WO2021204267A1 - 身份识别 - Google Patents

身份识别 Download PDF

Info

Publication number
WO2021204267A1
WO2021204267A1 PCT/CN2021/086266 CN2021086266W WO2021204267A1 WO 2021204267 A1 WO2021204267 A1 WO 2021204267A1 CN 2021086266 W CN2021086266 W CN 2021086266W WO 2021204267 A1 WO2021204267 A1 WO 2021204267A1
Authority
WO
WIPO (PCT)
Prior art keywords
target
image
imaging device
recognized
depth information
Prior art date
Application number
PCT/CN2021/086266
Other languages
English (en)
French (fr)
Inventor
李志荣
梁明杰
王浦林
刘源
Original Assignee
支付宝(杭州)信息技术有限公司
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 支付宝(杭州)信息技术有限公司 filed Critical 支付宝(杭州)信息技术有限公司
Publication of WO2021204267A1 publication Critical patent/WO2021204267A1/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
    • G06V40/161—Detection; Localisation; Normalisation
    • 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
    • G06V40/12—Fingerprints or palmprints
    • G06V40/13—Sensors therefor
    • 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

Definitions

  • This specification relates to the field of identity recognition technology, and in particular to methods, systems and devices for identity recognition based on in-depth information.
  • biometric identification ie, biometric identification
  • recognition of biological individuals based on their physical signs for example, fingerprints
  • identity verification identification
  • to identify a biological individual based on the physical signs of the biological individual requires the collection of image data of the biological individual in order to perform identity recognition based on the characteristics of the biological individual in the image data.
  • Image data collection and the quality of image data will greatly affect the speed and accuracy of identification.
  • One of the embodiments of this specification provides an identity-based identification method, the method includes: acquiring a first image acquired by a first imaging device, the first image includes one or more candidate targets; acquiring a second imaging device A second image collected by the device, where the second image includes depth information of at least one of the one or more candidate targets; based on the first image and the second image, from the second image Extracting the depth information of the one or more candidate targets in, determining at least one candidate target from the one or more candidate targets as the target to be identified based on the depth information of the one or more candidate targets; The depth information of at least a part of the target to be recognized acquires a third image collected by a third imaging device, where the third image includes the at least part of the target to be recognized; and based on the third image, The identification of the target to be identified is performed.
  • the device includes a candidate target image acquisition module, a depth information extraction module, a to-be-identified target determination module, and a to-be-identified target image acquisition module.
  • the candidate target image acquisition module is configured to acquire a first image, the first image includes one or more candidate targets; and to acquire a second image captured by a second imaging device, the second image includes the Depth information of at least one of the one or more candidate targets.
  • the depth information extraction module is configured to extract the depth information of the one or more candidate targets from the second image based on the first image and the second image.
  • the to-be-recognized target determination module is configured to determine at least one candidate target as the to-be-recognized target from the one or more candidate targets based on the depth information of the one or more candidate targets.
  • the image acquisition module of the target to be recognized is configured to acquire a third image collected by a third imaging device based on the depth information of at least a part of the target to be recognized, and the third image includes the at least Part.
  • the device also includes an identification module, which is used to identify the target to be identified based on the third image.
  • the system includes a first imaging device, a second imaging device, and a third imaging device.
  • the first imaging device is configured to collect a first image, and the first image includes one or more candidate targets.
  • the second imaging device is configured to acquire a second image, the second image including depth information of at least one candidate target among the one or more candidate targets.
  • the third imaging device is configured to acquire a third image, the third image including at least a part of at least one candidate target among the one or more candidate targets.
  • the system further includes a processor and a storage medium, where the storage medium is used to store executable instructions, and the processor is used to execute the executable instructions to implement the above-mentioned identification method.
  • One of the embodiments of this specification provides a computer-readable medium, the storage medium stores computer instructions, and when the computer instructions are executed by a processor, the above-mentioned identity recognition method is realized.
  • Fig. 1 is a schematic diagram of an application scenario of an identity recognition system according to some embodiments of this specification
  • Fig. 2 is an exemplary flowchart of an identity recognition method according to some embodiments of this specification
  • Fig. 3 is an exemplary flow chart of another identity recognition method according to some embodiments of the present specification.
  • Fig. 4 is an exemplary flowchart of another identity recognition method according to some embodiments of the present specification.
  • Fig. 5 is an exemplary module diagram of an identity recognition device according to some embodiments of the present specification.
  • system is a method for distinguishing different components, elements, parts, parts, or assemblies of different levels.
  • the words can be replaced by other expressions.
  • Fig. 1 is a schematic diagram of an application scenario of an identity recognition system according to some embodiments of this specification.
  • the identity recognition system 100 can recognize the identity information of the target to be recognized.
  • the identity recognition system 100 may include a processing device 110, an imaging device 120, a terminal 130, a storage device 140, and a network 150.
  • the processing device 110 may process data and/or information from at least one other component of the identity recognition system 100.
  • the processing device 110 may acquire image data from the imaging device 120.
  • the processing device 110 may extract depth information of a candidate target (for example, a human face) based on image data and determine the target to be recognized based on the depth information of the candidate target.
  • the processing device 110 may perform identity recognition on the target to be recognized based on at least a part of the image data (for example, an iris image) of the target to be recognized.
  • the processing device 110 may be a single processing device or a group of processing devices.
  • the processing device group may be a centralized processing device group connected to the network 150 via an access point, or a distributed processing device group respectively connected to the network 150 via at least one access point.
  • the processing device 110 may be locally connected to the network 150 or remotely connected to the network 150.
  • the processing device 110 may access information and/or data stored in the terminal 130 and/or the storage device 140 via the network 150.
  • the storage device 140 may be used as a back-end data storage of the processing device 110.
  • the processing device 110 may be implemented on a cloud platform.
  • the cloud platform may include private cloud, public cloud, hybrid cloud, community cloud, distributed cloud, internal cloud, multi-layer cloud, etc., or any combination thereof.
  • the processing device 110 may include a processing device.
  • the processing device can process information and/or data related to at least one function described in this specification.
  • the processing device may include at least one processing unit (for example, a single-core processing device or a multi-core processing device).
  • processing equipment includes central processing unit (CPU), application specific integrated circuit (ASIC), application specific instruction set processor (ASIP), graphics processing unit (GPU), physical processing unit (PPU), digital signal processor ( DSP), field programmable gate array (FPGA), programmable logic device (PLD), controller, microcontroller unit, reduced instruction set computer (RISC), microprocessor, etc., or any combination thereof.
  • the imaging device 120 may include multiple types of imaging devices having an image capturing function, for example, a first imaging device 120-1, a second imaging device 120-2, a third imaging device 120-3, and the like.
  • the first imaging device 120-1 may be used to collect planar images.
  • the first imaging device 120-1 may include one or any combination of a color camera, a digital camera, a camcorder, a PC camera, a web camera, a closed circuit television (CCTV), a PTZ camera, a video sensor device, etc.
  • the second imaging device 120-2 may be used to acquire a depth image.
  • the second imaging device 120-2 may include a structured light depth camera, a binocular stereo vision camera, a time-of-flight TOF camera, and the like.
  • the third imaging device 120-3 may be used to collect infrared images (for example, iris images).
  • the third imaging device 120-3 may include an infrared thermal imager, an infrared camera, and the like.
  • the field of view (FOV) of the first imaging device 120-1 and the field of view (FOV) of the second imaging device 120-2 overlap at least partially.
  • the field of view (FOV) of the second imaging device 120-2 and the field of view (FOV) of the third imaging device 120-3 overlap at least partially.
  • the first imaging device 120-1, the second imaging device 120-2, the third imaging device 120-3, etc. may be integrated in the same device.
  • the first imaging device 120-1, the second imaging device 120-2, the third imaging device 120-3, etc. may be different imaging modules in the same device.
  • the imaging device 120 may collect images containing candidate targets, and send the collected images to one or more devices in the identity recognition system 100.
  • the imaging device 120 may collect images containing multiple human faces, and send the images to the processing device 110 through the network 150 for subsequent processing.
  • the terminal 130 may communicate and/or connect with the processing device 110, the imaging device 120, and/or the storage device 140.
  • the terminal 130 may obtain image data acquired through the imaging device 120, and send the image data to the processing device 110 for processing.
  • the terminal 130 may obtain the result of identity recognition from the processing device 110.
  • the terminal 130 may include a mobile device, a tablet computer, a laptop computer, etc., or any combination thereof.
  • the user can interact with other components in the identity recognition system 100 through the terminal 130.
  • the user can view the image collected by the imaging device through the terminal 130.
  • the user can also view the identification result determined by the processing device 110 through the terminal 130.
  • the storage device 140 may store data and/or instructions.
  • the storage device 140 may store the image data collected by the imaging device 120, the coordinate system conversion relationship between the imaging devices, the identity information of the target to be recognized, the image processing model and/or algorithm, and the like.
  • the storage device 140 may store data and/or instructions that can be executed by the processing device 110, and the processing device 110 may execute or use the data and/or instructions to implement the exemplary methods described in this specification.
  • the storage device 140 may include mass storage, removable storage, volatile read-write storage, read-only storage (ROM), etc., or any combination thereof.
  • Exemplary mass storage devices may include magnetic disks, optical disks, solid state disks, and the like.
  • Exemplary removable storage may include flash drives, floppy disks, optical disks, memory cards, compact disks, magnetic tapes, and the like.
  • An exemplary volatile read-write memory may include random access memory (RAM).
  • Exemplary random access memory may include dynamic random access memory (DRAM), double data rate synchronous dynamic random access memory (DDRSDRAM), static random access memory (SRAM), thyristor random access memory (T-RAM) And zero capacitance random access memory (Z-RAM), etc.
  • Exemplary read-only memory may include mask-type read-only memory (MROM), programmable read-only memory (PROM), erasable programmable read-only memory (PEROM), electrically erasable programmable read-only memory (EEPROM), CD-ROM and digital versatile disk read-only memory, etc.
  • the storage device 140 may be implemented on a cloud platform.
  • the cloud platform may include a private cloud, a public cloud, a hybrid cloud, a community cloud, a distributed cloud, an internal cloud, a multi-layer cloud, etc., or any combination thereof.
  • the network 150 may facilitate the exchange of information and/or data.
  • at least one component for example, the processing device 110, the imaging device 120, the terminal 130, and the storage device 140
  • the processing device 110 may send information and/or data to other components via the network 150.
  • the processing device 110 may acquire an image from the imaging device 120 through the network 150.
  • the processing device 110 may send the acquired image to the terminal 130 via the network 150.
  • the processing device 110 may use the network 150 to obtain the identity information of multiple objects (for example, biological individuals) from the storage device 140.
  • the processing device 110 may send the processed image to the terminal 130 via the network 150.
  • the network 150 may be any form of wired or wireless network, or any combination thereof.
  • the network 150 may include a cable network, a wired network, an optical fiber network, a telecommunication network, an internal network, the Internet, a local area network (LAN), a wide area network (WAN), a wireless local area network (WLAN), a metropolitan area network (MAN), public switched telephone network (PSTN), Bluetooth network, ZigBee network, near field communication (NFC) network, etc. or any combination thereof.
  • the network 150 may include at least one network access point.
  • the network 150 may include wired or wireless network access points, such as base stations and/or Internet exchange points. Through these network access points, at least one component of the identification system 100 may be connected to the network 150 to exchange data and/or information. .
  • Fig. 2 is an exemplary flowchart of an identity recognition method according to some embodiments of the present specification.
  • the process 200 may be implemented by the identity recognition apparatus 500 or the processing device 110 shown in FIG. 1.
  • the process 200 may be stored in a storage device (such as the storage device 140) in the form of a program or instruction, and the program or instruction may implement the process 200 when the program or instruction is executed.
  • the process 200 may include the following steps.
  • Step 201 The processing device may acquire a first image collected by a first imaging device, where the first image includes one or more candidate targets. In some embodiments, this step may be performed by the candidate target image acquisition module 501.
  • the candidate target may also be referred to as a candidate to-be-recognized object.
  • the candidate target may include all or part of the biological individual.
  • the candidate target may include a human body or a face of a human body (ie, a human face).
  • the fact that the first image includes one or more candidate targets may mean that the first image includes an image representation of the candidate targets.
  • the image representation of the candidate target can also be referred to as the image description of the candidate target.
  • the first image may include feature information used to represent or describe the candidate target, for example, texture features, boundary features, color features, and so on.
  • the first image may be a two-dimensional image.
  • the image type of the first image includes at least one of the following: grayscale image, RGB image, etc., or any combination thereof.
  • the first imaging device may include an imaging device for acquiring a planar image (eg, a first image) of the candidate target.
  • the first imaging device may include, but is not limited to, one or any combination of a color camera, a digital camera, a camcorder, a PC camera, a network camera, a closed-circuit television (CCTV), a PTZ camera, a video sensor device, etc.
  • the processing device may acquire the first image from the first imaging device, the storage device 140, or other storage devices.
