device, and an electronic device.
  • the application provides an identity authentication method, including:
  • Step S1 projecting infrared flooding onto an object to be measured, and sensing a first infrared image of the object to be tested;
  • Step S2 According to the first infrared image obtained in step S1, whether the two-dimensional image information of the object to be tested matches the pre-stored two-dimensional image information;
  • Step S3 projecting infrared structured light onto the object to be tested, and sensing a second infrared image of the object to be tested;
  • Step S4 determining, according to the second infrared image obtained in step S3, whether the object to be tested is a three-dimensional object;
  • Step S5 According to the execution result of step S2 and step S4, it is confirmed whether the identity of the object to be tested is legal.
  • the identity of the object to be tested is authenticated by means of optical image sensing.
  • the 2D image information of the object to be tested can be obtained according to the first infrared image
  • the 3D image information of the object to be tested can be obtained according to the second infrared image, so that the 2D and 3D image information can be confirmed according to the 2D and 3D image information. Whether the identity of the object is legal.
  • the present application provides a novel optical sensing technology for identity authentication.
  • the optical sensing technology can be applied to sensing over long distances, and the sensing response speed is faster.
  • the longer distance is, for example, a distance within a range of 1 meter or even further.
  • the identity authentication method of the present application is to confirm whether the identity of the object to be tested is legal by comparing the two-dimensional image information and determining whether the object to be tested is a solid object, thereby saving power consumption, Reduce the time of sensing and reduce costs.
  • the application also provides an identity authentication device, including:
  • a first projector for projecting infrared flooding to an object to be tested
  • a second projector for projecting infrared structured light to the object to be tested
  • An image sensing device configured to capture infrared flooding reflected by the object to be tested, to obtain a first infrared image of the object to be tested, and to capture infrared structure light and reflection reflected by the object to be tested Obtaining a second infrared image of the object to be tested;
  • a processor configured to compare, according to the first infrared image, the two-dimensional image information of the object to be tested and the pre-stored two-dimensional image information, and obtain a comparison result; the processor is further configured to be used according to the first And determining, by the infrared image, whether the object to be tested is a three-dimensional object, and obtaining a determination result; and the processor confirms whether the identity of the object to be tested is legal according to the comparison result and the determination result.
  • the identity authentication device authenticates the identity of the object to be tested by means of optical image sensing.
  • the processor can obtain 2D image information of the object to be tested according to the first infrared image, and obtain 3D image information of the object to be tested according to the second infrared image, thereby, according to the first infrared image and the first The two infrared images can confirm whether the identity of the object to be tested is legal.
  • the present application provides a novel optical identity authentication device for identity authentication.
  • the optical identity authentication device can be applied to sensing over a long distance, and the sensing response speed is faster.
  • the longer distance is, for example, a distance within a range of 1 meter or even further.
  • the identity authentication device of the present application determines whether the identity of the object to be tested is legal by comparing the two-dimensional image information and determining whether the object to be tested is a solid object, thereby saving power consumption, Reduce the time of sensing and reduce costs.
  • the application further provides an electronic device comprising the identity authentication device according to any one of the above.
  • the electronic device is configured to correspond to whether to perform a corresponding function according to an identity authentication result of the identity authentication device.
  • the respective function includes unlocking, paying, launching any one or more of the pre-stored applications.
  • the electronic device includes any one or more of a consumer electronic product, a home electronic product, a vehicle-mounted electronic product, and a financial terminal product.
  • the electronic device of the present application includes the above-described identity authentication device, the electronic device can realize sensing of a longer distance of the object to be measured, and the sensing response speed is faster.
  • FIG. 1 is a schematic flowchart diagram of an identity authentication method according to the present application.
  • 2 is a schematic diagram showing the relationship between the radiation intensity of ambient light and the wavelength.
  • FIG. 3 is a schematic diagram of a refinement process of a first embodiment of an identity authentication method according to the present application.
  • FIG. 4 is a schematic diagram of a refinement process of a second embodiment of the identity authentication method of the present application.
  • FIG. 5 is a structural block diagram of an embodiment of an identity authentication apparatus according to the present application.
  • FIG. 6 is a schematic structural diagram of an embodiment of an electronic device of the present application.
  • first and second are used for descriptive purposes only and are not to be construed as indicating or implying a relative importance or implicitly indicating the number of technical features indicated.
  • features defining “first” or “second” may include one or more of the described features either explicitly or implicitly.
  • the meaning of "a plurality" is two or more unless specifically and specifically defined otherwise.
  • connection In the description of the present application, it should be noted that the terms “installation”, “connected”, and “connected” are to be understood broadly, and may be fixed or detachable, for example, unless otherwise specifically defined and defined. Connected, or integrally connected; may be mechanically connected, or may be electrically connected or may communicate with each other; may be directly connected or indirectly connected through an intermediate medium, may be internal communication of two elements or interaction of two elements relationship.
  • Connected, or integrally connected may be mechanically connected, or may be electrically connected or may communicate with each other; may be directly connected or indirectly connected through an intermediate medium, may be internal communication of two elements or interaction of two elements relationship.
  • the specific meanings of the above terms in the present application can be understood on a case-by-case basis.
  • step numbers S1, S2, S3, S4, and S5 referred to in the specification and the claims of the present application are only for clearly distinguishing the steps, and do not represent the order of execution of the steps.
  • FIG. 1 is a schematic flowchart diagram of an identity authentication method according to the present application.
  • the identity authentication method is applicable to, for example, but not limited to, an electronic device such as, but not limited to, a suitable type of electronic product such as a consumer electronic product, a home electronic product, a vehicle-mounted electronic product, a financial terminal product, or the like.
  • consumer electronic products such as but not limited to mobile phones, tablets, notebook computers, desktop monitors, computer integrated machines.
  • Home-based electronic products such as, but not limited to, smart door locks, televisions, refrigerators, wearable devices, and the like.
  • Vehicle-mounted electronic products such as, but not limited to, car navigation systems, car DVDs, and the like.
  • the financial terminal products are, for example, but not limited to ATM machines, terminals for self-service business, and the like.
  • the identity authentication method includes:
  • Step S1 projecting infrared flooding onto an object to be measured, and sensing a first infrared image of the object to be tested;
  • Step S2 According to the first infrared image obtained in step S1, whether the two-dimensional image information of the object to be tested matches the pre-stored two-dimensional image information;
  • Step S3 projecting infrared structured light onto the object to be tested, and sensing a second infrared image of the object to be tested;
  • Step S4 determining, according to the second infrared image obtained in step S3, whether the object to be tested is a three-dimensional object;
  • Step S5 According to the execution result of step S2 and step S4, it is confirmed whether the identity of the object to be tested is legal.
  • step S1 and step S3 are performed in time division to avoid aliasing of the first infrared image and the second infrared image.
  • step S2 and step S4 are performed sequentially.
  • the step S5 confirms that the identity of the object to be tested is illegal, that is, the identity authentication fails, the process ends, and no further execution is required. Unauthorized authentication steps. For example, when it is confirmed in step S2 that the two-dimensional image information of the object to be tested does not match the pre-stored two-dimensional image information, the identity authentication fails, and step S4 does not need to be executed again. Similarly, when the result obtained in step S4 is first executed and the result is negative, the identity authentication fails, and step S2 does not need to be executed again.
  • steps S2 and S4 can be performed simultaneously.
  • step S4 only needs to confirm whether the object to be tested is a solid object, and does not need to analyze and calculate the entire stereoscopic image information of the object to be measured, thereby reducing power consumption and saving cost.
  • the object to be tested by sensing infrared flooding and infrared structured light onto the object to be tested, different infrared images of the object to be tested are respectively obtained, thereby sensing and identifying the object to be tested.
  • the two-dimensional (2-Dimension, 2D) image information of the object to be tested can be obtained according to the first infrared image
  • the three-dimensional (3-Dimension, 3D) image information of the object to be tested can be obtained according to the second infrared image.
  • the 3D image information includes, for example, depth information.
  • the object to be tested is identified once by comparing the two-dimensional image information of the object to be tested with the pre-stored two-dimensional image information.
  • the recognition fails, the identity authentication fails, and the process ends.
  • determining whether the object to be tested is a three-dimensional object according to the second infrared image, to identify the object to be tested once, and when the recognition fails, the identity authentication fails, and the process ends.
  • step S5 confirms that the identity of the object to be tested is legal, and the identity authentication is successful.
  • identity authentication method of the above embodiment it is also feasible to additionally add certain steps.
  • a further step is added to confirm whether the object to be tested is a living body based on the first infrared image or/and the second infrared image.
  • the adding step of confirming whether the eye of the object to be tested is in a predetermined range in front of the electronic device or the like according to the first infrared image or/and the second infrared image.
  • the present application provides a novel optical sensing technology for identity authentication.
  • the optical sensing technology can be applied to sensing over a long distance, and the sensing response speed is faster.
  • the longer distance is, for example, a distance within a range of 1 meter or even further.
  • the object to be tested is, for example, a face of a human body. Accordingly, the identity authentication method of the present application is for recognizing a face. However, the application is not limited thereto, and the object to be tested may be, for example, other suitable parts of the human body, or even other suitable organisms or non-living bodies and the like.
  • the following is an example of face recognition.
  • the user Before performing face recognition, the user has registered his or her own face image template in advance and stored in, for example, a memory.
  • the face image template includes, for example, the two-dimensional image information and depth information.
  • step S1 for example, an infrared floodlight is used to project infrared floodlight to the object to be tested, and the infrared flooding reflected by the object to be tested is captured by the image sensing device, thereby obtaining a first infrared image.
  • step S2 the identification of the object to be tested is: whether the two-dimensional image information of the object to be tested matches the two-dimensional image information of the registered user's face, if the comparison is to know the object to be tested If the dimensional image information does not match the two-dimensional image information of the user's face, the identity authentication fails and the process ends.
  • step S2 If the comparison knows that the two-dimensional image information of the object to be tested matches the two-dimensional image information of the user's face, it is not yet determined that the identity of the object to be tested is legal, because: in step S2, because it is two-dimensional The judgment and recognition of the image information can also be successfully recognized if the photo of the user is utilized.
  • the comparison of the two-dimensional image information can be realized, for example, by comparing a plane picture of the object to be tested with a plane picture of the pre-stored object face.
  • the pre-stored two-dimensional image information includes facial feature information.
  • the step S2 further includes performing facial feature information extraction on the object to be measured, and comparing the extracted facial feature information and the pre-stored facial feature information to confirm whether the two-dimensional image information of the object to be tested and the pre-stored two-dimensional image information are match. Therefore, the calculation amount can be further reduced and the sensing efficiency can be improved by comparing the features with respect to the entire picture.
  • step S2 facial feature information of the object to be tested is extracted by a deep learning method.
  • the deep learning method comprises: establishing a deep convolutional neural network model, training the deep convolutional neural network model with a predetermined number of facial photos, and extracting characteristic parameters of the human face according to the trained deep convolutional neural network model.
  • the facial feature information includes, for example, any one or more of a nose, an eye, a mouth, an eyebrow, a forehead, a tibia, a chin, a face, or/and any combination of distance information therebetween.
  • the pre-stored facial feature information may also include the nose, eyes, mouth, eyebrows, forehead, cheekbones, chin, face, and even the width of the nose, the width of the chin, and the like.
  • facial features such as the nose, eyes, mouth, eyebrows, forehead, cheekbones, chin, face, width of the nose, width of the chin, etc., or/and, may also be extracted.
  • Distance information for any combination of nose, eyes, mouth, eyebrows, forehead, cheekbones, chin, etc. For example, the distance between the nose and the eye.
  • the facial feature information is not limited to the examples listed above, but may be other suitable feature information.
  • step S2 when it is confirmed that the matching coefficient of the two-dimensional image information of the object to be tested and the pre-stored two-dimensional image information is greater than or equal to a predetermined threshold, the two-dimensional image of the object to be tested can be confirmed.
  • the information matches the pre-stored two-dimensional image information.
  • it is confirmed that the matching coefficient is smaller than the predetermined threshold it can be confirmed that the two-dimensional image information of the object to be tested does not match the pre-stored two-dimensional image information.
  • the present application is not limited to the comparison manner of the two-dimensional image information referred to above, and may be other suitable comparison methods.
  • an optical component is used to project the infrared structured light to the object to be tested, and the infrared structure sensing device is used to capture the infrared structured light reflected by the object to be tested, and the second object of the object to be tested is sensed.
  • the optical component includes, for example, a light source, a collimating lens, and an optical diffraction element (DOE), wherein the light source is used to generate an infrared laser beam; the collimating lens calibrates the infrared laser beam to form approximately parallel light; and the optical diffraction element is aligned The infrared laser beam is modulated to form a corresponding speckle pattern.
  • DOE optical diffraction element
  • the speckle pattern is, for example but not limited to, one or more of a regular dot matrix, a stripe pattern, a mesh format, a speckle pattern, a coded pattern, and the like. Among them, speckle is also called random dot matrix.
  • the coded pattern consists, for example, of light of different waveforms, each waveform representing a number, the combination of which is the code.
  • the above is based on the principle of optical coding, projecting a known infrared structured light pattern onto the object to be tested.
  • the image sensing device or processor analyzes the depth information of the object to be tested according to the captured deformed infrared structured light pattern.
  • This type of infrared structured light is defined as spatially structured light.
  • the infrared structured light can be projected onto the object to be tested, for example, based on the Time of Flight (ToF) principle.
  • the image sensing device or processor calculates the depth information of the object to be measured, for example, by measuring the propagation delay time between the light pulses.
  • This type of infrared structured light is defined as time structured light.
  • the time structured light is, for example but not limited to, a combination of any one or both of a sine wave and a square wave.
  • step S4 based on the second infrared image obtained in step S3, it is determined whether the object to be tested is a solid object. For example, depth information can be obtained according to the second infrared image, thereby determining whether the object to be tested is a solid object based on the depth information.
  • the identity authentication fails and the process ends. In this case, it is possible that someone else uses a photo or video of a legitimate user for identification.
  • step S5 it is confirmed that the two-dimensional image information of the object to be tested in step S2 is successfully matched with the pre-stored two-dimensional image information, and that the object to be tested is determined to be a three-dimensional object in step S4. , identity authentication is successful.
  • the step S4 further includes: extracting the stereo feature information from the second infrared image, and determining, according to the extracted stereo feature information, whether the object to be tested is a solid object.
  • the stereo face feature information of the object to be tested is extracted by a deep learning method.