  • the processing device may obtain a single first image captured by a first imaging device.
  • the processing device may acquire multiple first images simultaneously acquired by multiple first imaging devices. At least one of the plurality of first images includes one or more candidate targets. For example, at least one of the multiple first images includes multiple human faces.
  • the candidate target image acquisition module 501 may acquire the first image from the first imaging device and store the first image in the storage device 140.
  • Step 203 The processing device may acquire a second image collected by the second imaging device, where the second image includes depth information of at least one of the candidate targets. In some embodiments, this step may be performed by the candidate target image acquisition module 501.
  • the depth information of the candidate target may indicate the distance between a point on the surface of the candidate target (for example, a human face) and the second imaging device.
  • the size of the pixel value of the second image may indicate the size of the distance between the surface of the candidate target and the second imaging device.
  • the second image may include a depth image.
  • the second image may include a point cloud image.
  • the depth image can be converted into a point cloud image to obtain the second image.
  • the point cloud image can be converted into a depth map to obtain the second image.
  • the second image may be a two-dimensional image, a three-dimensional image, or the like.
  • the second imaging device includes an imaging device that can collect depth information of a candidate target (for example, a human face).
  • the second imaging device may include one or more depth imaging devices.
  • the depth imaging device may include, but is not limited to: a structured light (Structured Light) depth camera, a binocular stereo vision (Binocular Stereo Vision) camera, a TOF (Time of flight) camera, etc., or any combination thereof.
  • the field of view (FOV) of the second imaging device and the field of view (FOV) of the first imaging device at least partially overlap.
  • the first imaging device and the second imaging device simultaneously capture the first image and the second image.
  • the candidate target image acquisition module 501 may acquire the second image from the second imaging device, the storage device 140 or other storage devices. In some embodiments, the candidate target image acquisition module 501 can acquire a single second image acquired by a second imaging device. In some embodiments, the candidate target image acquisition module 501 may acquire multiple second images simultaneously acquired by multiple second imaging devices. At least one second image among the plurality of second images includes depth information of at least one candidate target among the one or more candidate targets. Each second image can correspond to a first image. As described herein, the second image corresponding to the first image refers to a pixel in the second image corresponding to the same position or the same part of a certain pixel in the first image on the candidate target.
  • Step 205 The processing device may extract depth information of the one or more candidate targets based on the first image and the second image. In some embodiments, this step may be performed by the depth information extraction module 503.
  • the first imaging device and the second imaging device may be calibrated based on the same coordinate system (for example, the world coordinate system) before acquiring the first image and the second image, so that the first imaging device and the second imaging device
  • the equipment has a unified coordinate system.
  • the processing device may directly detect one or more candidate targets (for example, human faces) from the first image.
  • the processing device may extract the depth information of one or more candidate targets (for example, a human face) from the second image based on the position of the detected one or more candidate targets in the first image.
  • the processing device may register the first image with the second image to obtain the registration result.
  • the processing device may detect one or more candidate targets (for example, human faces) from the registered first image.
  • the depth information extraction module 503 may extract depth information of one or more candidate targets (for example, a human face) based on the detected one or more candidate targets and the registration result.
  • the processing device may register the first image with the second image through image registration technology.
  • Exemplary image registration techniques may include gray-scale and template-based matching algorithms, feature-based matching algorithms, domain transform-based algorithms, and the like.
  • the processing device may use methods such as image segmentation technology and model-based target detection technology to detect candidate targets from the registered first image.
  • Image segmentation techniques may include the use of edge-based segmentation algorithms, threshold-based segmentation algorithms, region-based segmentation algorithms, morphological watershed algorithms, etc., or a combination thereof.
  • Model-based target detection technology may include the use of machine learning models (R-CNN model, Fast RCNN model, SVM model, etc.) for target detection.
  • the processing device may perform mask processing on the registered first image based on the detected candidate target to obtain a mask image. For example, the pixel value of the region where the candidate target is located in the registered first image can be set to 1, and the pixel value of the remaining regions can be set to 0.
  • the depth information of one or more candidate targets may be extracted from the registered second image based on the position of the detected candidate target in the registered first image.
  • the position of the detected candidate target in the registered second image may be determined based on the position of the detected candidate target in the registered first image.
  • the mask image with the detected candidate target may be multiplied with the registered second image to determine the position of the detected candidate target in the registered second image.
  • the depth information of the position of the detected candidate target in the registered second image may be extracted from the registered second image.
  • Step 207 The processing device may determine at least one candidate target from the one or more candidate targets as the target to be identified based on the depth information of the one or more candidate targets. In some embodiments, this step may be performed by the to-be-identified target determination module 505.
  • the spatial position relationship of one or more candidate targets may be determined based on the depth information of one or more candidate targets (for example, human faces), and based on the depth information of one or more candidate targets.
  • the spatial position relationship determines the target to be identified.
  • the spatial position relationship of the candidate target may include the spatial position relationship between the candidate target and the second imaging device.
  • the spatial position relationship of the candidate target may be expressed as the distance between the candidate target and the second imaging device.
  • the processing device may determine at least one candidate target among the one or more candidate targets as the target to be recognized based on the distance between the candidate target (for example, a human face) and the second imaging device.
  • the candidate target for example, a human face
  • the candidate target can be determined as the target to be recognized.
  • the imaging device may set the distance to the second imaging device to be less than a certain threshold (for example, less than 1 meter, or less than 2 meters, or less than 3 meters, or less than 4 meters, etc.) or the distance of the second imaging device within a certain range
  • a certain threshold for example, less than 1 meter, or less than 2 meters, or less than 3 meters, or less than 4 meters, etc.
  • Candidate targets within are determined as targets to be identified.
  • the processing device may determine the candidate target with the smallest distance from the second imaging device as the target to be recognized.
  • the processing device may determine the target to be recognized from the two or more candidate targets based on a certain criterion. For example, the processing device may determine the target to be recognized based on the position in the first image of the candidate target having the same distance from the second imaging device. Further, the to-be-recognized target determining module 505 may determine the candidate target close to the left or right of the image in the first image as the to-be-recognized target.
  • Step 209 The processing device acquires a third image collected by a third imaging device based on the depth information of at least a part of the target to be recognized, the third image including the at least part of the target to be recognized. This step can be performed by the to-be-identified target image acquisition module 507.
  • the target to be recognized may include a human face, and at least a part of the target to be recognized may include at least one of a human eye, an iris, an eye pattern, and an eye circumference.
  • the third image including the at least part of the target to be recognized may also be referred to as an image representation of the third image including the at least part (for example, human eyes) of the target to be recognized.
  • the processing device may acquire one or more fourth images (for example, human eye images) acquired by the third imaging device.
  • the processing device may determine the third image from one or more fourth images based on depth information of at least a part of the target to be recognized (for example, human eyes).
  • the coordinate system conversion relationship between the third imaging device and the second imaging device ie, geometric mapping relationship or spatial projection relationship
  • the projection relationship between the fourth image and the depth information of at least a part of the target to be recognized satisfies the coordinate system conversion relationship between the third imaging device and the second imaging device as the third image.
  • the processing device may use the coordinate system conversion relationship between the third imaging device and the second imaging device to project the depth information of at least a part of the target to be recognized (for example, human eyes) onto the plane where each fourth image is located.
  • the processing device may designate a fourth image matching the projected information among the fourth images as the third image.
  • the coordinate system conversion relationship between the third imaging device and the second imaging device is related to the calibration parameters of the second imaging device and the third imaging, and is the default setting of the identity recognition system 100.
  • the processing device may locate at least a part of the target to be recognized (for example, human eyes) based on the depth information of at least a part of the target to be recognized (for example, human eyes) to determine at least a part of the target to be recognized (for example, Human eye) relative to the spatial position information of the third imaging device.
  • the distance and direction of at least a part of the target to be recognized (for example, the human eye) from the second imaging device may be determined based on the depth information of at least a part of the target to be recognized (for example, the human eye).
  • it is possible to determine at least a part of the target to be recognized from the third imaging device.
  • the third imaging device may image at least a part of the target to be recognized (for example, human eyes) based on the distance and direction of at least a part of the target to be recognized (for example, human eyes) and the third imaging device to acquire the third image.
  • the spatial position information of the target to be recognized relative to the third imaging device may be determined based on the depth information of at least a part of the target to be recognized.
  • the third imaging device may be activated to focus on at least a part of the target to be recognized based on the spatial position information of the target to be recognized relative to the third imaging device to acquire the third image. Refer to FIG. 4 for more description of the autofocus of the third device.
  • the processing device may determine whether the depth information of at least a part of the target to be recognized meets a certain condition, and determine whether to activate the third imaging device to acquire the third image based on the determination result. For more description about the activation of the third imaging device, refer to FIG. 3.
  • the third imaging device includes one or more infrared imaging devices.
  • the third device may include one or more image sensors, for example, CMOS image sensors, CCD image sensors, and so on.
  • at least one of the vertical viewing angle (FOV) or the horizontal viewing angle (FOV) of the third imaging device is greater than a threshold value or within a certain range.
  • the vertical FOV of the third imaging device can be within a certain range, for example, at 0-60 degrees, or at 0-90 degrees, or at 0-90 degrees. 120 degrees and other ranges.
  • multiple image sensors can be installed in the horizontal direction of the third imaging device, so that the horizontal FOV of the third imaging device is within a certain range, for example, at 0-60 degrees, or at 0-90 degrees, or at 0 degrees. -120 degrees and other ranges.
  • multiple image sensors can be installed in the vertical and horizontal directions of the third imaging device at the same time, so that the vertical FOV and horizontal FOV of the third imaging device are within a certain range, for example, at 0-60 degrees, or at 0-degrees. 90 degrees, or in the range of 0-120 degrees.
  • the image sensor in the third device can rotate along multiple degrees of freedom, for example, can rotate clockwise or counterclockwise.
  • the FOV in a certain direction can be changed by rotating the image sensor in the third device. For example, if the horizontal FOV is larger, the image sensor in the third device can be rotated 90 degrees to make the vertical FOV larger.
  • Step 211 The processing device may perform identity recognition on the target to be recognized based on the third image. This step may be performed by the identification module 509.
  • images of at least a part of a plurality of objects may be collected in advance and image features may be extracted.
  • the pre-extracted image features can be stored in the storage device 140 in the form of feature codes, or can be directly stored in an external database.
  • the recognition module 509 can use a feature extraction algorithm to extract features from the third image.
  • the recognition module 509 may preprocess the third image before performing feature extraction, for example, image smoothing, edge detection, image separation, and so on.
  • the recognition module 509 may further perform feature encoding on the features extracted from the third image.
  • the recognition module 509 can match the feature code obtained from the third image with the pre-stored feature code to perform identity recognition on the target to be recognized.
  • the target to be recognized may be determined based on the depth information of the candidate target.
  • the third image of the specified target may be filtered out of the images taken by the current third imaging device (for example, iris camera) based on the depth information of at least a part of the target to be recognized (for example, human eyes) for identity recognition (for example, Iris recognition), which can avoid false collection or collection of objects that should not be collected, improve the quality and efficiency of collected images, and further improve the speed and accuracy of identity recognition.
  • identity recognition for example, Iris recognition
  • the vertical or horizontal FOV for example, you can Obtain a larger vertical FOV in order to cover people of different heights, so as to achieve no need to increase the mechanical structure of the pitch angle adjustment.
  • the second imaging device can adopt a structured light depth camera or a TOF depth camera, which can effectively reduce the dependence on ambient light, and can improve the accuracy of depth information, thereby improving the accuracy of determining the target to be recognized ,
  • a structured light depth camera or a TOF depth camera which can effectively reduce the dependence on ambient light, and can improve the accuracy of depth information, thereby improving the accuracy of determining the target to be recognized .
  • the third image for example, iris image
  • identity recognition for example, iris recognition
  • Fig. 3 is an exemplary flowchart of another identity recognition method according to some embodiments of the present specification.
  • the process 300 may be implemented by the identity recognition apparatus 500 or the processing device 110 shown in FIG. 1.
  • the process 300 may be stored in a storage device (such as the storage device 140) in the form of a program or instruction, and when the program or instruction is executed, the process 300 may be implemented.
  • the process 300 may include the following steps.
  • the third imaging device may simultaneously acquire images with the first imaging device and the second imaging device. In some embodiments, the third imaging device may initiate image acquisition based on corresponding conditions. For example, the third imaging device may determine whether to start operations such as the acquisition of the third image based on the results of the first and second images acquired by the first imaging device and the second imaging device. For example, the candidate target image acquisition module 501 may acquire a first image acquired by a first imaging device and a second image acquired by a second imaging device.
  • the first and second images include image representations of one or more candidate targets and at least one
  • the depth information of the candidate target can be determined based on the result of whether the depth information of the target to be identified in the second image meets the corresponding condition, whether to start the third imaging device to collect the third image.
  • Step 301 The processing device may acquire a first image collected by the first imaging device, where the first image includes image representations of one or more candidate targets. In some embodiments, this step may be performed by the candidate target image acquisition module 501.
  • step 201 in the process 200 For a detailed description of acquiring the first image, reference may be made to step 201 in the process 200.
  • Step 303 The processing device may acquire a second image collected by the second imaging device, where the second image includes depth information of at least one candidate target among the one or more candidate targets. In some embodiments, this step may be performed by the candidate target image acquisition module 501.
  • step 203 in the process 200 For a detailed description of acquiring the second image, refer to step 203 in the process 200.
  • Step 305 The processing device may extract depth information of the one or more candidate targets from the second image based on the first image and the second image. In some embodiments, this step may be performed by the depth information extraction module 503.
  • step 205 For a specific description of extracting the depth information of one or more candidate targets from the second image, reference may be made to step 205 in the process 200.
  • Step 307 The processing device may determine at least one candidate target from the one or more candidate targets as the target to be identified based on the depth information of the one or more candidate targets. In some embodiments, this step may be performed by the to-be-identified target determination module 505.
  • step 207 For a specific description of determining at least one candidate target as the target to be recognized based on the depth information, reference may be made to step 207 in the process 200.
  • Step 309 The processing device may determine whether the depth information of the at least a part of the target to be identified meets a condition. In some embodiments, this step may be performed by the to-be-identified target image acquisition module 507 (for example, the activation unit (not shown)).
  • the depth information of at least a part (for example, human eyes, eye patterns, eye circumference, etc.) of the target to be recognized may indicate the difference between the point on the surface of the target to be recognized and the second imaging device.
  • the distance relationship between may be based on the distance between a point on the surface of at least a part of the target to be recognized (for example, human eyes, eye patterns, eye circumference, etc.) and the second imaging device, and the distance between the second imaging device and the third imaging device.
  • the spatial position relationship (for example, direction, distance, etc.) between the devices determines the distance between a point on the surface of at least a part of the target to be recognized (for example, human eyes, eye patterns, eye circumference, etc.) and the third imaging device. In some embodiments, it may be based on the distance between a point on the surface of at least a part of the target to be recognized (for example, human eyes, eye patterns, eye circumference, etc.) and the second imaging device, and the geographic coordinate system of the second imaging device.
  • the coordinate system conversion relationship between them determines the position of the point on the surface of at least a part of the target to be recognized (for example, human eyes, eye patterns, eye circumference, etc.) in the geographic coordinate system.
  • the to-be-recognized target may be determined based on the position of the point on the surface of at least a part of the target to be recognized (for example, human eyes, eye patterns, eye circumference, etc.) in the geographic coordinate system and the position of the third imaging device in the geographic coordinate system.
  • determining whether the depth information of at least a part of the target to be recognized satisfies the condition includes determining whether the distance between at least a part of the target to be recognized and the third imaging device satisfies a certain condition. For example, it can be determined whether the distance between at least a part of the target to be recognized and the third imaging device is within a certain distance range. If the distance between at least a part of the target to be recognized and the third imaging device is within a certain distance range, it can be determined that the depth information of at least a part of the target to be recognized satisfies the condition.
  • the distance between at least a part of the target to be recognized and the third imaging device is not within a certain distance range, it may be determined that the depth information of at least a part of the target to be recognized does not satisfy the condition.
  • the distance range can include 30-70cm, 20-80cm, 10-90cm, and so on.
  • the distance threshold may include 70cm, 80cm, 90cm, and so on.
  • the points on at least a part of the upper surface of the target to be recognized may not be on the same plane, that is, the distance between the points on at least a part of the surface of the target to be recognized and the third imaging device may be different.
  • the distance between at least a part of the target to be recognized and the third imaging device may be determined based on the distance between a point on the surface of at least a part of the target to be recognized and the third imaging device. For example, it may be determined that the average value of the distance between a point on the surface of at least a part of the target to be recognized and the third imaging device is the distance between at least a part of the target to be recognized and the third imaging device. For another example, it may be determined that the median value of the distance between a point on the surface of at least a part of the target to be recognized and the third imaging device is the distance between at least a part of the target to be recognized and the third imaging device.
  • the processing device may, in response to the depth information of at least a part of the target to be recognized satisfying the condition, start the third imaging device to collect a third image of at least a part of the target to be recognized.
  • this step may be performed by the to-be-identified target image acquisition module 507 (for example, the activation unit (not shown)).
  • one or more fourth images collected by the third imaging device may be activated.
  • the processing device for example, a screening unit (not shown)
  • the coordinate system conversion relationship between the third imaging device and the second imaging device ie, geometric mapping relationship or spatial projection relationship
  • the projection relationship between the fourth image and the depth information of at least a part of the target to be recognized satisfies the coordinate system conversion relationship between the third imaging device and the second imaging device as the third image.
  • the processing device may use the coordinate system conversion relationship between the third imaging device and the second imaging device to project the depth information of at least a part of the target to be recognized (for example, human eyes) onto the plane where each fourth image is located.
  • the processing device may designate a fourth image matching the projected information among the fourth images as the third image.
  • the coordinate system conversion relationship between the third imaging device and the second imaging device is related to the calibration parameters of the second imaging device and the third imaging, and is the default setting of the identity recognition system 100.
  • the processing device may locate at least a part of the target to be recognized (for example, human eyes) based on the depth information of at least a part of the target to be recognized (for example, human eyes) to determine at least a part of the target to be recognized (for example, Human eye) relative to the spatial position information of the third imaging device.
  • the distance and direction of at least a part (for example, human eyes) of the target to be recognized from the second imaging device may be determined based on the depth information of at least a part of the target to be recognized (for example, human eyes).
  • it is possible to determine at least a part of the target to be recognized for example, the distance and direction between the human eye
  • the third imaging device may locate at least a part of the target to be recognized (for example, human eyes) based on the depth information of at least a part of the target to be recognized (for example, human eyes) to determine at least a part of the target to be recognized (for example, Human eye) relative to the spatial position information of the third imaging device.
  • the third imaging device may be activated based on the distance and direction between at least a part of the target to be recognized (for example, human eyes) and the third imaging device. , The human eye) performs imaging to obtain a third image.
  • the processing device may return to perform steps 301-307 to reacquire the first image reacquired by the first imaging device and the reacquisition of the second imaging device.
  • the second image based on the re-acquired first image and the second image, extract the depth information of one or more candidate targets, and determine at least from the one or more candidate targets based on the depth information of the one or more candidate targets A candidate target is the target to be identified.
  • step 211 in FIG. 2 For a specific description of performing identity recognition based on the third image, reference may be made to step 211 in FIG. 2.
  • the third imaging device for example, the iris imaging device
  • the third imaging device When the distance between at least a part of the target to be recognized and the third imaging device satisfies the condition (for example, the distance to the third imaging device is relatively close), the third imaging device will start image acquisition, which can effectively avoid continuous acquisition by the third imaging device The image can cause the target that should not be collected to be collected. There is no need for the user to actively turn on or turn off the third imaging device, which improves user experience.
  • Fig. 4 is an exemplary flow chart of another identity recognition method according to some embodiments of the present specification.
  • the process 400 may be executed by the identity recognition apparatus 500 or implemented by the processing device 110 shown in FIG. 1.
  • the process 400 may be stored in a storage device (such as the storage device 140) in the form of a program or instruction, and when the program or instruction is executed, the process 400 may be implemented.
  • the process 400 may include the following steps.
  • the third imaging device that captures at least a part of the image of the target to be recognized needs to have a good focus function of the target to be recognized, so that the captured third image meets the quality requirements for identity recognition.
  • the third imaging device is activated to perform automatic focusing based on the depth information of at least a part of the target to be recognized and collect a third image.
  • Step 401 The processing device may obtain a first image collected by a first imaging device, where the first image includes image representations of one or more candidate targets. In some embodiments, this step may be performed by the candidate target image acquisition module 501.
  • step 201 in the process 200 For a detailed description of acquiring the first image, reference may be made to step 201 in the process 200.
  • Step 403 The processing device may acquire a second image collected by the second imaging device, where the second image includes depth information of at least one candidate target among the one or more candidate targets. In some embodiments, this step may be performed by the candidate target image acquisition module 501.
  • step 203 in the process 200 For a detailed description of acquiring the second image, refer to step 203 in the process 200.
  • Step 405 The processing device may extract depth information of the one or more candidate targets from the second image based on the first image and the second image. In some embodiments, this step may be performed by the depth information extraction module 503.
  • step 205 For a specific description of extracting the depth information of one or more candidate targets from the second image, reference may be made to step 205 in the process 200.
  • the processing device may determine at least one candidate target from the one or more candidate targets as the target to be recognized based on the depth information of the one or more candidate targets. In some embodiments, this step may be performed by the to-be-identified target determination module 505.
  • step 207 For a specific description of determining at least one candidate target as the target to be recognized based on the depth information, reference may be made to step 207 in the process 200.
  • Step 409 The processing device may determine the spatial position information of the at least part of the target to be recognized relative to the third imaging device based on the depth information of the at least part of the target to be recognized. In some embodiments, this step may be performed by the to-be-identified target image acquisition module 507 (for example, a focusing unit (not shown)).
  • the spatial position information of at least a part of the target to be recognized (for example, human face, eye, eye circumference, etc.) relative to the third imaging device may include the spatial position of at least a part of the target to be recognized and the spatial position of the third imaging device
  • the spatial position information of at least a part of the target to be recognized (for example, human face, eye, eye circumference, etc.) relative to the third imaging device may include the distance between at least a part of the target to be recognized and the third imaging device, and the target to be recognized At least a part of the direction with respect to the third imaging device, etc.
  • the spatial position relationship of at least a part of the target to be recognized relative to the third imaging device includes the distance of the human eye relative to the third imaging device.
  • the processing device may determine the spatial position relationship between at least a part of the target to be recognized and the second imaging device based on the depth information of at least a part of the target to be recognized extracted from the second image (also referred to as the first spatial position). relation). Further, the processing device may determine the spatial position relationship of at least a part of the target to be identified relative to the third imaging device based on the first spatial position relationship and the spatial position relationship between the second imaging device and the third imaging device (also referred to as the first The second spatial position relationship), that is, the spatial position information of at least a part of the target to be recognized (for example, human face, human eye, eye circumference, etc.) relative to the third imaging device.
  • the processing device may determine based on the depth information of at least a part of the target to be recognized extracted from the second image and the coordinate conversion relationship between the second imaging device and the geographic coordinate system to determine that at least part of the target to be recognized is geographically Spatial location information (for example, coordinates) in the coordinate system. Further, the processing device may determine, based on the spatial position information (for example, coordinates) of at least a part of the target to be recognized in the geographic coordinate system, and the spatial position information of the third imaging device in the geographic coordinate system, determine that at least a part of the target to be recognized is relative to Spatial location information of the third imaging device.
  • the spatial position relationship between the second imaging device and the third imaging device, the coordinate conversion relationship between the second imaging device and the geographic coordinate system, and/or the spatial position information of the third imaging device in the geographic coordinate system can be determined by the identity recognition system 100 pre-setting.
  • Step 411 The processing device may cause the third imaging device to focus on the at least part of the target to be recognized based on the spatial position information of the at least part of the target to be recognized relative to the third imaging device .
  • this step may be performed by the to-be-identified target image acquisition module 507 (for example, a focusing unit (not shown)).
  • the spatial position information of at least a part of the target to be recognized relative to the third imaging device may include at least a part of the target to be recognized (for example, human eyes, eye patterns, eye circumference) and collection
  • the processing device for example, the focusing unit
  • the correspondence relationship between the object distance interval and the focus position may be constructed in advance.
  • the correspondence between the object distance interval and the focus position includes multiple object distance intervals and corresponding focus positions.
  • the to-be-identified target image acquisition module 507 (for example, a focusing unit (not shown)) can be based on at least a part of the to-be-identified target (for example, human eyes, eye patterns, eye circumference) and the third imaging device (for example, an iris camera).
  • the distance determines the object distance interval to which it belongs from a plurality of object distance intervals. And determine the corresponding focus position according to the object distance interval to which it belongs.
  • the third imaging device includes a voice coil motor.
  • a voice coil motor can be used as a device that converts electrical energy into mechanical energy.
  • the voice coil motor can adjust the distance between the lens of the third imaging device and the image sensor according to the determined focus position to adjust the image distance and the object distance.
  • the third imaging device can adjust the image distance and the object distance during the focusing process by adjusting the position of the lens group, thereby achieving focusing.
  • Step 413 The processing device may obtain the third image and perform identity recognition. In some embodiments, this step may be performed by the target image acquisition module 507 and/or the recognition module 509 to be recognized.
  • the to-be-identified target image acquisition module 507 adjusts the lens position of the third imaging device to the focal length position, so that the lens of the third imaging device can focus on the to-be-identified target.
  • One or more fourth images including at least one candidate target can be acquired by using at least a part of the image of the target to be recognized collected by the third imaging device after focusing. Based on the depth information of at least a part of the target to be recognized, the third image may be obtained from one or more fourth images.