  • the deep learning method comprises: establishing a deep convolutional neural network model, training the deep convolutional neural network model with a predetermined number of facial photos, and extracting characteristic parameters of the human face according to the trained deep convolutional neural network model.
  • step S4 stereoscopic size information of any one or more of facial features such as nose, eyes, mouth, eyebrows, forehead, tibia, chin, and face may also be extracted.
  • the stereo size information is depth information.
  • the stereoscopic feature information is not limited to the examples listed above, but may be other suitable feature information.
  • stereoscopic image information may be constructed according to the second infrared image, and compared with the pre-stored stereoscopic image information. To determine whether the object to be tested is a solid object.
  • the identity authentication method of the present application can save sensing time, save power consumption, and reduce cost.
  • step S4 further includes: after confirming that the object to be tested is a solid object, further determining whether the stereoscopic information of the object to be tested conforms to a stereoscopic feature of the human body.
  • the object to be tested is a solid object
  • obtaining a stereoscopic feature such as a nose and an eye by calculating a distortion rate, and calculating a distance between the stereoscopic features, thereby determining that the object to be tested is known. Whether the stereoscopic information conforms to the three-dimensional features of the human body.
  • the identity authentication fails.
  • step S5 when it is confirmed that the two-dimensional image information of the object to be tested matches the pre-stored two-dimensional image information, and it is determined that the stereoscopic information of the object to be tested conforms to the facial three-dimensional feature of the human body, the identity The authentication was successful.
  • the industry usually projects near-infrared light with a wavelength of 850 nm to obtain an infrared image of an object to be measured.
  • the inventors of the present application have conducted a large amount of creative labor, analysis and research found that infrared luminescence with a projection wavelength of 940 nm and infrared structured light of 940 nm can be sensed, and a more accurate sensing effect can be obtained.
  • FIG. 2 is a schematic diagram showing the relationship between the radiation intensity of ambient light and the wavelength.
  • the wavelength is represented by the horizontal axis and is indicated by the letter ⁇
  • the radiation intensity is represented by the vertical axis and is indicated by the letter E.
  • Step S1 projects infrared flooding with a wavelength range of [920,960] nanometers to the object to be tested, and obtains the first infrared image of the object to be tested according to the captured infrared flooding, thereby being less susceptible to interference by ambient light, thereby improving image acquisition. Precision.
  • step S3 projects infrared structure light having a wavelength range of [920, 960] nanometers to the object to be measured, and obtaining a second infrared image of the object to be tested according to the captured infrared structured light, it can be less interfered by ambient light, thereby Improve the accuracy of image acquisition.
  • the infrared flooding projected in step S1 is further
  • the wavelength of the infrared structured light projected in step S3 is preferably 940 nm.
  • the wavelength of the infrared flood light projected in step S1 and the wavelength of the infrared structured light projected in step S3 may be deviated from 940 nm, for example, there may be (+15) nanometers or ( -15) The deviation around the nanometer. Therefore, the wavelength range of the infrared flood light projected in step S1 is, for example, [925, 955] nanometers, and the wavelength range of the infrared structured light projected in step S3 is, for example, [925, 955] nanometers. It can be seen that this wavelength range [925, 955] still falls within the wavelength range [920, 960].
  • the wavelength of the infrared flood light projected in step S1 and the wavelength of the infrared structure light projected in step S3 are any values falling within the above-mentioned wavelength range [920, 960] nanometers.
  • specific numerical values are not listed here, but any value falling within the wavelength range [920, 960] nanometers is feasible.
  • step S1 and step S3 of the identity authentication method of the present application may also be performed by using infrared flooding or infrared structured light having a wavelength of 850 nm or other suitable wavelength.
  • FIG. 3 is a schematic diagram of a refinement process of the first embodiment of the identity authentication method of the present application.
  • step S2 is performed prior to step S4.
  • the execution of step S4 is started, and in step S2, the two-dimensional image information of the object to be tested and the pre-stored two are confirmed.
  • the identity authentication fails and the process ends.
  • step S2 and step S3 are performed simultaneously. In this way, the sensing time can be further reduced and the work efficiency can be improved.
  • step S3 may be performed after step S2.
  • step S3 when it is confirmed in step S2 that the two-dimensional image information of the object to be tested matches the pre-stored two-dimensional image information, the execution of step S3 is started, and then step S4 is performed. In this way, power consumption can be reduced.
  • Step S3 can also be performed before step S2. For this case, it can be divided into two embodiments. Step S1 can be performed before step S3 or after step S3.
  • the identity authentication method of the present embodiment needs to identify the object to be tested twice, wherein the first recognition of the object to be tested is: a two-dimensional image of the object to be tested. Whether the information matches the two-dimensional image information of the registered user's face, and if the two-dimensional image information of the object to be tested does not match the two-dimensional image information of the user's face, the identity of the object to be tested is confirmed. Illegal, identity authentication failed, and the process ends.
  • step S2 If it is determined that the two-dimensional image information of the object to be tested matches the two-dimensional image information of the user's face, it is not yet determined that the identity of the object to be tested is legal, because: in step S2, because it is a two-dimensional image
  • the judgment and identification of information can also be successful if the photos of legitimate users are used.
  • step S4 by performing step S4, the above-described case of passing the photo recognition is avoided.
  • step S4 based on the second infrared image obtained in step S3, it is determined whether the object to be tested is a solid object. For example, depth information can be obtained according to the second infrared image, thereby determining whether the object to be tested is a solid object based on the depth information.
  • the identity authentication fails and the process ends. In this case, it is possible that someone else uses a photo or video of a legitimate user for identification.
  • step S5 when it is confirmed that the two-dimensional image information of the object to be tested matches the pre-stored two-dimensional image information, and that the object to be tested is a three-dimensional object, it is confirmed that the identity of the object to be tested is legal. Identity authentication was successful.
  • step S4 Since it is only necessary to determine whether the object to be tested is a solid object in step S4, without performing a large amount of analysis and calculation on the stereoscopic image information of the object to be tested, power consumption can be reduced, sensing time can be reduced, and the object can be reduced. cost.
  • step S2 when it is confirmed that the matching coefficient of the two-dimensional image information of the object to be tested and the two-dimensional image information of the registered user's face is greater than or equal to a predetermined threshold, the second object to be tested may be confirmed.
  • the dimensional image information matches the two-dimensional image information of the registered user's face.
  • the matching coefficient is smaller than the predetermined threshold, it can be confirmed that the two-dimensional image information of the object to be tested does not match the two-dimensional image information of the registered user's face.
  • the identity authentication method of the above embodiment it is also possible to additionally add certain steps.
  • a further step is added to confirm whether the object to be tested is a living body based on the first infrared image or/and the second infrared image.
  • the adding step of confirming whether the eye of the object to be tested is in a predetermined range in front of the electronic device or the like according to the first infrared image or/and the second infrared image.
  • FIG. 4 is a schematic diagram of a refinement process of a second embodiment of the identity authentication method according to the present application.
  • step S4 is performed prior to step S2.
  • step S2 is started, and when it is determined in step S4 that the object to be tested is not a solid object, the identity authentication fails, and the flow ends.
  • the sensing power consumption of the identity authentication method is lower.
  • step S4 is performed simultaneously with step S1. In this way, the sensing time can be reduced and the work efficiency can be improved.
  • step S1 can also be performed after the step S4.
  • step S4 when it is determined in step S4 that the object to be tested is a three-dimensional object, the execution of step S1 is started, and then step S2 is performed. In this way, power consumption can be reduced.
  • Step S1 can also be performed before step S4.
  • step S3 can be performed before step S1 or after step S1.
  • the identity authentication method of the present embodiment needs to identify the object to be tested twice, wherein the first recognition of the object to be tested is: determining whether the object to be tested is a three-dimensional object. If it is determined that the object to be tested is not a solid object, it is confirmed that the identity of the object to be tested is illegal, the identity authentication fails, and the process ends.
  • step S2 it is identified whether the object to be tested matches the pre-registered two-dimensional face information.
  • step S2 according to the first infrared image obtained in step S1, whether the two-dimensional image information of the object to be tested matches the pre-stored two-dimensional image information, if the two-dimensional image information of the object to be tested is confirmed with the user If the two-dimensional image information of the face does not match, the identity authentication fails.
  • step S5 when it is confirmed that the two-dimensional image information of the object to be tested matches the pre-stored two-dimensional image information, and that the object to be tested is a three-dimensional object, the identity authentication is successful.
  • the identity authentication method of the above embodiment it is also possible to additionally add certain steps.
  • a further step is added to confirm whether the object to be tested is a living body based on the first infrared image or/and the second infrared image.
  • the adding step of confirming whether the eye of the object to be tested is in a predetermined range in front of the electronic device or the like according to the first infrared image or/and the second infrared image.
  • FIG. 5 is a structural block diagram of an embodiment of an identity authentication apparatus according to the present application.
  • the identity authentication device 1 includes a first projector 10, a second projector 12, an image sensing device 14, a processor 16, and a memory 18.
  • the memory 18 is used to pre-store two-dimensional image information of one or more sample objects.
  • the first projector 10 is configured to project infrared flooding to an object to be tested.
  • the second projector 12 is configured to project infrared structured light to the object to be tested.
  • the image sensing device 14 is configured to capture infrared flooding reflected by the object to be tested, and obtain a first infrared image of the object to be tested according to the captured infrared flood light, and to capture back reflected by the object to be tested.
  • the processor 16 is configured to: according to the first infrared image, compare the two-dimensional image information of the object to be tested with the pre-stored two-dimensional image information, and obtain a comparison result, and further, according to the second infrared image, Determining whether the object to be tested is a solid object, and obtaining a judgment result, and determining whether the identity of the object to be tested is legal according to the comparison result and the judgment result.
  • the identity authentication device 1 further comprises a control circuit 17.
  • the control circuit 17 is for controlling the first projector 10, the second projector 12, and the image sensing device 14 to work together.
  • the control circuit 17 When the identity authentication is performed, the control circuit 17 is used to control the first projector 10 and the second projector 12 to work in time to avoid the first infrared image and the first sensed by the image sensing device 14. Two infrared images are aliased.
  • the control circuit 17 can control the first projector 10 to operate prior to the second projector 12, and can also control the second projector 12 to operate prior to the first projector 10.
  • the processor 16 determines whether the object to be tested is based on the second infrared image.
  • the processor 16 confirms that the identity of the object to be tested is illegal, that is, the identity authentication fails, and the process ends. The processor 16 does not need to perform other unauthenticated authentication procedures.
  • the processor 16 confirms that the two-dimensional image information of the object to be tested does not match the pre-stored two-dimensional image information, the identity authentication fails, and the processor 16 does not need to perform “according to the second infrared image, An unauthenticated authentication procedure is performed to determine whether the object to be tested is a solid object or the like.
  • the processor 16 only needs to confirm whether the object to be tested is a solid object, and does not need to analyze and calculate the entire stereoscopic image information of the object to be tested, thereby reducing power consumption and saving cost.
  • the identity authentication device 1 when the identity authentication is performed, the identity authentication device 1 obtains different infrared images of the object to be tested by projecting infrared flooding and infrared structured light to respectively obtain the different infrared images of the object to be tested. Sensing and identification of the object to be tested.
  • the processor 16 can obtain two-dimensional (2-Dimension, 2D) image information of the object to be tested according to the first infrared image, and the three-dimensional image of the object to be tested can be obtained according to the second infrared image (3-Dimension, 3D) Image information, which includes, for example, depth information.
  • the object to be tested is identified once by comparing the two-dimensional image information of the object to be tested with the pre-stored two-dimensional image information.
  • the recognition fails, the identity authentication fails, and the process ends.
  • determining whether the object to be tested is a three-dimensional object according to the second infrared image, to identify the object to be tested once, and when the recognition fails, the identity authentication fails, and the process ends.
  • the processor 16 confirms that the identity of the object to be tested is legal, and the identity authentication is successful.
  • the processor 16 may, for example, also perform an additional authentication procedure. For example, in the process of performing identity authentication, the processor 16 further performs: determining, according to the first infrared image or/and the second infrared image, whether the object to be tested is a living body; and, for example, according to the first infrared image or And a second infrared image confirming whether the eye of the object to be tested is in a predetermined range in front of the electronic device or the like.
  • the present invention is not limited to the technical solutions disclosed in the above, and any invention that is the same or similar to the technical idea of the present application should fall within the protection scope of the present application.
  • the present application provides a novel optical identity authentication device 1.
  • the optical sensing technology can be applied to sensing over a long distance, and the sensing response speed is faster.
  • the longer distance is, for example, a distance within a range of 1 meter or even further.
  • the first projector 10 is, for example but not limited to, an infrared floodlight.
  • the second projector 12 employs an optical component to project infrared structured light to the object to be tested.
  • the optical component includes, for example, a light source, a collimating lens, and an optical diffraction element (DOE), wherein the light source is used to generate an infrared laser beam; the collimating lens calibrates the infrared laser beam to form approximately parallel light; and the optical diffraction element is aligned
  • the infrared laser beam is modulated to form a corresponding speckle pattern.
  • the speckle pattern is, for example but not limited to, one or more of a regular dot matrix, a stripe pattern, a mesh format, a speckle pattern, a coded pattern, and the like.
  • speckle is also called random dot matrix.
  • the coded pattern consists, for example, of light of different waveforms, each waveform representing a number, the combination of which is the code.
  • the optical component can be constructed, for example, from other suitable optical components.
  • Image sensing device 14 or processor 16 analyzes the depth information of the object to be tested based on the captured deformed infrared structured light pattern.
  • This type of infrared structured light is defined as spatially structured light.
  • the second projector 12 projects infrared structured light onto the object to be tested.
  • the image sensing device 14 or the processor 16 calculates the depth information of the object to be measured, for example, by measuring the propagation delay time between the light pulses.
  • This type of infrared structured light is defined as time structured light.
  • the time structured light is, for example but not limited to, a combination of any one or both of a sine wave and a square wave.
  • the image sensing device 14 includes, for example, an infrared image sensor 141 for capturing infrared flooding and infrared reflected from the object to be tested. Structured light. Since the same infrared image sensor 141 is shared, the cost can be reduced.
  • the image sensing device 14 includes, for example, two infrared image sensors, the two infrared image sensors have different structures, different sensing principles, different resolutions, and the like. Wherein, an infrared image sensor is used to capture infrared flooding reflected by the object to be tested, and another infrared image sensor is used to capture infrared structured light reflected by the object to be tested.
  • the object to be tested is, for example, a face of a human body. Accordingly, the identity authentication device of the present application is for recognizing a human face. However, the application is not limited thereto, and the object to be tested may be, for example, other suitable parts of the human body, or even other suitable organisms or non-living bodies and the like.
  • the user Before the face recognition is performed, the user has registered his or her own face image template in advance and stored in, for example, the memory 18.
  • the face image template includes, for example, the two-dimensional image information and depth information.
  • the memory 18 can store, for example, one or more face image templates.