  • one or more fourth images collected by the third imaging device may be activated.
  • the to-be-recognized target image acquisition module 507 (for example, a screening unit (not shown)) may obtain the third image from one or more fourth images based on the depth information of at least a part of the to-be-recognized target (for example, human eyes).
  • the spatial projection relationship ie, geometric mapping relationship or coordinate system conversion relationship
  • step 211 in FIG. 2 For a specific description of performing identity recognition based on the third image, reference may be made to step 211 in FIG. 2.
  • the target to be recognized for example, the human eye
  • the depth information of at least a part of the target to be recognized for example, the human eye
  • the distance from the third imaging device for example, an iris camera
  • the third imaging device can further auto-focus on at least a part of the target to be recognized (for example, human eyes) according to the distance, so that fast and accurate auto-focusing can be realized , Improve the quality of the image collected by the third imaging device.
  • the third imaging device uses a voice coil motor to achieve auto focusing, which can avoid using a complicated stepping motor to drive a mechanical structure to achieve focusing.
  • Fig. 5 is an exemplary module diagram of an identity recognition device according to some embodiments of the present specification.
  • the identity recognition system may include a candidate target image acquisition module 501, a depth information extraction module 503, a target identification module 505, a target image acquisition module 507, a recognition module 509, and a storage module 511.
  • a candidate target image acquisition module 501 a depth information extraction module 503, a target identification module 505, a target image acquisition module 507, a recognition module 509, and a storage module 511.
  • the various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on the differences from other embodiments. Place.
  • the description is relatively simple, and for related parts, please refer to the part of the description of the method embodiment.
  • the candidate target image acquisition module 501 may be used to acquire the first image corresponding to one or more candidate targets acquired by the first imaging device and/or the second image acquired by the second imaging device.
  • the first image includes image representations of one or more candidate targets.
  • the first image may include feature information used to represent or describe the candidate target, for example, texture features, boundary features, color features, and so on.
  • the second image includes depth information of at least one of the candidate targets.
  • the depth information of the candidate target may indicate the distance relationship between a point on the surface of the candidate target (for example, a human face) and the second imaging device.
  • the size of the pixel value of the second image may be used to indicate the size of the distance between the surface of the at least one candidate target and the second imaging device.
  • the second image may include a depth image.
  • the second image may include a point cloud image.
  • the depth information extraction module 503 may extract depth information of one or more candidate targets based on the first image and the second image.
  • the depth information extraction module 503 can register the first image with the second image and obtain the registration result, including the registered first image and the second image.
  • the depth information extraction module 503 can detect one or more candidate targets from the first image after registration. And extract the depth information of one or more candidate targets (for example, human face) from the registered second image based on the detected candidate targets.
  • the to-be-recognized target determining module 505 may be configured to determine at least one candidate target as the to-be-recognized target from the one or more candidate targets based on the depth information of the one or more candidate targets. In some embodiments, the to-be-recognized target determination module 505 may determine at least one candidate target as the to-be-recognized target based on the distance between the candidate target (for example, a human face) and the second imaging device. For example, the to-be-recognized target determining module 505 may determine a candidate target whose distance from the second imaging device is less than a certain threshold or within a certain range as the to-be-recognized target. For another example, the to-be-recognized target determination module 505 may determine the candidate target with the smallest distance from the second imaging device as the to-be-recognized target.
  • the to-be-recognized target image acquisition module 507 may acquire the third image acquired by the third imaging device based on the depth information of at least a part of the to-be-recognized target.
  • the third image includes an image representation of at least a part of the target to be identified.
  • the to-be-identified target image acquisition module 507 includes a screening unit, an activation unit, and a focusing unit.
  • the activation unit may activate the third imaging device to acquire an image in response to a result that the depth information of at least a part of the target to be recognized satisfies the condition.
  • the activation unit may activate the third imaging device to collect one or more fourth images in response to a result that the depth information of at least a part of the target to be recognized meets the condition.
  • the to-be-recognized target image acquisition module 507 (for example, a screening unit) may filter the third image from one or more fourth images based on the depth information of at least a part of the to-be-recognized target.
  • the filtering unit may filter the third image from one or more fourth images based on depth information of at least a part of the target to be recognized (for example, human eyes).
  • the screening unit may obtain the spatial projection relationship between the third imaging device and the second imaging device, and determine that the projection relationship between the fourth image and the third image satisfies the relationship between the third imaging device and the second imaging device.
  • the fourth image of the spatial projection relationship is the third image.
  • the focusing unit may be used to focus the third imaging device according to the depth of at least a part of the target to be recognized, that is, to adjust the image distance of the third imaging device. For example, the focusing unit may determine the belonging object distance interval from multiple object distance intervals according to the distance between at least a part of the target to be identified and the third imaging device, and determine the corresponding focus position according to the belonging object distance interval.
  • the recognition module 509 may be used to perform identity recognition on the target to be recognized based on the third image.
  • the recognition module can be used to preprocess the third image, extract image features from the third image, perform feature encoding on the extracted features, and match the extracted feature codes with pre-stored feature codes to perform Identification.
  • the storage module 511 may be used to store the image data collected by the imaging device, the identity information of the target to be identified, the image processing model and/or algorithm, etc.
  • the storage module 511 may store the first image collected by the first imaging device and the second image collected by the second imaging device.
  • the storage module 511 may store image features of at least a part of a plurality of pre-collected targets for identity recognition.
  • the storage module 511 may store algorithms such as image preprocessing and target detection technology.
  • the storage module 511 includes an internal storage device, an external storage device, and the like.
  • the possible beneficial effects of the embodiments of this specification include, but are not limited to: (1)
  • the target to be recognized can be determined based on the depth information of the candidate target, and the depth information of at least a part of the target to be recognized (for example, human eyes) can be obtained from the current third
  • the third image of the specified target is selected from the images taken by the imaging device (for example, iris camera) for identity recognition (for example, iris recognition), so as to avoid false collection or collection of objects that should not be collected, and improve the quality of the collected images And efficiency, to further improve the speed and accuracy of identity recognition based on biological characteristics such as iris and eye patterns;
  • the number of image sensors in the third imaging device can be increased or the third imaging device (for example, iris camera) can be rotated In the image sensor, change the vertical or horizontal FOV (for example, you can obtain a larger vertical FOV to cover people of different heights), so as to achieve no need to increase the pitch angle adjustment mechanical structure;
  • second imaging The device can use structured light depth camera or TOF
  • the depth information of at least a part of the target to be recognized (for example, human eyes) determined based on the image collected by the second imaging device can determine at least the depth of the target to be recognized
  • the distance between a part (for example, a human eye) and the third imaging device (for example, an iris camera), and the third imaging device can further automatically focus on at least a part of the target to be recognized (for example, the human eye) according to the distance, Fast and accurate autofocus can be realized, and the quality of the image collected by the third imaging device can be improved, and the third imaging device adopts the autofocus realized by the voice coil motor, which can avoid the use of a complicated stepping motor to drive the mechanical structure to achieve focusing.
  • the possible beneficial effects may be any one or a combination of the above, or any other beneficial effects that may be obtained.
  • system and its modules shown in Figure 5 can be implemented in various ways.
  • the system and its modules may be implemented by hardware, software, or a combination of software and hardware.
  • the hardware part can be implemented using dedicated logic;
  • the software part can be stored in a memory and executed by an appropriate instruction execution system, such as a microprocessor or dedicated design hardware.
  • an appropriate instruction execution system such as a microprocessor or dedicated design hardware.
  • processor control codes for example on a carrier medium such as a disk, CD or DVD-ROM, such as a read-only memory (firmware Such codes are provided on a programmable memory or a data carrier such as an optical or electronic signal carrier.
  • the system and its modules in this specification can not only be implemented by hardware circuits such as very large-scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, etc., or programmable hardware devices such as field programmable gate arrays, programmable logic devices, etc. It may also be implemented by software executed by various types of processors, or may be implemented by a combination of the foregoing hardware circuit and software (for example, firmware).
  • modules in the identity recognition device 500 is only for convenience of description, and does not limit this specification within the scope of the embodiments mentioned. It can be understood that for those skilled in the art, after understanding the principle of the system, it is possible to arbitrarily combine various modules, or form a subsystem to connect with other modules without departing from this principle. For example, in some embodiments, for example, what is disclosed in FIG. 5 may be different modules in a system, or a module may implement the functions of two or more modules described above. For example, the depth information extraction module and the target determination module to be identified can be integrated into one module. Such deformations are within the protection scope of this manual
  • the computer storage medium may contain a propagated data signal containing a computer program code, for example on a baseband or as part of a carrier wave.
  • the propagated signal may have multiple manifestations, including electromagnetic forms, optical forms, etc., or a suitable combination.
  • the computer storage medium may be any computer readable medium other than the computer readable storage medium, and the medium may be connected to an instruction execution system, device, or device to realize communication, propagation, or transmission of the program for use.
  • the program code located on the computer storage medium can be transmitted through any suitable medium, including radio, cable, fiber optic cable, RF, or similar medium, or any combination of the above medium.
  • the computer program codes required for the operation of each part of this manual can be written in any one or more programming languages, including object-oriented programming languages such as Java, Scala, Smalltalk, Eiffel, JADE, Emerald, C++, C#, VB.NET, Python Etc., conventional programming languages such as C language, Visual Basic, Fortran 2003, Perl, COBOL 2002, PHP, ABAP, dynamic programming languages such as Python, Ruby and Groovy, or other programming languages.
  • the program code can be run entirely on the user's computer, or run as an independent software package on the user's computer, or partly run on the user's computer and partly run on a remote computer, or run entirely on the remote computer or server.
  • the remote computer can be connected to the user's computer through any network form, such as a local area network (LAN) or a wide area network (WAN), or connected to an external computer (for example, via the Internet), or in a cloud computing environment, or as a service Use software as a service (SaaS).
  • LAN local area network
  • WAN wide area network
  • SaaS service Use software as a service
  • numbers describing the number of ingredients and attributes are used. It should be understood that such numbers used in the description of the embodiments use the modifier "about”, “approximately” or “substantially” in some examples. Retouch. Unless otherwise stated, “approximately”, “approximately” or “substantially” indicates that the number is allowed to vary by ⁇ 20%.
  • the numerical parameters used in the specification and claims are approximate values, and the approximate values can be changed according to the required characteristics of individual embodiments. In some embodiments, the numerical parameter should consider the prescribed effective digits and adopt the method of general digit retention. Although the numerical ranges and parameters used to confirm the breadth of the ranges in some embodiments of this specification are approximate values, in specific embodiments, the setting of such numerical values is as accurate as possible within the feasible range.