  • the processor 16 matches whether the two-dimensional image information of the object to be tested matches the two-dimensional image information of the registered user's face, if the two-dimensional image information of the object to be tested is compared with the user's face. If the two-dimensional image information of the part does not match, it is confirmed that the identity of the object to be tested is illegal, the identity authentication fails, and the process ends.
  • the processor 16 matches that the two-dimensional image information of the object to be tested matches the two-dimensional image information of the user's face, it is not yet determined that the identity of the object to be tested is legal, because the reason is: because it is a two-dimensional image. Judgment and identification of information, if others use the photos of legitimate users can also identify success.
  • the comparison of the two-dimensional image information can be realized, for example, by comparing a plane picture of the object to be tested with a plane picture of the pre-stored object face.
  • the pre-stored two-dimensional image information includes facial feature information.
  • the processor 16 further includes performing facial feature information extraction on the object to be tested, and comparing the extracted facial feature information with the pre-stored facial feature information to confirm the two-dimensional image information of the object to be tested and the pre-stored two-dimensional image. Whether the image information matches. Therefore, the amount of calculation can be further reduced and the sensing efficiency can be improved by comparing the features with respect to the entire picture.
  • the processor 16 extracts facial feature information of the object to be tested by a deep learning method.
  • the deep learning method comprises: establishing a deep convolutional neural network model, training the deep convolutional neural network model with a predetermined number of facial photos, and extracting characteristic parameters of the human face according to the trained deep convolutional neural network model.
  • the processor 16 may also extract facial features such as nose, eyes, mouth, eyebrows, forehead, cheekbones, chin, face, width of the nose, width of the chin, or/and nose.
  • Distance information for any combination of eyes, mouth, eyebrows, forehead, cheekbones, chin, etc. For example, distance information between the nose and the eye.
  • the facial feature information is not limited to the examples listed above, but may be other suitable feature information.
  • the processor 16 when the processor 16 confirms that the matching coefficient of the two-dimensional image information of the object to be tested and the two-dimensional image information of the registered user's face is greater than or equal to a predetermined threshold, the object to be tested may be confirmed.
  • the two-dimensional image information matches the two-dimensional image information of the registered user's face.
  • it can be confirmed that the two-dimensional image information of the object to be tested does not match the two-dimensional image information of the registered user's face.
  • the present application is not limited to the comparison manner of the two-dimensional image information referred to above, and may be other suitable comparison methods.
  • the processor 16 performs, for example, extracting the stereoscopic feature information from the second infrared image, and determines whether the object to be tested is a solid object based on the extracted stereo feature information.
  • the processor 16 extracts stereo face feature information of the object to be tested, for example, by a deep learning method.
  • the deep learning method comprises: establishing a deep convolutional neural network model, training the deep convolutional neural network model with a predetermined number of facial photos, and extracting characteristic parameters of the human face according to the trained deep convolutional neural network model.
  • the processor 16 may also extract stereoscopic size information of any one or more of facial features such as nose, eyes, mouth, eyebrows, forehead, tibia, chin, face, and the like.
  • the stereo size information is depth information.
  • the stereoscopic feature information is not limited to the examples listed above, but may be other suitable feature information.
  • the processor 16 can obtain depth information according to the second infrared image, for example, so that the processor 16 determines whether the object to be tested is a solid object based on the depth information.
  • the identity authentication fails and the process ends. In this case, it is possible that someone else uses a photo or video of a legitimate user for identification.
  • the processor 16 confirms that the two-dimensional image information of the object to be tested matches the pre-stored two-dimensional image information, and the object to be tested is a three-dimensional object, the identity authentication is successful.
  • the present application is not limited to the above-described manner of determining a three-dimensional object.
  • the memory 18 may be further used to pre-store stereoscopic image information of the sample object.
  • the processor 16 may also construct stereoscopic image information according to the second infrared image, and compare with the pre-stored stereo image information to determine whether the object to be tested is a solid object.
  • the identity authentication device 1 of the present application can save sensing time, save power consumption, and reduce cost.
  • the processor 16 is further configured to: after confirming that the object to be tested is a solid object, further determining whether the stereo information of the object to be tested conforms to a stereoscopic feature of the human body.
  • the processor 16 confirms that the object to be tested is a solid object
  • the stereoscopic features of the nose, the eye, and the like are obtained by calculating the distortion rate, and the distance between the stereo features is calculated, so that the Whether the stereoscopic information of the object to be tested conforms to the stereoscopic feature of the human body.
  • the processor 16 determines that the stereoscopic information of the object to be tested does not conform to the facial feature of the human body, the identity authentication fails.
  • the processor 16 confirms that the two-dimensional image information of the object to be tested matches the pre-stored two-dimensional image information, and determines that the stereoscopic information of the object to be tested conforms to the stereoscopic feature of the human body, the identity is determined. The right is successful.
  • the industry usually projects near-infrared light with a wavelength of 850 nm to obtain an infrared image of an object to be measured.
  • the inventors of the present application have conducted a large amount of creative labor, analysis and research found that infrared luminescence with a projection wavelength of 940 nm and infrared structured light of 940 nm can be sensed, and a more accurate sensing effect can be obtained.
  • near-infrared light having a wavelength range of [920,960] nanometers in ambient light is easily absorbed by the atmosphere and has a large intensity attenuation.
  • the first projector 10 projects infrared floodlights with a wavelength range of [920,960] nanometers to
  • the image to be measured and the image sensing device 14 obtain the first infrared image of the object to be tested according to the captured infrared flooding, the image of the object to be tested can be less interfered by the ambient light, thereby improving the accuracy of image acquisition.
  • the image sensing device 14 obtains the second infrared image of the object to be tested according to the captured infrared structured light. It can be less subject to interference from ambient light, thereby improving image acquisition accuracy.
  • the first projector 10 projects The wavelength of the infrared flood light is preferably 940 nm, and the wavelength of the infrared structured light projected by the second projector 12 is preferably 940 nm.
  • the wavelength of the red external light projected by the first projector 10 and the wavelength of the infrared structured light projected by the second projector 12 may deviate from 940 nm, for example, There will be a deviation of (+15) nanometers or (-15) nanometers. Therefore, the wavelength range of the infrared flood light projected by the first projector 10 is, for example, [925, 955] nanometers, and the wavelength range of the infrared structured light projected by the second projector 12 is, for example, [925, 955] nanometers. It can be seen that this wavelength range [925, 955] still falls within the wavelength range [920, 960].
  • the wavelength of the infrared floodlight projected by the first projector 10 and the wavelength of the infrared structured light projected by the second projector 12 are any values falling within the wavelength range [920, 960] nanometers.
  • specific numerical values are not listed here, but any value falling within the wavelength range [920, 960] nanometers is feasible.
  • first projector 10 and the second projector 12 can also respectively project infrared flooding and infrared structured light having a wavelength of 850 nm or other suitable wavelengths.
  • the control circuit 17 controls the first projector 10 and the second projector 12 to operate in a time-sharing manner.
  • the processor 16 first compares the two-dimensional image information of the object to be tested with the pre-stored two-dimensional image information according to the first infrared image, and confirms the object to be tested.
  • the processor 16 confirms that the identity of the object to be tested is illegal, that is, the identity authentication fails, and the process ends; when the two-dimensional image information of the object to be tested is confirmed
  • the processor 16 determines whether the object to be tested is a solid object according to the second infrared image; when the processor 16 determines that the object to be tested is not a solid object, Identity authentication failed and the process ended.
  • the processor 16 determines that the object to be tested is a solid object, the identity authentication succeeds.
  • the control circuit 17 controls The second projector 12 projects infrared structure light to the object to be tested, and the image sensing device 14 senses a second infrared image of the object to be tested. In this way, the sensing time can be further reduced and the work efficiency can be improved.
  • the control circuit 17 controls the second projector 12 to project the infrared structure light to the standby The object is sensed, and the image sensing device 14 senses a second infrared image of the object to be tested. Then, the processor 16 determines, according to the second infrared image, whether the object to be tested is a solid object. In this way, power consumption can be reduced.
  • control circuit 17 controls the first projector 10 and the second projector 12 to work sequentially, and the image sensing device 14 sequentially senses the first infrared image and the second infrared image, and then The processor 16 further compares the two-dimensional image information of the object to be tested with the pre-stored two-dimensional image information according to the first infrared image, and confirms the two-dimensional image information of the object to be tested and the pre-stored two-dimensional image. When the information is used, it is determined according to the second infrared image whether the object to be tested is a three-dimensional object.
  • control circuit 17 can also control the second projector 12 to work with the first projector 10, and the image sensing device 14 sequentially senses the second infrared image and the first infrared image, and then the processor 16 according to the first infrared image, whether the two-dimensional image information of the object to be tested matches the pre-stored two-dimensional image information, and when confirming the two-dimensional image information of the object to be tested and the pre-stored two-dimensional image information And determining, according to the second infrared image, whether the object to be tested is a three-dimensional object.
  • the identity authentication device 1 of the present embodiment needs to identify the object to be tested twice, wherein the first recognition of the object to be tested is: the processor 16 is compared to the object to be tested. Whether the two-dimensional image information of the object matches the two-dimensional image information of the registered user's face, and if the comparison knows that the two-dimensional image information of the object to be tested does not match the two-dimensional image information of the registered user's face, the identity Authentication failed.
  • the processor 16 determines that the two-dimensional image information of the object to be tested matches the two-dimensional image information of the registered user's face, it is not yet determined that the identity of the object to be tested is legal. The reason is: because it is a two-dimensional image. The judgment and identification of information can also be successful if the photos of legitimate users are used.
  • the processor 16 avoids the above-mentioned use of photo recognition by re-determining whether the object to be tested is a solid object.
  • the depth information can be obtained based on the second infrared image, it is determined whether the object to be tested is a solid object based on the depth information. If it is determined that the object to be tested is not a solid object, the identity authentication fails and the process ends. In this case, it is possible that someone else uses a photo or video of a legitimate user for identification.
  • the processor 16 confirms that the two-dimensional image information of the object to be tested matches the pre-stored two-dimensional image information, and confirms that the object to be tested is a three-dimensional object, the identity authentication is successful.
  • the identity authentication device 1 of the present application is a comparison of two-dimensional image information and a determination of a three-dimensional object, the overall three-dimensional stereoscopic image information of the object to be tested is calculated, and then the object to be tested is subjected to identity determination.
  • the identity authentication device 1 of the present application has less power consumption, faster sensing response speed, and lower manufacturing cost.
  • the processor 16 performs the comparison of the two-dimensional image information first, and then determines whether to perform the determination of the solid object based on the comparison result.
  • the processor 16 may also perform the determination of the solid object first, and then determine whether to perform the comparison of the two-dimensional image information according to the determination result.
  • the processor 16 first determines whether the object to be tested is a solid object according to the second infrared image. When it is determined that the object to be tested is not a solid object, the processor 16 determines that the identity of the object to be tested is illegal. That is, the identity authentication fails, and the process ends; when the processor 16 determines that the object to be tested is a solid object, the processor 16 compares the two-dimensional object to the object according to the first infrared image.
  • the identity authentication fails, and the process fails; if the object to be tested is confirmed If the dimensional image information matches the pre-stored two-dimensional image information, the identity authentication is successful.
  • the processor 16 Compared with the comparison of the two-dimensional image information, the processor 16 performs less power consumption for the determination of the solid object, and therefore, when the processor 16 first performs the judgment on the solid object, once in this sense The measurement phase is passed, so there is no need to compare the subsequent two-dimensional image information, thereby reducing power consumption.
  • the processor 16 determines whether the object to be tested is a solid object according to the second infrared image: the control circuit 17 controls the first projector 10 to be turned on, and the image sensing The device 14 senses a first infrared image of the object to be tested. In this way, the sensing time can be further reduced and the work efficiency can be improved.
  • the control circuit 17 controls the first projector 10 to be turned on, and the image sensing device 14 senses the object to be tested. The first infrared image. Then, the processor 16 compares the two-dimensional image information of the object to be tested with the pre-stored two-dimensional image information according to the first infrared image. In this way, power consumption can be reduced.
  • control circuit 17 controls the first projector 10 and the second projector 12 to work sequentially, and the image sensing device 14 sequentially senses the first infrared image and the second infrared image, and then The processor 16 starts to perform "determining whether the object to be tested is a solid object according to the second infrared image", and when determining that the object to be tested is a solid object, then performing "based on the first infrared image, Whether the two-dimensional image information of the object to be tested matches the pre-stored two-dimensional image information.
  • control circuit 17 can also control the second projector 12 to work with the first projector 10, and the image sensing device 14 sequentially senses the second infrared image and the first infrared image, and then the processor 16 re-execution "determining whether the object to be tested is a three-dimensional object according to the second infrared image", and when determining that the object to be tested is a three-dimensional object, performing "according to the first infrared image, Whether the two-dimensional image information of the object to be tested matches the pre-stored two-dimensional image information.
  • FIG. 6 is a schematic structural diagram of an embodiment of an electronic device according to the present application.
  • the electronic device 100 is, for example but not limited to, a suitable type of electronic product such as a consumer electronic product, a home-based electronic product, a vehicle-mounted electronic product, a financial terminal product, or the like.
  • consumer electronic products such as but not limited to mobile phones, tablets, notebook computers, desktop monitors, computer integrated machines.
  • Home-based electronic products such as, but not limited to, smart door locks, televisions, refrigerators, wearable devices, and the like.
  • Vehicle-mounted electronic products such as, but not limited to, car navigation systems, car DVDs, and the like.
  • the financial terminal products are, for example, but not limited to ATM machines, terminals for self-service business, and the like.
  • the electronic device 100 includes the above-described identity authentication device 1.
  • the electronic device 100 corresponds to whether the corresponding function is executed according to the identity authentication result of the identity authentication device 1.
  • the respective functions are, for example but not limited to, any one or more of an application including unlocking, paying, and starting a pre-stored application.
  • an electronic device will be described as an example of a mobile phone.
  • the mobile phone is, for example, a full-screen mobile phone, and the identification device 1 is provided, for example, at the front top of the mobile phone.
  • the mobile phone is not limited to a full screen mobile phone.
  • the screen for lifting up the mobile phone or touching the mobile phone can function to wake up the identity authentication device 1.
  • the identity authentication device 1 is woken up and recognizes that the user in front of the mobile phone is a legitimate user, the screen is unlocked.
  • the electronic device 100 applies the identity authentication device 1, the electronic device 1 can realize sensing of a longer distance of the object to be measured, and the sensing response speed is faster.