Landscapes

  • Engineering & Computer Science (AREA)
  • Human Computer Interaction (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • Theoretical Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • Oral & Maxillofacial Surgery (AREA)
  • General Health & Medical Sciences (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Image Analysis (AREA)
  • Collating Specific Patterns (AREA)

Abstract

本说明书实施例公开了一种身份识别方法。该方法可以包括：获取由第一成像设备采集的第一图像，所述第一图像中包括一个或多个候选目标；获取由第二成像设备采集的第二图像，所述第二图像包括所述一个或多个候选目标中的至少一个候选目标的深度信息；基于所述第一图像与所述第二图像，从所述第二图像中提取所述一个或多个候选目标的深度信息；基于所述一个或多个候选目标的所述深度信息，从所述一个或多个候选目标中确定至少一个候选目标为待识别目标；基于所述待识别目标的至少一部分的深度信息获取由第三成像设备采集的第三图像，所述第三图像包括所述待识别目标的所述至少一部分；以及基于所述第三图像，识别所述待识别目标的身份信息。

Description

身份识别 技术领域
本说明书涉及身份识别技术领域，特别涉及基于深度信息的身份识别方法、***及装置。
背景技术
随着科技的进步和发展，生物个体识别(即生物识别)技术在人们的生产生活中发挥着重要作用。基于生物个体体征(例如，指纹)对生物个体进行识别广泛应用于需要身份验证(即身份识别)的领域，如基于指纹和/或人脸识别的手机解锁、指纹门锁、基于人脸识别的支付等。通常，基于生物个体体征对生物个体进行识别需要采集生物个体的图像数据，以便基于图像数据中的生物个体特征进行身份识别。图像数据采集以及图像数据的质量会较大影响身份识别的速度与准确性。
因此，需要提供一种身份识别方法，可以快速以及准确采集待识别目标的图像数据，以实现快速以及准确的身份识别。
发明内容
本说明书实施例之一提供一种基于身份识别方法，所述方法包括：获取由第一成像设备采集的第一图像，所述第一图像中包括一个或多个候选目标；获取由第二成像设备采集的第二图像，所述第二图像包括所述一个或多个候选目标中的至少一个候选目标的深度信息；基于所述第一图像与所述第二图像，从所述第二图像中提取所述一个或多个候选目标的深度信息；基于所述一个或多个候选目标的所述深度信息，从所述一个或多个候选目标中确定至少一个候选目标为待识别目标；基于所述待识别目标的至少一部分的深度信息获取由第三成像设备采集的第三图像，所述第三图像包括所述待识别目标的所述至少一部分；以及基于所述第三图像，对所述待识别目标进行身份识别。
本说明书实施例之一提供一种身份识别装置。所述装置包括候选目标图像获取模块、深度信息提取模块、待识别目标确定模块、待识别目标图像获取模块。所述候选目标图像获取模块，用于获取第一图像，所述第一图像中包括一个或多个候选目标；以及获取由第二成像设备采集的第二图像，所述第二图像包括所述一个或多个候选目标中的至少一个候选目标的深度信息。所述深度信息提取模块，用于基于所述第一图像与所述第二 图像，从所述第二图像中提取所述一个或多个候选目标的深度信息。所述待识别目标确定模块，用于基于所述一个或多个候选目标的所述深度信息，从所述一个或多个候选目标中确定至少一个候选目标为待识别目标。所述待识别目标图像获取模块，用于基于所述待识别目标的至少一部分的深度信息获取由第三成像设备采集的第三图像，所述第三图像包括所述待识别目标的所述至少一部分。所述装置还包括识别模块，用于基于所述第三图像，对所述待识别目标进行身份识别。
本说明书实施例之一提供一种身份识别***。所述***包括第一成像设备、第二成像设备、第三成像设备。所述第一成像设备，用于采集第一图像，所述第一图像中包括一个或多个候选目标。所述第二成像设备，用于采集第二图像，所述第二图像包括所述一个或多个候选目标中的至少一个候选目标的深度信息。所述第三成像设备，用于采集第三图像，所述第三图像包括所述一个或多个候选目标中的至少一个候选目标的至少一部分。所述***还包括处理器以及存储介质，其中，存储介质用于存储可执行的指令，所述处理器用于执行所述可执行的指令以实现以上所述的身份识别方法。
本说明书实施例之一提供一种计算机可读介质，所述存储介质存储计算机指令，当所述计算机指令被处理器执行时实现以上所述的身份识别方法。
附图说明
本说明书将以示例性实施例的方式进一步说明，这些示例性实施例将通过附图进行详细描述。这些实施例并非限制性的，在这些实施例中，相同的编号表示相同的结构。
图1是根据本说明书一些实施例所示的身份识别***的应用场景示意图；
图2是根据本说明书一些实施例所示的一种身份识别方法的示例性流程图；
图3是根据本说明书一些实施例所示的另一种身份识别方法的示例性流程图；
图4是根据本说明书一些实施例所示的另一种身份识别方法的示例性流程图；
图5是根据本说明书一些实施例所示的身份识别装置的示例性模块图。
具体实施方式
为了更清楚地说明本说明书实施例的技术方案，下面将对实施例描述中所需要使用的附图作简单的介绍。显而易见地，下面描述中的附图仅仅是本说明书的一些示例或实施例，对于本领域的普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这 些附图将本说明书应用于其它类似情景。除非从语言环境中显而易见或另做说明，图中相同标号代表相同结构或操作。
应当理解，本文使用的“***”、“装置”、“单元”和/或“模组”是用于区分不同级别的不同组件、元件、部件、部分或装配的一种方法。然而，如果其他词语可实现相同的目的，则可通过其他表达来替换所述词语。
如本说明书和权利要求书中所示，除非上下文明确提示例外情形，“一”、“一个”、“一种”和/或“该”等词并非特指单数，也可包括复数。一般说来，术语“包括”与“包含”仅提示包括已明确标识的步骤和元素，而这些步骤和元素不构成一个排它性的罗列，方法或者设备也可能包含其它的步骤或元素。
本说明书中使用了流程图用来说明根据本说明书的实施例的***所执行的操作。应当理解的是，前面或后面操作不一定按照顺序来精确地执行。相反，可以按照倒序或同时处理各个步骤。同时，也可以将其他操作添加到这些过程中，或从这些过程移除某一步或数步操作。
图1是根据本说明书一些实施例所示的身份识别***的应用场景示意图。该身份识别***100可以识别待识别目标的身份信息。
如图1所示，身份识别***100可以包括处理设备110、成像设备120、终端130、存储设备140以及网络150。
处理设备110可以处理来自身份识别***100的至少一个其他组件的数据和/或信息。例如，处理设备110可以从成像设备120获取图像数据。又例如，处理设备110可以基于图像数据提取候选目标(例如，人脸)的深度信息并基于候选目标的深度信息确定待识别目标。又例如，处理设备110可以基于待识别目标的至少一部分的图像数据(例如，虹膜图像)对待识别目标进行身份识别。
在一些实施例中，处理设备110可以是单个处理设备，也可以是处理设备组。处理设备组可以是经由接入点连接到网络150的集中式处理设备组，或者经由至少一个接入点分别连接到网络150的分布式处理设备组。在一些实施例中，处理设备110可以本地连接到网络150或者与网络150远程连接。例如，处理设备110可以经由网络150访问存储在终端130和/或存储设备140中的信息和/或数据。又例如，存储设备140可以用作处理设备110的后端数据存储器。在一些实施例中，处理设备110可以在云平台上实施。仅作为示例，所述云平台可以包括私有云、公共云、混合云、社区云、分布云、内 部云、多层云等或其任意组合。
在一些实施例中，处理设备110可以包括处理设备。处理设备可以处理与本说明书中描述的至少一个功能相关的信息和/或数据。在一些实施例中，处理设备可包括至少一个处理单元(例如，单核处理设备或多核处理设备)。仅作为示例，处理设备包括中央处理单元(CPU)、专用集成电路(ASIC)、专用应用指令集处理器(ASIP)、图形处理单元(GPU)、物理处理单元(PPU)、数字信号处理器(DSP)、现场可程序门阵列(FPGA)、可程序逻辑设备(PLD)、控制器、微控制器单元、精简指令集计算机(RISC)、微处理器等，或其任意组合。
成像设备120可以包括具有图像采集功能的多种类型的成像设备，例如，第一成像设备120-1、第二成像设备120-2、第三成像设备120-3等。在一些实施例中，第一成像设备120-1可以用于采集平面图像。例如，第一成像设备120-1可以包括彩色摄像机、数码摄像机、便携式摄像机、PC摄像机、网络摄像机、闭路电视(CCTV)、PTZ摄像相机、视频传感设备等中的一种或其任意组合。第二成像设备120-2可以用于采集深度图像。例如，第二成像设备120-2可以包括结构光深度摄像机、双目立体视觉摄像机、飞行时间TOF摄像机等。第三成像设备120-3可以用于采集红外图像(例如，虹膜图像)。例如，第三成像设备120-3可以包括红外热像仪、红外摄像机等。在一些实施例中，第一成像设备120-1的视野(FOV)和第二成像设备120-2的视野(FOV)有至少部分的重叠。在一些实施例中，第二成像设备120-2的视野(FOV)和第三成像设备120-3的视野(FOV)有至少部分的重叠。在一些实施例中，第一成像设备120-1、第二成像设备120-2、第三成像设备120-3等可以集成于同一设备中。例如，第一成像设备120-1、第二成像设备120-2、第三成像设备120-3等可以是同一设备中的不同成像模块。在一些实施例中，成像设备120可以采集包含候选目标的图像，并将采集到的图像发送至身份识别***100中的一个或多个设备中。例如，成像设备120可以采集包含多个人脸的图像，通过网络150将图像发送至处理设备110进行后续处理。
终端130可以与处理设备110、成像设备120和/或存储设备140通信和/或连接。例如，终端130可以获得通过成像设备120获取的图像数据，并将图像数据发送到处理设备110以进行处理。又例如，终端130可以从处理设备110获得身份识别的结果。在一些实施例中，终端130可以包括移动设备、平板计算机、膝上型计算机等或其任意组合。在一些实施例中，用户可以通过终端130与身份识别***100中的其他组件进行交互。例如，用户可以通过终端130查看成像设备采集的图像。用户也可以通过终端130查看 经过处理设备110确定的身份识别结果。存储设备140可以储存数据和/或指令。例如，存储设备140可以存储成像设备120采集的图像数据、成像设备之间的坐标系转换关系、待识别目标的身份信息、图像处理模型和/或算法等。在一些实施例中，存储设备140可以存储处理设备110可以执行的数据和/或指令，处理设备110可以通过执行或使用所述数据和/或指令以实现本说明书描述的示例性方法。在一些实施例中，存储设备140可包括大容量存储器、可移动存储器、易失性读写存储器、只读存储器(ROM)等或其任意组合。示例性的大容量存储器可以包括磁盘、光盘、固态磁盘等。示例性可移动存储器可以包括闪存驱动器、软盘、光盘、存储卡、压缩盘、磁带等。示例性易失性读写存储器可以包括随机存取存储器(RAM)。示例性随机存取存储器可包括动态随机存取存储器(DRAM)、双倍数据速率同步动态随机存取存储器(DDRSDRAM)、静态随机存取存储器(SRAM)、晶闸管随机存取存储器(T-RAM)和零电容随机存取存储器(Z-RAM)等。示例性只读存储器可以包括掩模型只读存储器(MROM)、可编程只读存储器(PROM)、可擦除可编程只读存储器(PEROM)、电可擦除可编程只读存储器(EEPROM)、光盘只读存储器(CD-ROM)和数字多功能磁盘只读存储器等。在一些实施例中，所述存储设备140可在云平台上实现。仅作为示例，所述云平台可以包括私有云、公共云、混合云、社区云、分布云、内部云、多层云等或其任意组合。
网络150可以促进信息和/或数据的交换。在一些实施例中，身份识别***100中的至少一个组件(例如，处理设备110、成像设备120、终端130、存储设备140)可以经由网络150将信息和/或数据发送到其他组件。例如，处理设备110可以通过网络150从成像设备120获取图像。又例如，处理设备110可以通过网络150将获取的图像发送至终端130。又例如，处理设备110可以用过网络150从存储设备140获取多个对象(例如，生物个体)的身份信息。又例如，处理设备110可以通过网络150将处理后的图像发送至终端130。
在一些实施例中，网络150可以为任意形式的有线或无线网络，或其任意组合。仅作为示例，网络150可以包括缆线网络、有线网络、光纤网络、远程通信网络、内部网络、互联网、局域网络(LAN)、广域网络(WAN)、无线局域网络(WLAN)、城域网(MAN)、公共开关电话网络(PSTN)、蓝牙网络、ZigBee网络、近场通讯(NFC)网络等或其任意组合。在一些实施例中，网络150可以包括至少一个网络接入点。例如，网络150可以包括有线或无线网络接入点，如基站和/或互联网交换点，通过这些网络接入点，身份识别***100的至少一个部件可以连接到网络150以交换数据和/或信息。
应当注意的是，上述身份识别方法的应用身份识别***100的描述仅仅是为了示例和说明，而不限定本说明书的适用范围。对于本领域技术人员来说，在本说明书的指导下可以对该应用身份识别***100进行各种修正和改变。然而，这些修正和改变仍在本说明书的范围之内。
图2是根据本说明书一些实施例所示的一种身份识别方法的示例性流程图。在一些实施例中，流程200可以由身份识别装置500或图1所示的处理设备110实现。例如，流程200可以以程序或指令的形式存储在存储装置(如存储设备140)中，所述程序或指令在被执行时，可以实现流程200。如图2所示，流程200可以包括以下步骤。
步骤201，处理设备可以获取由第一成像设备采集的第一图像，所述第一图像包括一个或多个候选目标。在一些实施例中，该步骤可以由候选目标图像获取模块501执行。
在一些实施例中，候选目标也可以称为候选待识别对象。候选目标可以包括生物个体的全部或一部分。例如，候选目标可以包括人体或人体的面部(即，人脸)。所述第一图像包括一个或多个候选目标可以指的是第一图像包括候选目标的图像表示。候选目标的图像表示也可以称为候选目标的图像描述。第一图像可以包括用于表示或描述候选目标的特征信息，例如，纹理特征、边界特征、颜色特征等。在一些实施例中，第一图像可以是二维图像。在一些实施例中，第一图像的图像类型包括以下中的至少一种：灰度图像、RGB图像等中的一种或其任意组合。
第一成像设备可包括用于采集候选目标的平面图像(如，第一图像)的成像设备。例如，第一成像设备可包括但不限于彩色摄像机、数码摄像机、便携式摄像机、PC摄像机、网络摄像机、闭路电视(CCTV)、PTZ摄像机、视频传感设备等中的一种或其任意组合。
在一些实施例中，处理设备(例如，候选目标图像获取模块501)可以从第一成像设备、存储设备140或其他存储设备中获取第一图像。在一些实施例中，处理设备可以获取由一个第一成像设备采集的单张第一图像。在一些实施例中，处理设备可以获取由多个第一成像设备同时采集的多张第一图像。多张第一图像中的至少一张第一图像包括一个或多个候选目标。例如，多张第一图像中的至少一张第一图像包括多个人脸。在一些实施例中，候选目标图像获取模块501可以从第一成像设备中获取第一图像，并将第一图像存储在存储设备140中。
步骤203，处理设备可以获取由第二成像设备采集的第二图像，第二图像包括候选 目标中的至少一个候选目标的深度信息。在一些实施例中，该步骤可以由候选目标图像获取模块501执行。
候选目标的深度信息可以表示候选目标(例如，人脸)表面上的点与第二成像设备之间的距离。例如，可以通过第二图像的像素值的大小表示候选目标的表面与第二成像设备之间的距离大小。在一些实施例中，第二图像可以包括深度图像。在一些实施例中，第二图像可以包括点云图。在一些实施例中，可以将深度图像转化为点云图以此获取第二图像。在一些实施例中，可以将点云图转化为深度图以此获取第二图像。第二图像可以是二维图像、三维图像等。
第二成像设备包括可以采集候选目标(例如，人脸)的深度信息的成像设备。在一些实施例中，第二成像设备可以包括一个或多个深度成像设备。深度成像设备可以包括但不限于：结构光(Structured Light)深度摄像机、双目立体视觉(Binocular Stereo Vision)摄像机、飞行时间TOF(Time of flight)摄像机等中的一种或其任意组合。在一些实施例中，第二成像设备的视野(FOV)和第一成像设备的视野(FOV)有至少部分的重叠。在一些实施例中，第一成像设备与第二成像设备同时采集第一图像和第二图像。
在一些实施例中，候选目标图像获取模块501可以从第二成像设备、存储设备140或其他存储设备中获取第二图像。在一些实施例中，候选目标图像获取模块501可以获取由一个第二成像设备采集的单张第二图像。在一些实施例中，候选目标图像获取模块501可以获取由多个第二成像设备同时采集的多张第二图像。多张第二图像中的至少一张第二图像包括一个或多个候选目标中的至少一个候选目标的深度信息。每张第二图像可以对应一张第一图像。如本文中所述，第二图像对应第一图像指的是第二图像中的像素与第一图像中的某个像素对应候选目标上相同的位置或相同的部分。
步骤205，处理设备可以基于所述第一图像与所述第二图像，提取所述一个或多个候选目标的深度信息。在一些实施例中，该步骤可以由深度信息提取模块503执行。
在一些实施例中，可以在采集第一图像和第二图像前对第一成像设备和第二成像设备基于同一坐标系(例如，世界坐标系)进行标定，使得第一成像设备和第二成像设备具有统一的坐标系。处理设备可以直接从第一图像中检测一个或多个候选目标(例如，人脸)。处理设备可以基于检测到的一个或多个候选目标在第一图像中的位置从第二图像中提取一个或多个候选目标(例如，人脸)的深度信息。
在一些实施例中，处理设备可以将第一图像与第二图像进行配准以获得配准结果。 处理设备可以从配准后的第一图像中检测一个或多个候选目标(例如，人脸)。深度信息提取模块503可以基于检测到的一个或多个候选目标以及配准结果提取一个或多个候选目标(例如，人脸)的深度信息。
在一些实施例中，处理设备可以通过图像配准技术将第一图像与第二图像进行配准。示例性图像配准技术可以包括基于灰度和模板匹配算法、基于特征的匹配算法、基于域变换的算法等。
在一些实施例中，处理设备可以利用图像分割技术、基于模型的目标检测技术等方法从配准后的第一图像中检测候选目标。图像分割技术可以包括利用基于边缘分割算法、基于阈值分割算法、基于区域分割算法、形态学分水岭算法等或其组合。基于模型的目标检测技术可以包括利用机器学习模型(R-CNN模型、Fast RCNN模型、SVM模型等)进行目标检测。在一些实施例中，处理设备可以基于检测到的候选目标对配准后的第一图像进行掩膜处理以获取掩膜图像。例如，可以将配准后的第一图像中候选目标所在区域的像素值设置为1，将其余区域的像素值设置为0。
在一些实施例中，可以基于检测到的候选目标在配准后的第一图像中的位置从配准后的第二图像中提取一个或多个候选目标的深度信息。可以基于检测到的候选目标在配准后的第一图像中的位置确定检测到的候选目标在配准后的第二图像中的位置。例如，可以将具有检测到的候选目标的掩膜图像与配准后的第二图像相乘，以确定检测到的候选目标在配准后的第二图像中的位置。进一步的，可以从配准后的第二图像中提取检测到的候选目标在配准后的第二图像中的位置处的深度信息。
步骤207，处理设备可以基于所述一个或多个候选目标的所述深度信息，从所述一个或多个候选目标中确定至少一个候选目标为待识别目标。在一些实施例中，该步骤可以由待识别目标确定模块505执行。
在本说明书的一个或多个实施例中，可以基于一个或多个候选目标(例如，人脸)的深度信息确定一个或多个候选目标的空间位置关系，并基于一个或多个候选目标的空间位置关系确定待识别目标。在一些实施例中，候选目标的空间位置关系可以包括候选目标与第二成像设备之间的空间位置关系。例如，候选目标的空间位置关系可以表示为候选目标与第二成像设备的距离。
在一些实施例中，处理设备可以基于候选目标(例如，人脸)与第二成像设备的距离确定一个或多个候选目标中至少一个候选目标为待识别目标。当候选目标与第二成像 设备的距离满足一定条件时可以确定该候选目标为待识别目标。例如，成像设备可以将与第二成像设备的距离小于一定阈值(例如，小于1米、或小于2米、或小于3米、或小于4米等)的或第二成像设备的距离在一定范围内(例如，1-2米范围内、或0.5-4米范围内、或0-6米范围内)的候选目标确定为待识别目标。又例如，处理设备可以将与第二成像设备的距离最小的候选目标确定为待识别目标。在一些实施例中，若两个或以上候选目标与第二成像设备之间的距离相同且满足上述条件，处理设备可以基于一定准则从两个或以上候选目标中确定待识别目标。例如，处理设备可以基于与第二成像设备之间距离相同的候选目标在第一图像中的位置确定待识别目标。进一步的，待识别目标确定模块505可以将在第一图像中靠近图像左边或右边的候选目标确定为待识别目标。
步骤209，处理设备基于所述待识别目标的至少一部分的深度信息获取由第三成像设备采集的第三图像，所述第三图像包括所述待识别目标的所述至少一部分。该步骤可以由待识别目标图像获取模块507执行。
在一些实施例中，所述待识别目标可以包括人脸，待识别目标的至少一部分可以包括人眼、虹膜、眼纹、眼周中的至少一个。所述第三图像包括所述待识别目标的所述至少一部分也可称为所述第三图像包括所述待识别目标的所述至少一部分(例如，人眼)的图像表示。
在一些实施例中，处理设备(例如，待识别目标图像获取模块507)可以获取由第三成像设备采集的一张或多张第四图像(例如，人眼图像)。处理设备可以基于待识别目标的至少一部分(例如，人眼)的深度信息从一张或多张第四图像中确定第三图像。例如，可以获取第三成像设备和第二成像设备之间的坐标系转换关系(即几何映射关系或空间投影关系)。指定第四图像中与待识别目标的至少一部分(例如，人眼)的深度信息之间的投影关系满足第三成像设备和第二成像设备之间的坐标系转换关系的第四图像为第三图像。又例如，处理设备可以利用第三成像设备和第二成像设备之间的坐标系转换关系将待识别目标的至少一部分(例如，人眼)的深度信息投影到每张第四图像所在的平面。处理设备可以将第四图像中与投影后的信息匹配的第四图像指定为第三图像。第三成像设备和第二成像设备之间的坐标系转换关系与第二成像设备以及第三成像的标定参数相关，为身份识别***100的默认设置。
在一些实施例中，处理设备可以基于待识别目标的至少一部分(例如，人眼)的深度信息对待识别目标的至少一部分(例如，人眼)进行定位以确定待识别目标的至少一部分(例如，人眼)相对于第三成像设备的空间位置信息。例如，可以基于待识别目标 的至少一部分(例如，人眼)的深度信息确定待识别目标的至少一部分(例如，人眼)距离第二成像设备的距离与方向。进一步的，可以基于待识别目标的至少一部分(例如，人眼)与第二成像设备的距离与方向以及第二成像设备与第三成像设备之间的空间位置关系确定待识别目标的至少一部分(例如，人眼)距离第三成像设备的距离与方向。第三成像设备可以基于待识别目标的至少一部分(例如，人眼)与第三成像设备的距离与方向对待识别目标的至少一部分(例如，人眼)进行成像以获取第三图像。
在一些实施例中，可以基于待识别目标的至少一部分的深度信息确定待识别目标相对于第三成像设备的空间位置信息。可以进一步基于待识别目标相对于第三成像设备的空间位置信息启动第三成像设备对待识别目标的至少一部分对焦，以获取第三图像。关于第三设备的自动对焦的更多描述参考图4。
在一些实施例中，处理设备可以判断待识别目标的至少一部分的深度信息是否满足一定条件，并基于判断结果确定是否启动第三成像设备采集第三图像。关于第三成像设备的启动更多描述可以参考图3。
第三成像设备包括一个或多个红外成像设备。在一些实施例中，第三设备可以包括一个或多个图像传感器，例如，CMOS图像传感器、CCD图像传感器等。在一些实施例中，第三成像设备的垂直视角(FOV)或水平视角(FOV)中的至少一个大于阈值或在一定范围内。例如，可以通过在第三成像设备的垂直方向安装多个图像传感器，使得第三成像设备的垂直FOV在一定范围内，例如，在0-60度，或在0-90度，或在0-120度等范围。又例如，可以通过在第三成像设备的水平方向安装多个图像传感器，使得第三成像设备的水平FOV在一定范围内，例如，在0-60度，或在0-90度，或在0-120度等范围。再例如，可以在第三成像设备的垂直方向和水平方向同时安装多个图像传感器，使得第三成像设备的垂直FOV和水平FOV在一定范围内，例如，在0-60度，或在0-90度，或在0-120度等范围。在一些实施例中，第三设备中的图像传感器可以沿着多个自由度旋转，例如，可以顺时针或逆时针旋转。可以通过旋转第三设备中的图像传感器以改变某一方向的FOV。例如，若水平FOV更大，则可以通过将第三设备中的图像传感器旋转90度以使得垂直方向FOV更大。
步骤211，处理设备可以基于所述第三图像，对所述待识别目标进行身份识别。该步骤可以由识别模块509执行。
在一些实施例中，可以预先采集多个对象的至少一部分(例如，虹膜)的图像并提取图像特征。预先提取的图像特征可以以特征编码的形式存储在存储设备140中，也可 以直接存储在外部数据库中。识别模块509可以利用特征提取算法从第三图像中提取特征。在一些实施例中，识别模块509在进行特征提取前可以对第三图像进行预处理，例如，图像平滑、边缘检测、图像分离等。在一些实施例中，识别模块509可以进一步对从第三图像中提取的特征进行特征编码。识别模块509可以将从第三图像中获取的特征编码与预先存储的特征编码进行匹配以便对待识别目标进行身份识别。