Abstract

本申请公开了一种身份鉴权方法、身份鉴权装置、以及电子设备。该身份鉴权方法包括：步骤S1：投射红外泛光至一待测物体上，感测该待测物体的第一红外图像；步骤S2：根据步骤S1获得的第一红外图像，比对该待测物体的二维图像信息与预存的二维图像信息是否匹配；步骤S3：投射红外结构光至该待测物体上，感测该待测物体的第二红外图像；步骤S4：根据步骤S3获得的第二红外图像，判断该待测物体是否为立体物体；步骤S5：根据步骤S2与步骤S4的执行结果，确认所述待测物体的身份是否合法。该身份鉴权装置运行该身份鉴权方法。该电子设备包括该身份鉴权装置。

Description

身份鉴权方法、身份鉴权装置、和电子设备 技术领域
本申请涉及一种身份鉴权方法、身份鉴权装置、和电子设备。
背景技术
随着科技的发展，越来越多的场合开始采用各种传感技术对物体进行识别。例如，指纹识别技术、虹膜识别技术等。然，指纹识别技术与虹膜识别技术等都有其各自的局限性，例如，指纹识别技术并不能进行较远距离的感测，虹膜识别技术的感测响应速度较慢等。
因此，有必要提供一种新型的传感技术，用于身份鉴权。
发明内容
本申请实施方式旨在至少解决现有技术中存在的技术问题之一。为此，本申请实施方式需要提供一种身份鉴权方法、身份鉴权装置、和电子设备。
首先，本申请提供一种身份鉴权方法，包括：
步骤S1：投射红外泛光至一待测物体上，感测该待测物体的第一红外图像；
步骤S2：根据步骤S1获得的第一红外图像，比对该待测物体的二维图像信息与预存的二维图像信息是否匹配；
步骤S3：投射红外结构光至该待测物体上，感测该待测物体的第二红外图像；
步骤S4：根据步骤S3获得的第二红外图像，判断该待测物体是否为立体物体；
步骤S5：根据步骤S2与步骤S4的执行结果，确认所述待测物体的身份是否合法。
在本申请的实施方式中，通过光学图像传感的方式对该待测物体的身份进行鉴权。其中，根据该第一红外图像能够获得该待测物体的2D图像信息，根据该第二红外图像能够获得该待测物体的3D图像信息，从而，根据该2D与3D图像信息是可以确认该待测物体的身份是否合法。由此可见，本申请提供了一种新型的光学传感技术来实现身份鉴权。
另外，该光学传感技术可适用于较远距离的感测，且感测响应速度较快。所述较远距离例如为1米范围内或甚至更远一些的距离。
进一步地，由于本申请的身份鉴权方法是通过比对二维图像信息与判断该待测物体是否为立体物体，来对该待测物体的身份是否合法进行确认，从而可以达到节省功耗、减少感测时间、降低成本的功效。
本申请还提供一种身份鉴权装置，包括：
第一投射器，用于投射红外泛光至一待测物体；
第二投射器，用于投射红外结构光至该待测物体；
图像传感装置，用于捕获由该待测物体反射回来的红外泛光、感测获得该待测物体的第一红外图像，以及用于捕获由该待测物体反射回来的红外结构光、感测获得该待测物体的第二红外图像；
处理器，用于根据该第一红外图像，比对所述待测物体的二维图像信息与预存的二维图像信息是否匹配，并获得比对结果；所述处理器还用于根据该第二红外图像，判断所述待测物体是否为立体物体，并获得判断结果；所述处理器根据该比对结果与判断结果，确认所述待测物体的身份是否合法。
在本申请的实施方式中，该身份鉴权装置通过光学图像传感的方式对该待测物体的身份进行鉴权。其中，该处理器根据该第一红外图像能够获得该待测物体的2D图像信息，根据该第二红外图像能够获得该待测物体的3D图像信息，从而，根据该第一红外图像和该第二红外图像，能够确认该待测物体的身份是否合法。由此可见，本申请提供了一种新型的光学式身份鉴权装置来实现身份鉴权。
另外，该光学式身份鉴权装置可适用于较远距离的感测，且感测响应速度较快。所述较远距离例如为1米范围内或甚至更远一些的距离。
进一步地，由于本申请的身份鉴权装置是通过比对二维图像信息与判断该待测物体是否为立体物体，来对该待测物体的身份是否合法进行确认，从而可以达到节省功耗、减少感测时间、降低成本的功效。
本申请还提供一种电子设备，包括上述中任意一项所述的身份鉴权装置。
在某些实施方式中，所述电子设备用于根据所述身份鉴权装置的身份鉴权结果来对应是否执行相应的功能。
在某些实施方式中，所述相应的功能包括解锁、支付、启动预存的应用程序中的任意一种或几种。
在某些实施方式中，所述电子设备包括消费性电子产品、家居式电子产品、车载式电子产品、金融终端产品中的任意一种或几种。
由于本申请的电子设备包括上述的身份鉴权装置，因此，该电子设备能够实现对待测物体的较远距离的感测，且感测响应速度较快。
本申请实施方式的附加方面和优点将在下面的描述中部分给出，部分将从下面的描述中变得明显，或通过本申请实施方式的实践了解到。
附图说明
本申请实施方式的上述和/或附加的方面和优点从结合下面附图对实施方式的描述中将变得明显和容易理解，其中：
图1为本申请的身份鉴权方法的流程示意图。
图2为环境光的辐射强度与波长之间的关系示意图。
图3为本申请的身份鉴权方法的第一实施方式的细化流程示意图。
图4为本申请的身份鉴权方法的第二实施方式的细化流程示意图。
图5是本申请的身份鉴权装置一实施方式的结构框图。
图6是本申请的电子设备一实施方式的结构示意图。
具体实施方式
下面详细描述本申请的实施方式，所述实施方式的示例在附图中示出，其中自始至终相同或类似的标号表示相同或类似的元件或具有相同或类似功能的元件。下面通过参考附图描述的实施方式是示例性的，仅用于解释本申请，而不能理解为对本申请的限制。
在本申请的描述中，需要理解的是，术语“第一”、“第二”仅用于描述目的，而不能理解为指示或暗示相对重要性或者隐含指明所指示的技术特征的数量。由此，限定有“第一”、“第二”的特征可以明示或者隐含地包括一个或者更多个所述特征。在本申请的描述中，“多个”的含义是两个或两个以上，除非另有明确具体的限定。
在本申请的描述中，需要说明的是，除非另有明确的规定和限定，术语“安装”、“相连”、“连接”应做广义理解，例如，可以是固定连接，也可以是可拆卸连接，或一体地连接；可以是机械连接，也可以是电连接或可以相互通信；可以是直接相连，也可以通过中间媒介间接相连，可以是两个元件内部的连通或两个元件的相互作用关系。对于本领域的普通技术人员而言，可以根据具体情况理解上述术语在本申请中的具体含义。
下文的公开提供了许多不同的实施方式或例子用来实现本申请的不同结构。为了简化本申请的公开，下文中对特定例子的部件和设定进行描述。当然，它们仅仅为示例，并且目的不在于限制本申请。此外，本申请可以在不同例子中重复参考数字和/或参考字 母，这种重复是为了简化和清楚的目的，其本身不指示所讨论各种实施方式和/或设定之间的关系。
进一步地，所描述的特征、结构可以以任何合适的方式结合在一个或更多实施方式中。在下面的描述中，提供许多具体细节从而给出对本申请的实施方式的充分理解。然而，本领域技术人员应意识到，没有所述特定细节中的一个或更多，或者采用其它的结构、组元等，也可以实践本申请的技术方案。在其它情况下，不详细示出或描述公知结构或者操作以避免模糊本申请。
更进一步地，需要提前说明的是，本申请的说明书以及权利要求书中涉及的步骤编号S1、S2、S3、S4、S5只是为了清楚区分各步骤，并不代表步骤执行的先后顺序。
请参阅图1，图1为本申请的身份鉴权方法的流程示意图。该身份鉴权方法例如但不局限于应用在电子设备上，所述电子设备例如但不局限于为消费性电子产品、家居式电子产品、车载式电子产品、金融终端产品等合适类型的电子产品。其中，消费性电子产品例如但不局限为手机、平板电脑、笔记本电脑、桌面显示器、电脑一体机等。家居式电子产品例如但不局限为智能门锁、电视、冰箱、穿戴式设备等。车载式电子产品例如但不局限为车载导航仪、车载DVD等。金融终端产品例如但不局限为ATM机、自助办理业务的终端等。该身份鉴权方法包括：
步骤S1：投射红外泛光至一待测物体上，感测该待测物体的第一红外图像；
步骤S2：根据步骤S1获得的第一红外图像，比对该待测物体的二维图像信息与预存的二维图像信息是否匹配；
步骤S3：投射红外结构光至该待测物体上，感测该待测物体的第二红外图像；
步骤S4：根据步骤S3获得的第二红外图像，判断该待测物体是否为立体物体；和
步骤S5：根据步骤S2与步骤S4的执行结果，确认所述待测物体的身份是否合法。
需要说明的是，步骤S1与步骤S3分时进行，以避免第一红外图像和第二红外图像混叠。
较佳地，步骤S2与步骤S4先后进行。当步骤S2与S4中的任意一个步骤先被执行而获得的结果是否定的结果时，则步骤S5确认所述待测物体的身份非法，即，身份鉴权失败，流程结束，无需再执行其它未进行的鉴权步骤。举例，当步骤S2确认该待测物体的二维图像信息与预存的二维图像信息不匹配时，则身份鉴权失败，步骤S4则不需要再被执行。类似地，当步骤S4先被执行而获得的结果是否定的结果时，则身份鉴权失败，步骤S2不需要再被执行。
相对地，当步骤S2与S4中的任意一个步骤先被执行而获得的结果是肯定的结果时，则继续执行其它未进行的鉴权步骤。如此，可以节省感测时间、提高感测响应速度。
当然，可变更地，步骤S2与步骤S4也可同时进行。
另外，例如步骤S4只需确认该待测物体是否为立体物体即可，而无需对待测物体的整个立体图像信息进行分析与计算，从而能够降低功耗，节省成本。
在本申请的实施方式中，通过投射红外泛光与红外结构光到该待测物体上，分别获得该待测物体的不同红外图像，来实现对该待测物体的感测与识别。其中，根据该第一红外图像能够获得该待测物体的二维(2-Dimension,2D)图像信息，根据该第二红外图像能够获得该待测物体的三维(3-Dimension,3D)图像信息，该3D图像信息例如包括深度信息。
通过比对该待测物体的二维图像信息与预存的二维图像信息是否匹配，以对该待测物体进行一次识别，当识别不通过时，则身份鉴权失败，流程结束。
另外，根据该第二红外图像判断该待测物体是否为立体物体，以对该待测物体进行一次识别，当识别不通过时，则身份鉴权失败，流程结束。
当至少上述两次识别都通过时，步骤S5确认该待测物体的身份合法，身份鉴权成功。
由上述内容可知，当上述两次识别都通过后，可以确认该待测物体的身份合法，身份鉴权成功。
另外，需要说明的是，在上述实施方式的身份鉴权方法中，额外增加某些步骤也是可行的。例如，在执行身份鉴权的过程，进一步增加步骤：根据该第一红外图像或/和第二红外图像，确认该待测物体是否为活体。又例如，增加步骤：根据该第一红外图像或/和第二红外图像，确认该待测物体的眼睛是否注视在电子设备前方的预定范围内等。从而，使得本申请的身份鉴权方法的效果更好。由前述可知，本申请并不限于以上内容所公开的技术方案，只要是技术思想与本申请的技术思想相同或相似的发明，均应落在本申请的保护范围。
可见，本申请提供了一种新型的光学传感技术来实现身份鉴权。
进一步地，由于红外泛光和红外结构光的光学特性，该光学传感技术可适用于较远距离的感测，且感测响应速度较快。所述较远距离例如为1米范围内或甚至更远一些的距离。
该待测物体例如为人体的面部。相应地，本申请的身份鉴权方法用于识别人脸。然， 本申请并不局限于此，该待测物体例如也可为人体的其它合适部位，甚至是其它合适的生物体或非生物体等。
下面以人脸识别为例进行说明。在进行脸部识别之前，用户已提前注册好其本人的脸部图像模板，并存储在例如一存储器中。该脸部图像模板例如包括所述二维图像信息以及深度信息。
在步骤S1中，例如采用红外泛光灯投射红外泛光至该待测物体，并利用图像传感装置捕获由该待测物体反射回来的红外泛光，从而获得第一红外图像。
在步骤S2中，对该待测物体的识别就是：比对该待测物体的二维图像信息是否与已注册用户脸部的二维图像信息匹配，如果比对得知该待测物体的二维图像信息与该用户脸部的二维图像信息不匹配，则身份鉴权失败，流程结束。
如果比对得知该待测物体的二维图像信息与该用户脸部的二维图像信息匹配，则还不能确定该待测物体的身份合法，理由是：在步骤S2中，因为是二维图像信息的判断与识别，如果利用该用户的照片同样可以识别成功。
进一步地，关于二维图像信息的比对，例如，可以通过比对该待测物体的平面图片与预存的物体面部的平面图片来实现。
然而，本申请提出了用于二维图像信息比对的另一种实现方式。所述预存的二维图像信息包括面部特征信息。所述步骤S2进一步包括对待测物体进行面部特征信息提取，通过比对提取到的面部特征信息与预存的面部特征信息，来确认该待测物体的二维图像信息与预存的二维图像信息是否匹配。从而，相对于比对整幅图片来说，通过比对特征的方式能够进一步减少计算量，提高感测效率。
在一示例中，在步骤S2中，通过深度学习方法提取该待测物体的面部特征信息。该深度学习方法包括：建立深度卷积神经网络模型，使用预定数量的人脸照片训练该深度卷积神经网络模型，根据训练好的该深度卷积神经网络模型提取人脸的特征参数。
具体地，所述面部特征信息例如包括鼻子、眼睛、嘴巴、眉毛、额头、颧骨、下巴、脸庞中的任意一种或几种、或/和它们之间任意组合的距离信息。可扩展地，所述预存的面部特征信息也可包括鼻子、眼睛、嘴巴、眉毛、额头、颧骨、下巴、脸庞，甚至包括鼻子的宽度、下巴的宽度等等。
可变更地，在其它实施方式中，在步骤S2中，也可提取鼻子、眼睛、嘴巴、眉毛、额头、颧骨、下巴、脸庞、鼻子的宽度、下巴的宽度等面部特征，或/和，鼻子、眼睛、嘴巴、眉毛、额头、颧骨、下巴等中任意组合的距离信息。例如，鼻子和眼睛之间的距 离信息。当然，所述面部特征信息并不局限于上面所列举的例子，也可为其它合适的特征信息。
具体地，例如，在步骤S2中，当确认该待测物体的二维图像信息与预存的二维图像信息的匹配系数大于或等于一预定阈值时，则可确认该待测物体的二维图像信息与预存的二维图像信息匹配。相对地，如果确认匹配系数小于该预定阈值时，则可确认该待测物体的二维图像信息与预存的二维图像信息不匹配。
本申请不限于上面所涉及的二维图像信息的比对方式，也可为其它合适的比对方式。
在步骤S3中，例如采用光学组件来投射红外结构光至该待测物体，并利用红外图像传感装置捕获由该待测物体反射回来的红外结构光，感测获得该待测物体的第二红外图像。所述光学组件例如包括光源、准直镜头以及光学衍射元件(DOE)，其中光源用于产生一红外激光束；准直镜头将红外激光束进行校准，形成近似平行光；光学衍射元件对校准后的红外激光束进行调制，形成相应的散斑图案。该散斑图案例如但不局限于包括规则点阵式、条纹式、网格式、散斑式、编码式等中的一种或几种。其中，散斑式又称为随机点阵式。编码式图案例如由不同波形的光组成，每种波形代表一种数字，各波形的组合即为编码。
上述是利用基于光编码原理，投射已知的红外结构光图案到该待测物体上。图像传感装置或处理器根据捕获到的变形的红外结构光图案来分析确定该待测物体的深度信息。定义此类红外结构光为空间结构光。
可变更地，例如也可利用基于飞行时间(Time of Flight,ToF)原理，投射红外结构光至该待测物体。图像传感装置或处理器例如通过测量光脉冲之间的传输延迟时间来计算待测物体的深度信息。定义此类红外结构光为时间结构光。
该时间结构光例如但不局限于呈正弦波、方波中的任意一种或两种的结合。
在步骤S4中，根据步骤S3获得的第二红外图像，判断该待测物体是否为立体物体。例如，根据该第二红外图像能够获得深度信息，从而根据深度信息来判断该待测物体是否为立体物体。
如果判断该待测物体不是立体物体，则身份鉴权失败，流程结束。在此种情况下，就有可能是他人利用合法用户的照片或者视频等进行身份识别。
在本实施方式中，在步骤S5中，当确认：步骤S2中的待测物体的二维图像信息与预存的二维图像信息匹配成功、步骤S4中判断得知该待测物体为立体物体时，则身份鉴权成功。
可选地，步骤S4进一步包括：对该第二红外图像进行立体特征信息的提取，并根据提取到的立体特征信息来判断所述待测物体是否为立体物体。
在一示例中，通过深度学习方法提取该待测物体的立体人脸特征信息。该深度学习方法包括：建立深度卷积神经网络模型，使用预定数量的人脸照片训练该深度卷积神经网络模型，根据训练好的该深度卷积神经网络模型提取人脸的特征参数。
可变更地，在其它实施方式中，在步骤S4中，也可提取鼻子、眼睛、嘴巴、眉毛、额头、颧骨、下巴、脸庞等面部特征中的任意一种或几种的立体尺寸信息。所述立体尺寸信息即为深度信息。当然，所述立体特征信息并不局限于上面所列举的例子，也可为其它合适的特征信息。
本申请并不限于上述对立体物体的判断方式，例如，可变更地，在其它实施方式中，也可以根据第二红外图像，构建出立体图像信息，并与预存的立体图像信息进行比对，来判断该待测物体是不是立体物体。
由上述内容可知，本申请的身份鉴权方法能够节省感测时间、节省功耗、并能降低成本。
然，可变更地，在某些实施方式中，步骤S4还包括：当确认该待测物体是立体物体后，进一步判断该待测物体的立体信息是否符合人体的面部立体特征。
在一示例中，当确认该待测物体是立体物体后，通过计算扭曲率来获得鼻子、眼睛等立体特征，以及计算获得这些立体特征之间的距离，从而可以判断得知该待测物体的立体信息是否符合人体的面部立体特征。
当判断得知该待测物体的立体信息不符合人体的面部立体特征时，则身份鉴权失败。
可选地，在步骤S5中，当确认该待测物体的二维图像信息与预存的二维图像信息匹配，且判断得知该待测物体的立体信息符合人体的面部立体特征时，则身份鉴权成功。
现有的，业界通常投射波长为850纳米的近红外光，来获得待测物体的红外图像。然而，本申请的发明人经过大量的创造性劳动，分析与研究发现：投射波长为940纳米的红外泛光、940纳米的红外结构光进行感测，可以获得较准确的感测效果。
请参阅图2，图2为环境光的辐射强度与波长之间的关系示意图。其中，波长用横轴表示，且被标示为字母λ，辐射强度用纵轴表示，且被标示为字母E。发明人通过理论研究、结合大量的实验测试、验证并反复进行分析与研究等，创造性地发现：环境光中波长范围为[920,960]纳米的近红外光易被大气吸收、强度衰减较大，当步骤S1投射波长范围为[920,960]纳米的红外泛光到待测物体，根据捕获的红外泛光获得该待测物体 的第一红外图像时，能够少受环境光的干扰，从而提高图像的获取精度。类似地，当步骤S3投射波长范围为[920,960]纳米的红外结构光到待测物体，根据捕获的红外结构光获得该待测物体的第二红外图像时，能够少受环境光的干扰，从而提高图像的获取精度。
进一步地，在波长范围为[920,960]纳米的红外光中，波长为940纳米的近红外光更易被大气吸收、强度衰减最大，因此，在本申请的实施方式中，步骤S1投射的红外泛光的波长优选为940纳米，步骤S3投射的红外结构光的波长优选为940纳米。
然而，在实际应用中，步骤S1所投射的红外泛光的波长和步骤S3所投射的红外结构光的波长在940纳米的基础上会有一定的偏差，例如会有(+15)纳米或(-15)纳米左右的偏差。因此，步骤S1投射的红外泛光的波长范围例如为[925,955]纳米，步骤S3投射的红外结构光的波长范围例如为[925,955]纳米。可见，该波长范围[925,955]仍然落在波长范围[920,960]内。
需要说明的是，步骤S1所投射的红外泛光的波长和步骤S3所投射的红外结构光的波长为落在上述波长范围[920,960]纳米中的任意一数值。本申请为了叙述简洁清楚，在此处并未一一列举各具体数值，但落在这波长范围[920,960]纳米中的任意一数值都是可行的。
当然，可变更地，本申请的身份鉴权方法的步骤S1和步骤S3也可采用波长为850纳米或者其它合适波长的红外泛光、红外结构光进行感测。
请参阅图3，图3为本申请的身份鉴权方法的第一实施方式的细化流程示意图。在此实施方式中，步骤S2先于步骤S4执行。当步骤S2中确认该待测物体的二维图像信息与预存的二维图像信息匹配时，则开始启动执行步骤S4，而当步骤S2中确认该待测物体的二维图像信息与预存的二维图像信息不匹配时，则身份鉴权失败，流程结束。
其中，在一具体实施例中，步骤S2与步骤S3同时进行。如此，能够更进一步减少感测时间，提升工作效率。
然，可变更地，步骤S3也可于步骤S2之后进行。对于此种情况，当步骤S2中确认该待测物体的二维图像信息与预存的二维图像信息匹配时，则开始启动执行步骤S3，然后再执行步骤S4。如此，能够减少功耗。
步骤S3也可于步骤S2之前进行。对于此种情况，又可分为两种实施方式，步骤S1可先于步骤S3进行，也可后于步骤S3进行。
以人脸识别为例，本实施方式的身份鉴权方法需对该待测物体进行两次识别，其中，对该待测物体的第一次识别就是：比对该待测物体的二维图像信息是否与已注册用户脸 部的二维图像信息匹配，如果比对得知该待测物体的二维图像信息与该用户脸部的二维图像信息不匹配，则确认该待测物体的身份非法，身份鉴权失败，流程结束。
如果判断得知该待测物体的二维图像信息与该用户脸部的二维图像信息匹配，则还不能确定该待测物体的身份合法，理由是：在步骤S2中，因为是二维图像信息的判断与识别，如果利用合法用户的照片同样可以识别成功。
接下来，通过执行步骤S4，来避免上述利用照片识别通过的情况。
在步骤S4中，根据步骤S3获得的第二红外图像，判断该待测物体是否为立体物体。例如，根据该第二红外图像能够获得深度信息，从而根据深度信息来判断该待测物体是否为立体物体。
如果判断得知该待测物体不是立体物体，则身份鉴权失败，流程结束。在此种情况下，就有可能是他人利用合法用户的照片或者视频等进行身份识别。
可选地，在步骤S5中，当确认该待测物体的二维图像信息与预存的二维图像信息匹配、以及确认该待测物体是立体物体时，则确认该待测物体的身份合法，身份鉴权成功。
由于在步骤S4中只需判断该待测物体是不是立体物体即可，而无需对该待测物体的立体图像信息进行大量的分析与计算，从而可以降低功耗、减少感测时间、以及降低成本。
具体地，在步骤S2中，当确认该待测物体的二维图像信息与已注册用户脸部的二维图像信息的匹配系数大于或等于一预定阈值时，则可确认该待测物体的二维图像信息与已注册用户脸部的二维图像信息匹配。相对地，如果确认匹配系数小于该预定阈值时，则可确认该待测物体的二维图像信息与已注册用户脸部的二维图像信息不匹配。
如前所述，在上述实施方式的身份鉴权方法中，额外增加某些步骤也是可行的。例如，在执行身份鉴权的过程，进一步增加步骤：根据该第一红外图像或/和第二红外图像，确认该待测物体是否为活体。又例如，增加步骤：根据该第一红外图像或/和第二红外图像，确认该待测物体的眼睛是否注视在电子设备前方的预定范围内等。从而，使得本申请的身份鉴权方法的效果更好。由前述可知，本申请并不限于以上内容所公开的技术方案，只要是技术思想与本申请的技术思想相同或相似的发明，均应落在本申请的保护范围。
请参阅图4，图4为本申请的身份鉴权方法的第二实施方式的细化流程示意图。在本实施方式中，步骤S4先于步骤S2执行。当步骤S4中判断得知该待测物体为立体物 体时，则开始启动执行步骤S2，而当步骤S4中判断得知该待测物体不是立体物体时，则身份鉴权失败，流程结束。
由于步骤S4的功耗相较步骤S2的功耗要低，因此，当步骤S4先于步骤S2执行时，且步骤S4不通过时，则该身份鉴权方法的感测功耗要较低。
其中，在一具体实施例中，步骤S4与步骤S1同时进行。如此，能够减少感测时间，提升工作效率。
然，可变更地，步骤S1也可于步骤S4之后进行。对于此种情况，当步骤S4中判断得知该待测物体为立体物体时，开始启动执行步骤S1，然后再执行步骤S2。如此，能够减少功耗。
步骤S1也可于步骤S4之前进行。对于此种情况，又可分为两种实施方式，步骤S3可先于步骤S1执行，也可后于步骤S1执行。
以人脸识别为例，本实施方式的身份鉴权方法需对该待测物体进行两次识别，其中，对该待测物体的第一次识别就是：判断该待测物体是否为立体物体，如果判断得知该待测物体不是立体物体时，则确认该待测物体的身份不合法，身份鉴权失败，流程结束。
如果判断得知该待测物体是立体物体，则还不能确定该待测物体的身份合法。接下来，通过执行步骤S2，来识别该待测物体与预先注册的二维脸部信息是否匹配。
在步骤S2中，根据步骤S1获得的第一红外图像，比对该待测物体的二维图像信息与预存的二维图像信息是否匹配，如果确认该待测物体的二维图像信息与该用户脸部的二维图像信息不匹配，则身份鉴权失败。
可选地，在步骤S5中，当确认该待测物体的二维图像信息与预存的二维图像信息匹配、以及判断得知该待测物体是立体物体时，则身份鉴权成功。
另外，如前所述，在上述实施方式的身份鉴权方法中，额外增加某些步骤也是可行的。例如，在执行身份鉴权的过程，进一步增加步骤：根据该第一红外图像或/和第二红外图像，确认该待测物体是否为活体。又例如，增加步骤：根据该第一红外图像或/和第二红外图像，确认该待测物体的眼睛是否注视在电子设备前方的预定范围内等。从而，使得本申请的身份鉴权方法的效果更好。由前述可知，本申请并不限于以上内容所公开的技术方案，只要是技术思想与本申请的技术思想相同或相似的发明，均应落在本申请的保护范围。
请参阅图5，图5是本申请的身份鉴权装置的一实施方式的结构框图。该身份鉴权装置1包括第一投射器10、第二投射器12、图像传感装置14、处理器16、和存储器18。 其中，该存储器18用于预存一个或多个样本物体的二维图像信息。该第一投射器10用于投射红外泛光至一待测物体。该第二投射器12用于投射红外结构光至该待测物体。该图像传感装置14用于捕获由该待测物体反射回来的红外泛光、并根据捕获的红外泛光获得该待测物体的第一红外图像，以及用于捕获由该待测物体反射回来的红外结构光、并根据捕获的红外结构光获得该待测物体的第二红外图像。该处理器16用于根据该第一红外图像，比对该待测物体的二维图像信息与预存的二维图像信息是否匹配，并获得比对结果，还用于根据该第二红外图像，判断所述待测物体是否为立体物体，并获得判断结果，以及用于根据该比对结果与判断结果，确认所述待测物体的身份是否合法。
可选地，该身份鉴权装置1进一步包括控制电路17。该控制电路17用于控制该第一投射器10、第二投射器12、和该图像传感装置14协同工作。
当进行身份鉴权时，所述控制电路17例如用于控制该第一投射器10与该第二投射器12分时工作，以避免该图像传感装置14所感测到的第一红外图像和第二红外图像发生混叠。该控制电路17可控制该第一投射器10先于该第二投射器12工作，也可控制该第二投射器12先于该第一投射器10工作。
较佳地，“根据该第一红外图像、比对该待测物体的二维图像信息与预存的二维图像信息是否匹配”，以及“根据该第二红外图像、判断该待测物体是否为立体物体”这二者可被该处理器16先后执行。当这二者中的任意一者先被该处理器16执行而获得的结果是否定的结果时，则该处理器16确认所述待测物体的身份非法，即，身份鉴权失败，流程结束，该处理器16无需再执行其它未进行的鉴权程序。举例，当该处理器16确认该待测物体的二维图像信息与预存的二维图像信息不匹配时，则身份鉴权失败，该处理器16则无需再执行“根据该第二红外图像、判断该待测物体是否为立体物体”等未进行的鉴权程序。
相对地，当这二者中的任意一者先被该处理器16执行而获得的结果是肯定的结果时，则该处理器16继续执行其它未进行的鉴权程序。如此，可以节省感测时间、提高感测响应速度。
当然，可变更地，这二者也可被该处理器16同时执行。
另外，例如该处理器16只需确认该待测物体是否为立体物体即可，而无需对该待测物体的整个立体图像信息进行分析与计算，从而能够降低功耗，节省成本。
在本申请的实施方式中，当进行身份鉴权时，该身份鉴权装置1通过投射红外泛光与红外结构光到该待测物体上，分别获得该待测物体的不同红外图像，来实现对该待测 物体的感测与识别。其中，该处理器16根据该第一红外图像能够获得该待测物体的二维(2-Dimension,2D)图像信息，根据该第二红外图像能够获得该待测物体的三维(3-Dimension,3D)图像信息，该3D图像信息例如包括深度信息。
通过比对该待测物体的二维图像信息与预存的二维图像信息是否匹配，以对该待测物体进行一次识别，当识别不通过时，则身份鉴权失败，流程结束。
另外，根据该第二红外图像判断该待测物体是否为立体物体，以对该待测物体进行一次识别，当识别不通过时，则身份鉴权失败，流程结束。
当至少上述两次识别都通过时，该处理器16确认该待测物体的身份合法，身份鉴权成功。
另外，需要说明的是，在上述实施方式的身份鉴权装置1中，该处理器16例如也可执行额外的鉴权程序。例如，在执行身份鉴权的过程，该处理器16进一步执行：根据该第一红外图像或/和第二红外图像，确认该待测物体是否为活体；又例如，根据该第一红外图像或/和第二红外图像，确认该待测物体的眼睛是否注视在电子设备前方的预定范围内等。从而，使得本申请的身份鉴权装置1的效果更好。由前述可知，本申请并不限于以上内容所公开的技术方案，只要是技术思想与本申请的技术思想相同或相似的发明，均应落在本申请的保护范围。
由上述内容可知，本申请提供了一种新型的光学式身份鉴权装置1。
进一步地，由于红外泛光和红外结构光的光学特性，该光学传感技术可适用于较远距离的感测，且感测响应速度较快。所述较远距离例如为1米范围内或甚至更远一些的距离。
所述第一投射器10例如但不局限于为红外泛光灯。
所述第二投射器12例如采用光学组件来投射红外结构光至该待测物体。所述光学组件例如包括光源、准直镜头以及光学衍射元件(DOE)，其中光源用于产生一红外激光束；准直镜头将红外激光束进行校准，形成近似平行光；光学衍射元件对校准后的红外激光束进行调制，形成相应的散斑图案。该散斑图案例如但不局限于包括规则点阵式、条纹式、网格式、散斑式、编码式等中的一种或几种。其中，散斑式又称为随机点阵式。编码式图案例如由不同波形的光组成，每种波形代表一种数字，各波形的组合即为编码。另外，所述光学组件例如也可由其它合适的光学元件构成。
上述是利用基于光编码原理，该第二投射器12投射已知的红外结构光图案到该待测物体上。图像传感装置14或处理器16根据捕获到的变形的红外结构光图案来分析确 定该待测物体的深度信息。定义此类红外结构光为空间结构光。
可变更地，例如也可利用基于飞行时间(Time of Flight,ToF)原理，该第二投射器12投射红外结构光至该待测物体。图像传感装置14或处理器16例如通过测量光脉冲之间的传输延迟时间来计算待测物体的深度信息。定义此类红外结构光为时间结构光。
该时间结构光例如但不局限于呈正弦波、方波中的任意一种或两种的结合。
当该第二投射器12投射的是空间结构光时，该图像传感装置14例如包括一红外图像传感器141，该红外图像传感器141用于捕获由该待测物体反射回来的红外泛光和红外结构光。由于共用同一红外图像传感器141，从而可以降低成本。