根据本说明中的一些实施例，可以基于候选目标的深度信息确定待识别目标。进一步的，可以基于待识别目标的至少一部分(例如，人眼)的深度信息从当前第三成像设备(例如，虹膜摄像机)拍摄的图像中筛选出指定目标的第三图像进行身份识别(例如，虹膜识别)，从而可以避免误采集或采集到不应该采集到的对象，提高采集的图像质量以及效率，进一步提高身份识别的速度以及准确性。
根据本说明书中的一些实施例，可以通过增加第三成像设备中图像传感器的个数或旋转第三成像设备(例如，虹膜摄像机)中的图像传感器，扩大垂直或水平方向的FOV(例如，可以获取更大的垂直方向FOV以便覆盖不同身高的人群)，从而达到不需要增加调节俯仰角的机械结构件。
根据本说明书中的一些实施例，第二成像设备可以采用结构光深度摄像头或者TOF深度摄像头，可以有效降低对环境光的依赖，可以提高深度信息的精确度，从而提高待识别目标确定的准确度，进一步提高第三图像(例如，虹膜图像)采集的精确性以及质量，提高身份识别(例如，虹膜识别)的速度以及准确性。
应当注意的是，上述有关流程200的描述仅仅是为了示例和说明，而不限定本说明书的适用范围。对于本领域技术人员来说，在本说明书的指导下可以对流程200进行各种修正和改变。然而，这些修正和改变仍在本说明书的范围之内上述对本说明书特定实施例进行了描述。其它实施例在所附权利要求书的范围内。在一些情况下，在权利要求书中记载的动作或步骤可以按照不同于实施例中的顺序来执行并且仍然可以实现期望的结果。另外，在附图中描绘的过程不一定要求示出的特定顺序或者连续顺序才能实现期望的结果。在某些实施方式中，多任务处理和并行处理也是可以的或者可能是有利的。
图3是根据本说明书一些实施例所示的另一种身份识别方法的示例性流程图。在一些实施例中，流程300可以由身份识别装置500或图1所示的处理设备110实现。例如，流程300可以以程序或指令的形式存储在存储装置(如存储设备140)中，所述程序或指令在被执行时，可以实现流程300。如图3所示，流程300可以包括以下步骤。
在一些实施例中，第三成像设备可以与第一成像设备、第二成像设备同时采集图像，在一些实施例中，第三成像设备可以基于相应的条件启动图像采集。例如，第三成像设备可以基于第一成像设备与第二成像设备采集的第一、第二图像的结果确定是否启动进行第三图像的采集等操作。例如，候选目标图像获取模块501可以获取第一成像设备采集的第一图像与第二成像设备采集的第二图像，第一、第二图像中包括一个或多个候选目标的图像表示以及至少一个候选目标的深度信息，基于第二图像中待识别目标的深度信息是否满足相应条件的结果，可以确定是否启动第三成像设备采集第三图像。
步骤301，处理设备可以获取由第一成像设备采集的第一图像，所述第一图像中包括一个或多个候选目标的图像表示。在一些实施例中，该步骤可以由候选目标图像获取模块501执行。
关于获取第一图像的具体描述可以参考流程200中的步骤201。
步骤303，处理设备可以获取由第二成像设备采集的第二图像，所述第二图像包括一个或多个候选目标中的至少一个候选目标的深度信息。在一些实施例中，该步骤可以由候选目标图像获取模块501执行。
关于获取第二图像的具体描述可以参考流程200中的步骤203。
步骤305，处理设备可以基于所述第一图像与所述第二图像，从所述第二图像中提取所述一个或多个候选目标的深度信息。在一些实施例中，该步骤可以由深度信息提取模块503执行。
关于从第二图像中提取一个或多个候选目标的深度信息的具体描述可以参考流程200中的步骤205。
步骤307，处理设备可以基于所述一个或多个候选目标的所述深度信息，从所述一个或多个候选目标中确定至少一个候选目标为待识别目标。在一些实施例中，该步骤可以由待识别目标确定模块505执行。
关于基于深度信息确定至少一个候选目标为待识别目标的具体描述可以参考流程200中的步骤207。
步骤309，处理设备可以判断所述待识别目标的所述至少一部分的所述深度信息是否满足条件。在一些实施例中，该步骤可以由待识别目标图像获取模块507(例如，启动单元(未示出))执行。
如本文中所述，待识别目标(例如，人脸)的至少一部分(例如，人眼、眼纹、眼周等)的深度信息可以表示待识别目标的表面上的点与第二成像设备之间的距离关系。在一些实施例中，可以基于待识别目标的至少一部分(例如，人眼、眼纹、眼周等)的表面上的点与第二成像设备之间的距离以及第二成像设备与第三成像设备之间的空间位置关系(例如，方向、距离等)确定待识别目标的至少一部分(例如，人眼、眼纹、眼周等)表面上的点与第三成像设备之间的距离。在一些实施例中，可以基于待识别目标的至少一部分(例如，人眼、眼纹、眼周等)的表面上的点与第二成像设备之间的距离以及第二成像设备与地理坐标系之间的坐标系转换关系确定待识别目标的至少一部分(例如，人眼、眼纹、眼周等)表面上的点在地理坐标系中的位置。进一步的，可以基于待识别目标的至少一部分(例如，人眼、眼纹、眼周等)表面上的点在地理坐标系中的位置以及第三成像设备在地理坐标系中的位置确定待识别目标的至少一部分(例如，人眼、眼纹、眼周等)表面上的点与第三成像设备之间的距离。
在一些实施例中，判断待识别目标的至少一部分的深度信息是否满足条件包括判断待识别目标的至少一部分与第三成像设备之间的距离是否满足一定条件。例如，可以判断待识别目标的至少一部分与第三成像设备之间的距离是否在一定距离范围内。若待识别目标的至少一部分与第三成像设备之间的距离在一定距离范围内，则可以确定待识别目标的至少一部分的深度信息满足条件。若待识别目标的至少一部分与第三成像设备之间的距离不在一定距离范围内，则可以确定待识别目标的至少一部分的深度信息不满足条件。该距离范围可以包括30-70cm、20-80cm、10-90cm等。又例如，可以判断待识别目标的至少一部分与第三成像设备之间的距离是否小于一定距离阈值。若待识别目标的至少一部分与第三成像设备之间的距离小于一定距离阈值，则可以确定待识别目标的至少一部分的深度信息满足条件。若待识别目标的至少一部分与第三成像设备之间的距离不小于一定距离阈值，则可以确定待识别目标的至少一部分的深度信息不满足条件。该距离阈值可以包括70cm、80cm、90cm等。
在一些实施例中，待识别目标的至少一部分上表面上的点可以不在同一个平面上，即待识别目标的至少一部分的表面上的点与第三成像设备之间的距离可以不一样。可以基于待识别目标的至少一部分的表面上的点与第三成像设备之间的距离确定待识别目标的至少一部分与第三成像设备之间的距离。例如，可以确定待识别目标的至少一部分的表面上的点与第三成像设备之间的距离平均值为待识别目标的至少一部分与第三成像设备之间的距离。又例如，可以确定待识别目标的至少一部分的表面上的点与第三成像设备之间的距离的中值为待识别目标的至少一部分与第三成像设备之间的距离。
步骤311，处理设备可以响应于待识别目标的至少一部分的深度信息满足条件，启动第三成像设备采集待识别目标的至少一部分的第三图像。在一些实施例中，该步骤可以由待识别目标图像获取模块507(例如，启动单元(未示出))执行。
在一些实施例中，响应于待识别目标的至少一部分的深度信息满足条件，可以启动第三成像设备采集的一张或多张第四图像(例如，人眼图像)。处理设备(例如，筛选单元(未示出))可以基于待识别目标的至少一部分(例如，人眼)的深度信息从一张或多张第四图像中获得第三图像。例如，可以获取第三成像设备和第二成像设备之间的坐标系转换关系(即几何映射关系或空间投影关系)。指定第四图像中与待识别目标的至少一部分(例如，人眼)的深度信息之间的投影关系满足第三成像设备和第二成像设备之间的坐标系转换关系的第四图像为第三图像。又例如，处理设备可以利用第三成像设备和第二成像设备之间的坐标系转换关系将待识别目标的至少一部分(例如，人眼)的深度信息投影到每张第四图像所在的平面。处理设备可以将第四图像中与投影后的信息匹配的第四图像指定为第三图像。第三成像设备和第二成像设备之间的坐标系转换关系与第二成像设备以及第三成像的标定参数相关，为身份识别***100的默认设置。
在一些实施例中，处理设备可以基于待识别目标的至少一部分(例如，人眼)的深度信息对待识别目标的至少一部分(例如，人眼)进行定位以确定待识别目标的至少一部分(例如，人眼)相对于第三成像设备的空间位置信息。例如，可以基于待识别目标的至少一部分(例如，人眼)的深度信息确定待识别目标的至少一部分(例如，人眼)距离第二成像设备的距离与方向。进一步的，可以基于待识别目标的至少一部分(例如，人眼)与第二成像设备的距离与方向以及第二成像设备与第三成像设备之间的空间位置关系确定待识别目标的至少一部分(例如，人眼)与第三成像设备之间的距离与方向。响应于待识别目标的至少一部分的深度信息满足条件，可以启动第三成像设备基于待识别目标的至少一部分(例如，人眼)与第三成像设备的距离与方向对待识别目标的至少一部分(例如，人眼)进行成像以获取第三图像。
在一些实施例中，响应于待识别目标的至少一部分的深度信息不满足条件，处理设备可以返回执行步骤301-307重新获取第一成像设备重新采集的第一图像以及第二成像设备重新采集的第二图像、基于重新获取的第一图像与第二图像，提取一个或多个候选目标的深度信息、基于所述一个或多个候选目标的深度信息，从一个或多个候选目标中确定至少一个候选目标为待识别目标。
关于基于第三图像进行身份识别的具体描述可以参考图2中的步骤211。
根据本说明书中的一个或多个实施例，可以根据第二成像设备采集的图像确定的候选目标的至少一部分(例如，人脸或人眼)的深度信息判断当前是否有待识别或待验证的目标，以此判断是否启动第三成像设备(例如，虹膜成像设备)。当待识别目标的至少一部分与第三成像设备之间的距离不满足条件(例如，距离第三成像设备较远)时，第三成像设备将不会启动。当待识别目标的至少一部分与第三成像设备之间的距离满足条件(例如，距离第三成像设备较近)时，第三成像设备将启动进行图像采集，可以有效避免第三成像设备不断采集图像以造成采集不到应该采集的目标。无需用户主动开启或关闭第三成像设备，提高用户体验。
应当注意的是，上述有关流程300的描述仅仅是为了示例和说明，而不限定本说明书的适用范围。对于本领域技术人员来说，在本说明书的指导下可对流程300进行各种修正和改变。然而，这些修正和改变仍在本说明书的范围之内。上述对本说明书特定实施例进行了描述。其它实施例在所附权利要求书的范围内。在一些情况下，在权利要求书中记载的动作或步骤可以按照不同于实施例中的顺序来执行并且仍然可以实现期望的结果。另外，在附图中描绘的过程不一定要求示出的特定顺序或者连续顺序才能实现期望的结果。在某些实施方式中，多任务处理和并行处理也是可以的或者可能是有利的。
图4是根据本说明书一些实施例所示的另一种身份识别方法的示例性流程图。在一些实施例中，流程400可以由身份识别装置500执行，或图1所示的处理设备110实现。例如，流程400可以以程序或指令的形式存储在存储装置(如存储设备140)中，所述程序或指令在被执行时，可以实现流程400。如图4所示，流程400可以包括以下步骤。
为了快速、准确地采集待识别目标的至少一部分(例如，人眼、虹膜、眼纹、眼周等)的第三图像，需要获取质量较高的目标图像。采集待识别目标的至少一部分的图像的第三成像设备需要良好的对待识别目标的对焦功能，以使采集的第三图像满足身份识别的质量要求。在本说明书的一个或多个实施例中，基于待识别目标的至少一部分的深度信息，启动第三成像设备基于待识别目标的至少一部分的深度信息进行自动对焦并采集第三图像。
步骤401，处理设备可以获取由第一成像设备采集的第一图像，所述第一图像中包括一个或多个候选目标的图像表示。在一些实施例中，该步骤可以由候选目标图像获取模块501执行。
关于获取第一图像的具体描述可以参考流程200中的步骤201。
步骤403，处理设备可以获取由第二成像设备采集的第二图像，所述第二图像包括一个或多个候选目标中的至少一个候选目标的深度信息。在一些实施例中，该步骤可以由候选目标图像获取模块501执行。
关于获取第二图像的具体描述可以参考流程200中的步骤203。
步骤405，处理设备可以基于所述第一图像与所述第二图像，从所述第二图像中提取所述一个或多个候选目标的深度信息。在一些实施例中，该步骤可以由深度信息提取模块503执行。
关于从第二图像中提取一个或多个候选目标的深度信息的具体描述可以参考流程200中的步骤205。
步骤407，处理设备可以基于所述一个或多个候选目标的所述深度信息，从所述一个或多个候选目标中确定至少一个候选目标为待识别目标。在一些实施例中，该步骤可以由待识别目标确定模块505执行。
关于基于深度信息确定至少一个候选目标为待识别目标的具体描述可以参考流程200中的步骤207。
步骤409，处理设备可基于所述待识别目标的所述至少一部分的所述深度信息确定所述待识别目标的所述至少一部分相对于所述第三成像设备的空间位置信息。在一些实施例中，该步骤可由待识别目标图像获取模块507(例如，对焦单元(未示出))执行。