然而，当该第二投射器12投射的是时间结构光时，该图像传感装置14例如包括二红外图像传感器，该二红外图像传感器的结构不同，感测原理不同，分辨率不同等。其中，一红外图像传感器用于捕获由该待测物体反射回来的红外泛光，另一红外图像传感器用于捕获由该待测物体反射回来的红外结构光。
该待测物体例如为人体的面部。相应地，本申请的身份鉴权装置用于识别人脸。然，本申请并不局限于此，该待测物体例如也可为人体的其它合适部位，甚至是其它合适的生物体或非生物体等。
下面以人脸识别为例进行说明。在进行脸部识别之前，用户已提前注册好其本人的脸部图像模板，并存储在例如该存储器18中。该脸部图像模板例如包括所述二维图像信息以及深度信息。其中，该存储器18例如可以存储一张或多张人脸图像模板。
相应地，该处理器16比对该待测物体的二维图像信息是否与已注册用户脸部的二维图像信息匹配，如果比对得知该待测物体的二维图像信息与该用户脸部的二维图像信息不匹配，则确认该待测物体的身份非法，身份鉴权失败，流程结束。
如果该处理器16比对得知该待测物体的二维图像信息与该用户脸部的二维图像信息匹配，则还不能确定该待测物体的身份合法，理由是：因为是二维图像信息的判断与识别，如果他人利用合法用户的照片同样可以识别成功。
进一步地，关于二维图像信息的比对，例如，可以通过比对该待测物体的平面图片与预存的物体面部的平面图片来实现。
然而，本申请提出了用于二维图像信息比对的另一种实现方式。所述预存的二维图像信息包括面部特征信息。所述处理器16进一步包括对该待测物体进行面部特征信息提取，通过比对提取到的面部特征信息与预存的面部特征信息，来确认该待测物体的二维图像信息与预存的二维图像信息是否匹配。从而，相对于比对整幅图片来说，通过比 对特征的方式能够进一步减少计算量，提高感测效率。
在一示例中，所述处理器16通过深度学习方法提取该待测物体的面部特征信息。该深度学习方法包括：建立深度卷积神经网络模型，使用预定数量的人脸照片训练该深度卷积神经网络模型，根据训练好的该深度卷积神经网络模型提取人脸的特征参数。
可变更地，在其它实施方式中，该处理器16也可提取鼻子、眼睛、嘴巴、眉毛、额头、颧骨、下巴、脸庞、鼻子的宽度、下巴的宽度等面部特征，或/和，鼻子、眼睛、嘴巴、眉毛、额头、颧骨、下巴等中任意组合的距离信息。例如，鼻子和眼睛之间的距离信息。当然，所述面部特征信息并不局限于上面所列举的例子，也可为其它合适的特征信息。
具体地，例如，当该处理器16确认该待测物体的二维图像信息与已注册用户脸部的二维图像信息的匹配系数大于或等于一预定阈值时，则可确认该待测物体的二维图像信息与已注册用户脸部的二维图像信息匹配。相对地，如果确认匹配系数小于该预定阈值时，则可确认该待测物体的二维图像信息与已注册用户脸部的二维图像信息不匹配。
本申请不限于上面所涉及的二维图像信息的比对方式，也可为其它合适的比对方式。
进一步地，该处理器16例如对该第二红外图像进行立体特征信息的提取，并根据提取到的立体特征信息来判断所述待测物体是否为立体物体。
在一示例中，该处理器16例如通过深度学习方法提取该待测物体的立体人脸特征信息。该深度学习方法包括：建立深度卷积神经网络模型，使用预定数量的人脸照片训练该深度卷积神经网络模型，根据训练好的该深度卷积神经网络模型提取人脸的特征参数。
可变更地，在其它实施方式中，该处理器16也可提取鼻子、眼睛、嘴巴、眉毛、额头、颧骨、下巴、脸庞等面部特征中的任意一种或几种的立体尺寸信息。所述立体尺寸信息即为深度信息。当然，所述立体特征信息并不局限于上面所列举的例子，也可为其它合适的特征信息。
该处理器16例如根据该第二红外图像能够获得深度信息，从而，该处理器16根据深度信息来判断该待测物体是否为立体物体。
如果判断得知该待测物体不是立体物体，则身份鉴权失败，流程结束。在此种情况下，就有可能是他人利用合法用户的照片或者视频等进行身份识别。
在本实施方式中，当该处理器16确认：该待测物体的二维图像信息与预存的二维图像信息匹配、且该待测物体为立体物体时，则身份鉴权成功。
然，本申请并不限于上述对立体物体的判断方式，例如，可变更地，在其它实施方式中，该存储器18可进一步用于预存该样本物体的立体图像信息。相应地，该处理器16也可以根据第二红外图像，构建出立体图像信息，并与预存的立体图像信息进行比对，来判断该待测物体是不是立体物体。
由上述内容可知，本申请的身份鉴权装置1能够节省感测时间、节省功耗、并能降低成本。
然，可变更地，在某些实施方式中，该处理器16还用于：当确认该待测物体是立体物体后，进一步判断该待测物体的立体信息是否符合人体的面部立体特征。
在一示例中，当该处理器16确认该待测物体是立体物体后，通过计算扭曲率来获得鼻子、眼睛等立体特征，以及计算获得这些立体特征之间的距离，从而可以判断得知该待测物体的立体信息是否符合人体的面部立体特征。
当该处理器16判断得知该待测物体的立体信息不符合人体的面部立体特征时，则身份鉴权失败。
可选地，当该处理器16确认该待测物体的二维图像信息与预存的二维图像信息匹配，且判断得知该待测物体的立体信息符合人体的面部立体特征时，则身份鉴权成功。
现有的，业界通常投射波长为850纳米的近红外光，来获得待测物体的红外图像。然而，本申请的发明人经过大量的创造性劳动，分析与研究发现：投射波长为940纳米的红外泛光、940纳米的红外结构光进行感测，可以获得较准确的感测效果。
请再参阅图2，环境光中波长范围为[920,960]纳米的近红外光易被大气吸收、强度衰减较大，当该第一投射器10投射波长范围为[920,960]纳米的红外泛光到待测物体、该图像传感装置14根据捕获的红外泛光获得该待测物体的第一红外图像时，能够少受环境光的干扰，从而提高图像的获取精度。类似地，当该第二投射器12投射波长范围为[920,960]纳米的红外结构光到待测物体、该图像传感装置14根据捕获的红外结构光获得该待测物体的第二红外图像时，能够少受环境光的干扰，从而提高图像的获取精度。
进一步地，在波长范围为[920,960]纳米的红外光中，波长为940纳米的近红外光更易被大气吸收、强度衰减最大，因此，在本申请的实施方式中，该第一投射器10投射的红外泛光的波长优选为940纳米，该第二投射器12投射的红外结构光的波长优选为940纳米。
然而，在实际应用中，该第一投射器10所投射的红泛外光的波长和该第二投射器12所投射的红外结构光的波长在940纳米的基础上会有一定的偏差，例如会有(+15)纳 米或(-15)纳米左右的偏差。因此，该第一投射器10所投射的红外泛光的波长范围例如为[925,955]纳米，该第二投射器12所投射的红外结构光的波长范围例如为[925,955]纳米。可见，该波长范围[925,955]仍然落在波长范围[920,960]内。
需要说明的是，该第一投射器10所投射的红外泛光的波长和该第二投射器12所投射的红外结构光的波长为落在上述波长范围[920,960]纳米中的任意一数值。本申请为了叙述简洁清楚，在此处并未一一列举各具体数值，但落在这波长范围[920,960]纳米中的任意一数值都是可行的。
当然，可变更地，该第一投射器10和第二投射器12也可分别投射波长为850纳米或者其它合适波长的红外泛光、红外结构光。
由于第一投射器10投射的是红外泛光，第二投射器12投射的是红外结构光，为了避免图像传感装置14感测获得的第一红外图像与第二红外图像发生混叠，因此，所述控制电路17控制所述第一投射器10和第二投射器12分时工作。
在某些实施方式中，例如，该处理器16先根据该第一红外图像，比对该待测物体的二维图像信息与预存的二维图像信息是否匹配，当确认该待测物体的二维图像信息与预存的二维图像信息不匹配时，则该处理器16确认该待测物体的身份不合法，即，身份鉴权失败，流程结束；当确认该待测物体的二维图像信息与预存的二维图像信息匹配时，则该处理器16再根据该第二红外图像判断该待测物体是否为立体物体；当该处理器16判断得知该待测物体不是立体物体时，则身份鉴权失败，流程结束。
可选地，当该处理器16判断得知该待测物体是立体物体时，则身份鉴权成功。
其中，在一具体实施例中，当该处理器16在根据该第一红外图像，比对该待测物体的二维图像信息与预存的二维图像信息是否匹配的同时：该控制电路17控制该第二投射器12投射红外结构光至该待测物体，该图像传感装置14感测该待测物体的第二红外图像。如此，能够更进一步减少感测时间，提升工作效率。
或，可变更地，当该处理器16确认该待测物体的二维图像信息与预存的二维图像信息匹配之后，该控制电路17再控制该第二投射器12投射红外结构光至该待测物体，该图像传感装置14感测该待测物体的第二红外图像。接着，该处理器16再根据该第二红外图像判断该待测物体是否为立体物体。如此，能够减少功耗。
另外，可变更地，该控制电路17控制该第一投射器10与第二投射器12先后工作，该图像传感装置14先后感测获得该第一红外图像与第二红外图像，然后，该处理器16再根据该第一红外图像，比对该待测物体的二维图像信息与预存的二维图像信息是否匹 配，并在确认该待测物体的二维图像信息与预存的二维图像信息时，再根据该第二红外图像判断该待测物体是否为立体物体。
当然，该控制电路17也可控制该第二投射器12与第一投射器10先后工作，该图像传感装置14先后感测获得该第二红外图像与第一红外图像，然后，该处理器16再根据该第一红外图像，比对该待测物体的二维图像信息与预存的二维图像信息是否匹配，并在确认该待测物体的二维图像信息与预存的二维图像信息时，再根据该第二红外图像判断该待测物体是否为立体物体。
以人脸识别为例，本实施方式的身份鉴权装置1需对该待测物体进行两次识别，其中，对该待测物体的第一次识别就是：该处理器16比对该待测物体的二维图像信息是否与已注册用户脸部的二维图像信息匹配，如果比对得知该待测物体的二维图像信息与该注册用户脸部的二维图像信息不匹配，则身份鉴权失败。
如果该处理器16判断得知该待测物体的二维图像信息与该注册用户脸部的二维图像信息匹配，则还不能确定该待测物体的身份合法，理由是：因为是二维图像信息的判断与识别，如果利用合法用户的照片同样可以识别成功。
接下来，该处理器16通过再判断该待测物体是否为立体物体，来避免上述利用照片识别通过的情况。
例如，由于根据该第二红外图像能够获得深度信息，从而根据深度信息来判断该待测物体是否为立体物体。如果判断该待测物体不是立体物体，则身份鉴权失败，流程结束。在此种情况下，就有可能是他人利用合法用户的照片或者视频等进行身份识别。
可选地，当该处理器16确认该待测物体的二维图像信息与预存的二维图像信息匹配、以及确认该待测物体是立体物体时，则身份鉴权成功。
由于本申请的身份鉴权装置1是对二维图像信息的比对与立体物体的判断，相对于计算出该待测物体的整体三维立体图像信息，然后再对该待测物体进行身份判断，本申请的身份鉴权装置1的功耗更少、感测响应速度更快、制造成本更低。
在上面的各实施方式中，简而言之，所述处理器16是先执行二维图像信息的比对，然后根据比对结果再确定是否执行立体物体的判断。然而，该处理器16也可先执行对立体物体的判断，然后根据判断结果确定是否再执行二维图像信息的比对。具体说明如下。
例如，该处理器16先根据第二红外图像，判断该待测物体是否为立体物体，当判断得知该待测物体不是立体物体时，则该处理器16明确该待测物体的身份不合法，即， 身份鉴权失败，流程结束；当该处理器16判断得知该待测物体是立体物体时，则该处理器16再根据该第一红外图像，比对该待测物体的二维图像信息与预存的二维图像信息是否匹配，如果确认该待测物体的二维图像信息与预存的二维图像信息不匹配，则身份鉴权失败，流程失败；如果确认该待测物体的二维图像信息与预存的二维图像信息匹配，则身份鉴权成功。
相较于对二维图像信息的比对，该处理器16执行对立体物体的判断所需消耗的功耗更少，因此，当该处理器16先执行对立体物体的判断，一旦在这个感测阶段通不过，则无需再进行后面的二维图像信息的比对，从而可减少功耗。
其中，在一具体实施例中，该处理器16在根据该第二红外图像，判断该待测物体是否为立体物体的同时：该控制电路17控制该第一投射器10开启，该图像传感装置14感测该待测物体的第一红外图像。如此，能够更进一步减少感测时间，提升工作效率。
或，可变更地，当该处理器16判断得知该待测物体为立体物体之后，该控制电路17再控制该第一投射器10开启，该图像传感装置14感测该待测物体的第一红外图像。接着，该处理器16再根据该第一红外图像，比对该待测物体的二维图像信息与预存的二维图像信息是否匹配。如此，能够减少功耗。
另外，可变更地，该控制电路17控制该第一投射器10与第二投射器12先后工作，该图像传感装置14先后感测获得该第一红外图像与第二红外图像，然后，该处理器16再开始执行“根据该第二红外图像判断该待测物体是否为立体物体”，并当判断得知该待测物体为立体物体时，则再执行“根据该第一红外图像，比对该待测物体的二维图像信息与预存的二维图像信息是否匹配”。
当然，该控制电路17也可控制该第二投射器12与第一投射器10先后工作，该图像传感装置14先后感测获得该第二红外图像与第一红外图像，然后，该处理器16再开始执行“根据该第二红外图像判断该待测物体是否为立体物体”，并当判断得知该待测物体为立体物体时，则再执行“根据该第一红外图像，比对该待测物体的二维图像信息与预存的二维图像信息是否匹配”。
请参阅图6，图6为本申请的电子设备的一实施方式的结构示意图。所述电子设备100例如但不局限于为消费性电子产品、家居式电子产品、车载式电子产品、金融终端产品等合适类型的电子产品。其中，消费性电子产品例如但不局限为手机、平板电脑、笔记本电脑、桌面显示器、电脑一体机等。家居式电子产品例如但不局限为智能门锁、电视、冰箱、穿戴式设备等。车载式电子产品例如但不局限为车载导航仪、车载DVD 等。金融终端产品例如但不局限为ATM机、自助办理业务的终端等。所述电子设备100包括上述身份鉴权装置1。所述电子设备100根据所述身份鉴权装置1的身份鉴权结果来对应是否执行相应的功能。所述相应的功能例如但不局限于包括解锁、支付、启动预存的应用程序中的任意一种或几种。
在本实施方式中，以电子设备为手机为例进行说明。所述手机例如为全面屏的手机，所述身份识别装置1例如设置在手机的正面顶端。当然，所述手机也并不限制于全面屏手机。
例如，当用户需要进行开机解锁时，抬起手机或触摸手机的屏幕都可以起到唤醒该身份鉴权装置1的作用。当该身份鉴权装置1被唤醒之后，识别该手机前方的用户是合法的用户时，则解锁屏幕。
可见，由于该电子设备100应用了该身份鉴权装置1，该电子设备1能够实现对待测物体的较远距离的感测，且感测响应速度较快。
在本说明书的描述中，参考术语“一个实施方式”、“某些实施方式”、“示意性实施方式”、“示例”、“具体示例”、或“一些示例”等的描述意指结合所述实施方式或示例描述的具体特征、结构、材料或者特点包含于本申请的至少一个实施方式或示例中。在本说明书中，对上述术语的示意性表述不一定指的是相同的实施方式或示例。而且，描述的具体特征、结构、材料或者特点可以在任何的一个或多个实施方式或示例中以合适的方式结合。
尽管上面已经示出和描述了本申请的实施方式，可以理解的是，上述实施方式是示例性的，不能理解为对本申请的限制，本领域的普通技术人员在本申请的范围内可以对上述实施方式进行变化、修改、替换和变型。

Claims (57)

  1. 一种身份鉴权方法，包括：
    步骤S1：投射红外泛光至一待测物体上，感测该待测物体的第一红外图像；
    步骤S2：根据步骤S1获得的第一红外图像，比对该待测物体的二维图像信息与预存的二维图像信息是否匹配；
    步骤S3：投射红外结构光至该待测物体上，感测该待测物体的第二红外图像；
    步骤S4：根据步骤S3获得的第二红外图像，判断该待测物体是否为立体物体；
    步骤S5：根据步骤S2与步骤S4的执行结果，确认所述待测物体的身份是否合法。
  2. 如权利要求1所述的身份鉴权方法，其特征在于：步骤S1与步骤S3分时进行，其中，步骤S1先于或后于步骤S3进行。
  3. 如权利要求2所述的身份鉴权方法，其特征在于：在步骤S5中，当步骤S2与步骤S4中的任意一个步骤先被执行而获得结果是否定的结果时，则身份鉴权失败。
  4. 如权利要求1所述的身份鉴权方法，其特征在于：在步骤S5中：当步骤S2中确认该待测物体的二维图像信息与预存的二维图像信息匹配、步骤S4中判断得知该待测物体为立体物体时，则身份鉴权成功。
  5. 如权利要求1所述的身份鉴权方法，其特征在于：步骤S1中的红外泛光的波长范围为[920,960]纳米，或/和，步骤S3中的红外结构光的波范围为[920,960]纳米。
  6. 如权利要求1所述的身份鉴权方法，其特征在于：步骤S1中的红外泛光的波长为940纳米、或/和，步骤S3中的红外结构光的波长为940纳米。
  7. 如权利要求1-6中任意一项所述的身份鉴权方法，其特征在于：当所述预存的二维图像信息包括物体的面部图像信息时，在步骤S2中比对的是：该待测物体的二维图像信息与该面部图像信息是否匹配。
  8. 如权利要求7所述的身份鉴权方法，其特征在于：所述物体的面部图像信息为人体的面部图像信息。
  9. 如权利要求8所述的身份鉴权方法，其特征在于：该面部图像信息包括面部特征信息，该步骤S2进一步包括：对该第一红外图像进行特征提取，通过比对提取到的特征信息与预存的面部特征信息，来确认该待测物体的二维图像信息与预存的二维图像信息是否匹配。
  10. 如权利要求9所述的身份鉴权方法，其特征在于：在步骤S2中，通过深度学习方法提取该待测物体的面部特征信息。
  11. 如权利要求10所述的身份鉴权方法，其特征在于：该深度学习方法包括：建立深度卷积神经网络模型，使用预定数量的人脸照片训练该深度卷积神经网络模型，根据训练好的该深度卷积神经网络模型提取人脸的特征参数。
  12. 如权利要求9所述的身份鉴权方法，其特征在于：所述面部特征信息包括鼻子、眼睛、嘴巴、眉毛、额头、颧骨、下巴、脸庞、鼻子的宽度、下巴的宽度中的任意一种或几种，或/和，鼻子、眼睛、嘴巴、眉毛、额头、颧骨、下巴中任意组合的距离信息。
  13. 如权利要求8所述的身份鉴权方法，其特征在于：该步骤S4进一步包括：对第二红外图像进行立体特征信息提取，根据提取到的立体特征信息来判断该待测物体是否为立体物体。
  14. 如权利要求13所述的身份鉴权方法，其特征在于：在步骤S4中，通过深度学习方法提取该待测物体的立体人脸特征信息。
  15. 如权利要求14所述的身份鉴权方法，其特征在于：该深度学习方法包括：建立深度卷积神经网络模型，使用预定数量的人脸照片训练该深度卷积神经网络模型，根据训练好的该深度卷积神经网络模型提取人脸的特征参数。
  16. 如权利要求13所述的身份鉴权方法，其特征在于：该立体特征信息包括鼻子、眼睛、嘴巴、眉毛、额头、颧骨、下巴、脸庞中的任意一种或几种的立体尺寸信息。
  17. 如权利要求1所述的身份鉴权方法，其特征在于：先执行步骤S1，然后同时执行步骤S2和S3，其中，当步骤S2中确认该待测物体的二维图像信息与预存的二维图像信息匹配后，则启动执行步骤S4，否则，当步骤S2中确认该待测物体的二维图像信息与预存的二维图像信息不匹配时，则身份鉴权失败。
  18. 如权利要求1所述的身份鉴权方法，其特征在于：先执行步骤S3，然后同时执行步骤S4和S1，其中，当步骤S4中判断得知该待测物体为立体物体后，则启动执行步骤S2，否则，当步骤S4中判断得知该待测物体不是立体物体时，则身份鉴权失败。
  19. 如权利要求1所述的身份鉴权方法，其特征在于：先执行步骤S1；执行完步骤S1后再执行步骤S2；当步骤S2中确认该待测物体的二维图像信息与预存的二维图像信息匹配时，则启动执行步骤S3，执行完步骤S3后再执行步骤S4；而当步骤S2中确认该待测物体的二维图像信息与预存的二维图像信息不匹配时，则身份鉴权失败；或，
    先执行步骤S3；执行完步骤S3再执行步骤S4；当步骤S4中判断得知该待测物体为立体物体时，则执行步骤S1，执行完步骤S1后再执行步骤S2；而当步骤S4中判断得知该待测物体不是立体物体时，则身份鉴权失败；或，
    步骤S1、步骤S3、步骤S2依次执行；当步骤S2中确认该待测物体的二维图像信息与预存的二维图像信息匹配后，则启动执行步骤S4，否则，当步骤S2中确认该待测物体的二维图像信息与预存的二维图像信息不匹配时，则身份鉴权失败；或，
    步骤S1、步骤S3、步骤S4依次执行；当步骤S4中判断得知该待测物体为立体物体时，则启动执行步骤S2，否则，当步骤S4中判断得知该待测物体不是立体物体时，则身份鉴权失败；或，
    步骤S3、步骤S1、步骤S2依次执行；当步骤S2中确认该待测物体的二维图像信息与预存的二维图像信息匹配后，则启动执行步骤S4，否则，当步骤S2中确认该待测物体的二维图像信息与预存的二维图像信息不匹配时，则身份鉴权失败；或，
    步骤S3、步骤S1、步骤S4依次执行；当步骤S4中判断得知该待测物体为立体物体时，则启动执行步骤S2，否则，当步骤S4中判断得知该待测物体不是立体物体时，则身份鉴权失败。
  20. 如权利要求1所述的身份鉴权方法，其特征在于：在步骤S3中，投射至待测物体的红外结构光形成图案，所述图案呈规则点阵式、条纹式、散斑式、网格式、编码式中的任意一种或几种的结合。
  21. 如权利要求1所述的身份鉴权方法，其特征在于：步骤S2与步骤S4分时或同时进行。
  22. 如权利要求1所述的身份鉴权方法，其特征在于：步骤S4还包括:当判断得知该待测物体为立体物体时，进一步判断该待测物体的立体信息是否符合人的面部立体特征。
  23. 如权利要求22所述的身份鉴权方法，其特征在于：在步骤S5中，当步骤S2与步骤S4中的任意一个步骤先被执行而获得结果是否定的结果时，则身份鉴权失败。
  24. 如权利要求22所述的身份鉴权方法，其特征在于：在步骤S5中，当步骤S2中确认该待测物体的二维图像信息与预存的二维图像信息匹配，步骤S4中判断得知该待测物体的立体信息符合人的面部立体特征时，则确认该待测物体合法，身份鉴权成功。
  25. 一种身份鉴权装置，包括：
    第一投射器，用于投射红外泛光至一待测物体；
    第二投射器，用于投射红外结构光至该待测物体；
    图像传感装置，用于捕获由该待测物体反射回来的红外泛光、感测获得该待测物体的第一红外图像，以及用于捕获由该待测物体反射回来的红外结构光、感测获得该待测物体的第二红外图像；
    处理器，用于根据该第一红外图像，比对所述待测物体的二维图像信息与预存的二维图像信息是否匹配，并获得比对结果；所述处理器还用于根据该第二红外图像，判断所述待测物体是否为立体物体，并获得判断结果；所述处理器根据该比对结果与判断结果，确认所述待测物体的身份是否合法。
  26. 如权利要求25所述的身份鉴权装置，其特征在于：该身份鉴权装置进一步包括控制电路，该控制电路用于控制该第一投射器、该第二投射器、和该图像传感装置协同工作。
  27. 如权利要求26所述的身份鉴权装置，其特征在于：该控制电路用于控制所述第一投射器和所述第二投射器分时工作。
  28. 如权利要求25所述的身份鉴权装置，其特征在于：该处理器用于分时或同时执行“根据该第一红外图像，比对所述待测物体的二维图像信息与预存的二维图像信息是否匹配”与“根据该第二红外图像，判断所述待测物体是否为立体物体”。
  29. 如权利要求28所述的身份鉴权装置，其特征在于：当该处理器分时执行“根据该第一红外图像，比对所述待测物体的二维图像信息与预存的二维图像信息是否匹配”与“根据该第二红外图像，判断所述待测物体是否为立体物体”时，“根据该第一红外图像，比对所述待测物体的二维图像信息与预存的二维图像信息是否匹配”与“根据该第二红外图像，判断所述待测物体是否为立体物体”中的任意一者先被所述处理器执行而获得结果是否定的结果时，则确认该待测物体的身份不合法，身份鉴权失败。
  30. 如权利要求29所述的身份鉴权装置，其特征在于：当所述处理器确认所述待测物体的二维图像信息与预存的二维图像信息匹配、所述待测物体为立体物体时，则身份鉴权成功。
  31. 如权利要求25-30中任意一项所述的身份鉴权装置，其特征在于：该身份鉴权装置进一步包括存储器，该存储器用于预存二维图像信息，当所述预存的二维图像信息包括物体的面部图像信息时，所述处理器用于比对该待测物体的二维图像信息与预存的面部图像信息是否匹配。
  32. 如权利要求31所述的身份鉴权装置，其特征在于：该物体的面部图像信息为人体的面部图像信息。
  33. 如权利要求32所述的身份鉴权装置，其特征在于：预存的面部图像信息包括面部特征信息，所述处理器还用于对该第一红外图像进行特征提取，通过比对提取到的特征信息与预存的面部特征信息，来确认该待测物体的二维图像信息与预存的二维图像信 息是否匹配。
  34. 如权利要求33所述的身份鉴权装置，其特征在于：该处理器通过深度学习方法提取该待测物体的面部特征信息。
  35. 如权利要求34所述的身份鉴权装置，其特征在于：该深度学习方法包括：建立深度卷积神经网络模型，使用预定数量的人脸照片训练该深度卷积神经网络模型，根据训练好的该深度卷积神经网络模型提取人脸的特征参数。
  36. 如权利要求33所述的身份鉴权装置，其特征在于：该面部特征信息包括鼻子、眼睛、嘴巴、眉毛、额头、颧骨、下巴、脸庞、鼻子的宽度、下巴的宽度中的任意一种或几种，或/和，鼻子、眼睛、嘴巴、眉毛、额头、颧骨、下巴中任意组合的距离信息。
  37. 如权利要求32所述的身份鉴权装置，其特征在于：所述处理器还用于对第二红外图像进行立体特征信息提取，根据提取到的立体特征信息来判断该待测物体是否为立体物体。
  38. 如权利要求37所述的身份鉴权装置，其特征在于：该处理器通过深度学习方法提取该待测物体的立体人脸特征信息。
  39. 如权利要求38所述的身份鉴权装置，其特征在于：该深度学习方法包括：建立深度卷积神经网络模型，使用预定数量的人脸照片训练该深度卷积神经网络模型，根据训练好的该深度卷积神经网络模型提取人脸的特征参数。
  40. 如权利要求37所述的身份鉴权装置，其特征在于：该立体特征信息包括鼻子、眼睛、嘴巴、眉毛、额头、颧骨、下巴、脸庞中的任意一种或几种的立体尺寸信息。
  41. 如权利要求27所述的身份鉴权装置，其特征在于：在执行身份鉴权时，所述控制电路控制该第一投射器先投射红外泛光至待测物体，该图像传感装置感测获得该待测物体的第一红外图像，然后在所述处理器根据该第一红外图像比对所述待测物体的二维图像信息与预存的二维图像信息是否匹配的同时：所述控制电路控制该第二投射器投射红外结构光至该待测物体，该图像传感装置感测获得该待测物体的第二红外图像。
  42. 如权利要求41所述的身份鉴权装置，其特征在于：当所述处理器确认所述待测物体的二维图像信息与预存的二维图像信息匹配时，则所述处理器再根据该第二红外图像判断所述待测物体是否为立体物体；当所述处理器确认所述待测物体的二维图像信息与预存的二维图像信息不匹配时，则所述处理器确认该待测物体的身份不合法，身份鉴权失败。
  43. 如权利要求27所述的身份鉴权装置，其特征在于：在执行身份鉴权时，所述控 制电路控制所述第二投射器先投射红外结构光至待测物体，该图像传感装置感测获得该待测物体的第二红外图像，然后在所述处理器根据该第二红外图像判断所述待测物体是否为立体物体的同时：所述控制电路控制所述第一投射器投射红外泛光至该待测物体，所述图像传感装置感测获得该待测物体的第一红外图像。
  44. 如权利要求43所述的身份鉴权装置，其特征在于：当所述处理器判断得知所述待测物体为立体物体时，则所述处理器再根据该第一红外图像比对所述待测物体的二维图像信息与预存的二维图像信息是否匹配；当所述处理器判断得知所述待测物体不是立体物体时，则身份鉴权失败。
  45. 如权利要求25所述的身份鉴权装置，其特征在于：所述处理器用于先根据该第一红外图像，比对该待测物体的二维图像信息与预存的二维图像信息是否匹配，当确认该待测物体的二维图像信息与预存的二维图像信息匹配之后，再根据该第二红外图像，判断该待测物体是否为立体物体。
  46. 如权利要求25所述的身份鉴权装置，其特征在于：所述处理器用于先根据该第二红外图像，判断该待测物体是否为立体物体，当判断得知该待测物体为立体物体之后，再根据该第一红外图像，比对该待测物体的二维图像信息与该预存的二维图像信息是否匹配。
  47. 如权利要求25所述的身份鉴权装置，其特征在于：所述第一投射器投射的红外泛光的波长范围为[920,960]纳米，和/或，所述第二投射器投射的红外结构光的波长范围为[920,960]纳米。
  48. 如权利要求25所述的身份鉴权装置，其特征在于：所述第一投射器投射的红外泛光的波长为940纳米，或/和，所述第二投射器投射的红外结构光的波长为940纳米。
  49. 如权利要求25所述的身份鉴权装置，其特征在于：所述第二投射器投射至待测物体的红外结构光形成图案，所述图案呈点阵式、条纹式、散斑式、网格式、编码式中的任意一种或几种的结合。
  50. 如权利要求25所述的身份鉴权装置，其特征在于：该身份鉴权装置进一步包括高速数据传送链路，用于把图像传感装置中表示该第一红外图像的信号和表示该第二红外图像的信号传送到该处理器中进行处理。
  51. 如权利要求25所述的身份鉴权装置，其特征在于：当判断得知该待测物体为立体物体时，该处理器还用于进一步判断该待测物体的立体信息是否符合人的面部立体特征。
  52. [根据细则91更正 23.05.2018]　
    如权利要求51所述的身份鉴权装置，其特征在于：当该处理器判断得知该待测物体的立体信息不符人的面部立体特征时，则身份鉴权失败。
  53. [根据细则91更正 23.05.2018]　
    如权利要求51所述的身份鉴权装置，其特征在于：当该处理器确认该待测物体的二维图像信息与预存的二维图像信息匹配，且判断得知该待测物体的立体信息符合人的面部立体特征时，则确认该待测物体合法，身份鉴权成功。
  54. [根据细则91更正 23.05.2018]　
    一种电子设备，包括权利要求25-53中任意一项所述的身份鉴权装置。
  55. [根据细则91更正 23.05.2018]　
    如权利要求54所述的电子设备，其特征在于：所述电子设备用于根据所述身份鉴权装置的身份鉴权结果来对应是否执行相应的功能。
  56. [根据细则91更正 23.05.2018]　
    如权利要求55所述的电子设备，其特征在于：所述相应的功能包括解锁、支付、启动预存的应用程序中的任意一种或几种。
  57. [根据细则91更正 23.05.2018]　
    如权利要求54所述的电子设备，其特征在于：所述电子设备包括消费性电子产品、家居式电子产品、车载式电子产品、金融终端产品中的任意一种或几种。
PCT/CN2018/083615 2018-04-18 2018-04-18 身份鉴权方法、身份鉴权装置、和电子设备 WO2019200572A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
CN201880000342.XA CN108566777A (zh) 2018-04-18 2018-04-18 身份鉴权方法、身份鉴权装置、和电子设备
PCT/CN2018/083615 WO2019200572A1 (zh) 2018-04-18 2018-04-18 身份鉴权方法、身份鉴权装置、和电子设备

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2018/083615 WO2019200572A1 (zh) 2018-04-18 2018-04-18 身份鉴权方法、身份鉴权装置、和电子设备

Publications (1)

Publication Number Publication Date
WO2019200572A1 true WO2019200572A1 (zh) 2019-10-24

Family

ID=63555980

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2018/083615 WO2019200572A1 (zh) 2018-04-18 2018-04-18 身份鉴权方法、身份鉴权装置、和电子设备

Country Status (2)

Country Link
CN (1) CN108566777A (zh)
WO (1) WO2019200572A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112861583A (zh) * 2019-11-27 2021-05-28 深圳市万普拉斯科技有限公司 人脸验证方法、电子设备及可读存储介质

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109740568A (zh) * 2019-01-21 2019-05-10 江西阳光安全设备集团有限公司 带有自动挡板的智能型密集架
CN109816907A (zh) * 2019-01-21 2019-05-28 江西阳光安全设备集团有限公司 智能型密集架的自动开关装置
CN109801462A (zh) * 2019-01-21 2019-05-24 江西阳光安全设备集团有限公司 智能型密集架的防盗报警装置

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20130236066A1 (en) * 2012-03-06 2013-09-12 Gary David Shubinsky Biometric identification, authentication and verification using near-infrared structured illumination combined with 3d imaging of the human ear
CN105574518A (zh) * 2016-01-25 2016-05-11 北京天诚盛业科技有限公司 人脸活体检测的方法和装置
CN106815914A (zh) * 2017-01-25 2017-06-09 辛明江 一种基于人脸识别技术的门禁系统及解锁方法
CN107105217A (zh) * 2017-04-17 2017-08-29 深圳奥比中光科技有限公司 多模式深度计算处理器以及3d图像设备

Family Cites Families (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN2828935Y (zh) * 2005-03-22 2006-10-18 北京数字奥森识别科技有限公司 一种用于人脸识别的图像获取识别装置
CN202887214U (zh) * 2012-11-13 2013-04-17 成都迪特福科技有限公司 基于人脸识别的人体身份识别装置
CN105740778B (zh) * 2016-01-25 2020-01-03 北京眼神智能科技有限公司 一种改进的三维人脸活体检测方法及其装置
CN105956518A (zh) * 2016-04-21 2016-09-21 腾讯科技（深圳）有限公司 一种人脸识别方法、装置和系统
CN106991377B (zh) * 2017-03-09 2020-06-05 Oppo广东移动通信有限公司 结合深度信息的人脸识别方法、人脸识别装置和电子装置
CN107633165B (zh) * 2017-10-26 2021-11-19 奥比中光科技集团股份有限公司 3d人脸身份认证方法与装置

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20130236066A1 (en) * 2012-03-06 2013-09-12 Gary David Shubinsky Biometric identification, authentication and verification using near-infrared structured illumination combined with 3d imaging of the human ear
CN105574518A (zh) * 2016-01-25 2016-05-11 北京天诚盛业科技有限公司 人脸活体检测的方法和装置
CN106815914A (zh) * 2017-01-25 2017-06-09 辛明江 一种基于人脸识别技术的门禁系统及解锁方法
CN107105217A (zh) * 2017-04-17 2017-08-29 深圳奥比中光科技有限公司 多模式深度计算处理器以及3d图像设备

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112861583A (zh) * 2019-11-27 2021-05-28 深圳市万普拉斯科技有限公司 人脸验证方法、电子设备及可读存储介质

Also Published As

Publication number Publication date
CN108566777A (zh) 2018-09-21

Similar Documents

Publication Publication Date Title
WO2019200574A1 (zh) 身份鉴权方法、身份鉴权装置、和电子设备
US11551481B2 (en) Living body detection method and apparatus, electronic device, storage medium, and related system to which living body detection method is applied
WO2019200572A1 (zh) 身份鉴权方法、身份鉴权装置、和电子设备
US9607138B1 (en) User authentication and verification through video analysis
CN104298909B (zh) 电子装置、身份验证系统及方法
US9158904B1 (en) Facial recognition
EP2680191B1 (en) Facial recognition
EP2680192B1 (en) Facial recognition
AU2022203880B2 (en) Methods and systems for determining user liveness and verifying user identities
US10210380B2 (en) Methods and systems for enhancing user liveness detection
US11115408B2 (en) Methods and systems for determining user liveness and verifying user identities
KR20150079583A (ko) 카메라 장착 디바이스들과 관련된 생체인식 인증
US11869272B2 (en) Liveness test method and apparatus and biometric authentication method and apparatus
EP3761222B1 (en) Living body detection method and apparatus, electronic device, storage medium and related system using living body detection method
KR20160006772A (ko) 액세스 제어되는 환경으로의 액세스를 인가하는 시스템 및 방법
WO2019200573A1 (zh) 身份鉴权方法、身份鉴权装置、和电子设备
US20160274670A1 (en) Gesture input apparatus, gesture input method, and program for wearable terminal
US20180046871A1 (en) Methods and systems for enhancing user liveness detection
WO2019200576A1 (zh) 身份鉴权方法、身份鉴权装置、和电子设备
Sharp et al. Authentication for drone delivery through a novel way of using face biometrics
KR20130133676A (ko) 카메라를 통한 얼굴인식을 이용한 사용자 인증 방법 및 장치
Aleluya et al. Faceture ID: face and hand gesture multi-factor authentication using deep learning
WO2019200571A1 (zh) 身份鉴权方法、身份鉴权装置、和电子设备
WO2019200575A1 (zh) 身份鉴权方法、身份鉴权装置、和电子设备
US20200401827A1 (en) Spoof detection by estimating subject motion from captured image frames

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 18915692

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 18915692

Country of ref document: EP

Kind code of ref document: A1