待识别目标的至少一部分(例如，人脸、人眼、眼周等)相对于第三成像设备的空间位置信息可包括待识别目标的至少一部分所处的空间位置与第三成像设备的空间位置之间的关系，即识别目标的至少一部分相对于第三成像设备的空间位置关系。例如，待识别目标的至少一部分(例如，人脸、人眼、眼周等)相对于第三成像设备的空间位置信息可包括待识别目标的至少一部分与第三成像设备的距离、待识别目标的至少一部分相对于第三成像设备的方向等。以待识别目标的至少一部分为人眼为例，待识别目标的至少一部分相对于第三成像设备的空间位置关系包括人眼相对于第三成像设备的距离。
在一些实施例中，处理设备可以基于第二图像中提取的待识别目标的至少一部分的深度信息确定待识别目标的至少一部分与第二成像设备的空间位置关系(也可称为第一空间位置关系)。进一步的，处理设备可以基于第一空间位置关系以及第二成像设备与第三成像设备的空间位置关系获得确定待识别目标的至少一部分相对于第三成像设备 的空间位置关系(也可称为第二空间位置关系)，即待识别目标的至少一部分(例如，人脸、人眼、眼周等)相对于第三成像设备的空间位置信息。
在一些实施例中，处理设备可以基于第二图像中提取的待识别目标的至少一部分的深度信息确定以及第二成像设备与地理坐标系之间的坐标转换关系确定待识别目标的至少一部分在地理坐标系中的空间位置信息(例如，坐标)。进一步的，处理设备可以基于待识别目标的至少一部分在地理坐标系中的空间位置信息(例如，坐标)以及第三成像设备在地理坐标系中的空间位置信息确定待识别目标的至少一部分相对于第三成像设备的空间位置信息。
第二成像设备与第三成像设备的空间位置关系、第二成像设备与地理坐标系之间的坐标转换关系和/或第三成像设备在地理坐标系中的空间位置信息可以由身份识别***100预先设定。
步骤411，处理设备可以基于所述待识别目标的所述至少一部分相对于所述第三成像设备的空间位置信息，使所述第三成像设备对所述待识别目标的所述至少一部分进行对焦。在一些实施例中，该步骤可以由待识别目标图像获取模块507(例如，对焦单元(未示出))执行。
在本说明书的一个或多个实施例中，待识别目标的至少一部分相对于第三成像设备的空间位置信息可以包括待识别目标的至少一部分(例如，人眼、眼纹、眼周)与采集待识别目标图像的第三成像设备(例如，虹膜摄像机)的距离。处理设备(例如，对焦单元)可以根据待识别目标与第三成像设备的距离进行对焦，即调节第三成像设备的物距和/或像距。
在本说明书的一个或多个实施例中，可以预先构建物距区间与对焦位置的对应关系。物距区间与对焦位置的对应关系包括多个物距区间以及相对应的对焦位置。待识别目标图像获取模块507(例如，对焦单元(未示出))可以根据待识别目标的至少一部分(例如，人眼、眼纹、眼周)与第三成像设备(例如，虹膜摄像机)的距离从多个物距区间确定所属的物距区间。并根据所属的物距区间确定对应的对焦位置。在一些实施例中，第三成像设备包括音圈马达。音圈马达可以用于将电能转化为机械能的装置。音圈马达可以根据确定的对焦位置调整第三成像设备的镜头与图像传感器之间的距离，以调整像距和物距。第三成像设备可以通过调节镜头组的位置实现对焦过程中的像距和物距调节，从而实现对焦。
步骤413，处理设备可以获取第三图像并进行身份识别。在一些实施例中，该步骤可以由待识别目标图像获取模块507和/或识别模块509执行。
待识别目标图像获取模块507通过将第三成像设备中的镜头位置调节到焦距位置，可实现第三成像设备的镜头对待识别目标的对焦。利用对焦后第三成像设备采集到的待识别目标的至少一部分的图像，可获取包括至少一个候选目标的一张或多张第四图像。再基于待识别目标的至少一部分的深度信息，可从一张或多张第四图像中获取第三图像。
在一些实施例中，基于对焦后的第三成像设备，可以启动第三成像设备采集的一张或多张第四图像(例如，人眼图像)。待识别目标图像获取模块507(例如，筛选单元(未示出))可以基于待识别目标的至少一部分(例如，人眼)的深度信息从一张或多张第四图像中获得第三图像。例如，可以获取第三成像设备和第二成像设备之间的空间投影关系(即几何映射关系或坐标系转换关系)。指定一张或多张第四图像中与第三图像之间的投影关系满足第三成像设备和第二成像设备之间的空间投影关系的第四图像为第三图像。
关于基于第三图像进行身份识别的具体描述可以参考图2中的步骤211。
根据本说明书的一个或多个实施例，可以基于第二成像设备采集的图像确定的待识别目标的至少一部分(例如，人眼)的深度信息确定待识别目标的至少一部分(例如，人眼)与第三成像设备(例如，虹膜摄像机)之间的距离，进一步可以根据该距离对第三成像设备针对待识别目标的至少一部分(例如，人眼)进行自动对焦，可以实现快速准确的自动对焦，提高第三成像设备采集的图像质量。
根据本说明书的一个或多个实施例，第三成像设备采用音圈马达实现的自动对焦，可以避免使用复杂的步进电机驱动机械结构实现对焦。
应当注意的是，上述有关流程400的描述仅仅是为了示例和说明，而不限定本说明书的适用范围。对于本领域技术人员来说，在本说明书的指导下可对流程300进行各种修正和改变。然而，这些修正和改变仍在本说明书的范围之内。上述对本说明书特定实施例进行了描述。其它实施例在所附权利要求书的范围内。在一些情况下，在权利要求书中记载的动作或步骤可以按照不同于实施例中的顺序来执行并仍然可实现期望的结果。另外，在附图中描绘的过程不一定要求示出的特定顺序或者连续顺序才能实现期望的结果。在某些实施方式中，多任务处理和并行处理也是可以的或者可能是有利的。
图5是根据本说明书一些实施例所示的身份识别装置的示例性模块图。
如图5所示，该身份识别***可以包括候选目标图像获取模块501、深度信息提取模块503、待识别目标确定模块505、待识别目标图像获取模块507、识别模块509、存储模块511。需要注意的是，本说明书中的各个实施例均采用递进的方式描述，各个实施例之间相同相似的部分互相参见即可，每个实施例重点说明的都是与其他实施例的不同之处。尤其，对于装置实施例而言，由于其基本相似于方法实施例，所以描述的比较简单，相关之处参见方法实施例的部分说明即可。
候选目标图像获取模块501可以用于获取由第一成像设备采集的一个或多个候选目标对应的第一图像和/或由第二成像设备采集的第二图像。
第一图像包括一个或多个候选目标的图像表示。第一图像可以包括用于表示或描述候选目标的特征信息，例如，纹理特征、边界特征、颜色特征等。第二图像包括候选目标中的至少一个候选目标的深度信息。候选目标的深度信息可以表示候选目标(例如，人脸)表面上的点与第二成像设备之间的距离关系。例如，可以通过第二图像的像素值的大小表示至少一个候选目标的表面与第二成像设备之间的距离大小。在一些实施例中，第二图像可以包括深度图像。在一些实施例中，第二图像可以包括点云图。关于第一成像设备以及第二成像设备的更多描述可以参考本说明书其他部分。
深度信息提取模块503可以基于第一图像和第二图像，提取一个或多个候选目标的深度信息。在一些实施例中，深度信息提取模块503可以将第一图像与第二图像进行配准并获取配准结果，包括配准后的第一图像和第二图像。深度信息提取模块503可以从配准后第一图像中检测一个或多个候选目标。并基于检测到的候选目标从配准后的第二图像中提取一个或多个候选目标(例如，人脸)的深度信息。
待识别目标确定模块505可以用于基于一个或多个候选目标的深度信息，从一个或多个候选目标中确定至少一个候选目标为待识别目标。在一些实施例中，待识别目标确定模块505可以基于候选目标(例如，人脸)与第二成像设备的距离确定至少一个候选目标为待识别目标。例如，待识别目标确定模块505可以将与第二成像设备的距离小于一定阈值的或距离在一定范围内的候选目标确定为待识别目标。又例如，待识别目标确定模块505可以将与第二成像设备的距离最小的候选目标确定为待识别目标。
待识别目标图像获取模块507可基于待识别目标的至少一部分的深度信息获取由第三成像设备采集的第三图像。第三图像包括待识别目标的至少一部分的图像表示。在一些实施例中，待识别目标图像获取模块507包括筛选单元、启动单元和对焦单元。
在一些实施例中，启动单元可以响应于待识别目标的至少一部分的深度信息满足条件的结果，启动第三成像设备采集图像。例如，启动单元可以响应于待识别目标的至少一部分的深度信息满足条件的结果，启动第三成像设备采集一张或多张第四图像。待识别目标图像获取模块507(例如，筛选单元)可以基于待识别目标的至少一部分的深度信息从一张或多张第四图像中筛选第三图像。
在一些实施例中，筛选单元可基于待识别目标的至少一部分(例如，人眼)的深度信息从一张或多张第四图像中筛选第三图像。例如，筛选单元可获取第三成像设备和第二成像设备之间的空间投影关系，并通过确定第四图像中与第三图像之间的投影关系满足第三成像设备和第二成像设备之间的空间投影关系的第四图像为第三图像。
在一些实施例中，对焦单元可以用于根据待识别目标的至少一部分的深度进行对第三成像设备进行对焦，即调节第三成像设备的像距。例如，对焦单元可以根据待识别目标的至少一部分与第三成像设备的距离从多个物距区间确定所属的物距区间，并根据所属的物距区间确定对应的对焦位置。
识别模块509可以用于基于第三图像，对待识别目标进行身份识别。在一些实施例中，识别模块可以用于对第三图像进行预处理、从第三图像中提取图像特征、对提取的特征进行特征编码、将提取的特征编码与预存的特征编码进行匹配以进行身份识别。
存储模块511可用于存储由成像设备采集的图像数据、待识别目标的身份信息、图像处理模型和/或算法等。例如，存储模块511可存储由第一成像设备采集的第一图像、由第二成像设备采集的第二图像。例如，存储模块511可存储预采集的多个目标的至少一部分的图像特征，用于身份识别。又例如，存储模块511可存储图像预处理、目标检测技术等算法。在一些实施例中，存储模块511包括内部存储设备、外部存储设备等。
本说明书实施例可能带来的有益效果包括但不限于：(1)可以基于候选目标的深度信息确定待识别目标并基于待识别目标的至少一部分(例如，人眼)的深度信息从当前第三成像设备(例如，虹膜摄像机)拍摄的图像中筛选出指定目标的第三图像进行身份识别(例如，虹膜识别)，从而可以避免误采集或采集到不应该采集到的对象，提高采集的图像质量以及效率，进一步提高基于虹膜、眼纹等生物特征进行身份识别的速度以及准确性；(2)可以通过增加第三成像设备中图像传感器的个数或旋转第三成像设备(例如，虹膜摄像机)中的图像传感器，改变垂直或水平方向的FOV(例如，可以获取更大的垂直方向FOV以便覆盖不同身高的人群)，从而达到不需要增加调节俯仰角的机械结构件；(3)第二成像设备可以采用结构光深度摄像头或者TOF深度摄像头， 可以有效降低对环境光的依赖，可以提高深度信息的精确度，从而提高待识别目标确定的准确度，进一步提高第三图像(例如，虹膜图像)采集的精确性以及质量，提高身份识别(例如，虹膜识别)的速度以及准确性；(4)可以根据第二成像设备采集的图像确定的候选目标的至少一部分(例如，人脸或人眼)的深度信息判断当前是否有待识别或待验证的目标，以此判断是否启动第三成像设备(例如，虹膜成像设备)，可有效避免第三成像设备不断采集图像以造成采集不到应该采集的目标。无需用户主动开启或关闭第三成像设备，提高用户体验；(5)可以基于第二成像设备采集的图像确定的待识别目标的至少一部分(例如，人眼)的深度信息确定待识别目标的至少一部分(例如，人眼)与第三成像设备(例如，虹膜摄像机)之间的距离，进一步可以根据该距离对第三成像设备针对待识别目标的至少一部分(例如，人眼)进行自动对焦，可以实现快速准确的自动对焦，提高第三成像设备采集的图像质量，并且第三成像设备采用音圈马达实现的自动对焦，可以避免使用复杂的步进电机驱动机械结构实现对焦。需要说明的是，不同实施例可能产生的有益效果不同，在不同的实施例里，可能产生的有益效果可以是以上任意一种或几种的组合，也可以是其他任何可能获得的有益效果。
图5所示的***及其模块可利用各种方式来实现。例如，在一些实施例中，***及其模块可以通过硬件、软件或者软件和硬件的结合来实现。其中，硬件部分可利用专用逻辑来实现；软件部分则可存储在存储器中，由适当的指令执行***，例如微处理器或者专用设计硬件来执行。本领域技术人员可以理解上述的方法和***可以使用计算机可执行指令和/或包含在处理器控制代码中来实现，例如在诸如磁盘、CD或DVD-ROM的载体介质、诸如只读存储器(固件)的可编程的存储器或者诸如光学或电子信号载体的数据载体上提供了这样的代码。本说明书的***及其模块不仅可以有诸如超大规模集成电路或门阵列、诸如逻辑芯片、晶体管等的半导体、或者诸如现场可编程门阵列、可编程逻辑设备等的可编程硬件设备的硬件电路实现，也可以用例如由各种类型的处理器所执行的软件实现，还可由上述硬件电路和软件的结合(例如，固件)来实现。
需要注意的是，以上对于身份识别装置500中的模块的描述，仅为描述方便，并不能把本说明书限制在所举实施例范围之内。可以理解，对于本领域的技术人员来说，在了解该***的原理后，可能在不背离这一原理的情况下，对各个模块进行任意组合，或者构成子***与其他模块连接。例如，在一些实施例中，例如，图5中披露的可以是一个***中的不同模块，也可以是一个模块实现上述的两个或两个以上模块的功能。例如，深度信息提取模块和待识别目标确定模块可以集成为一个模块。诸如此类的变形，均在本说明书的保护范围
上文已对基本概念做了描述，显然，对于本领域技术人员来说，上述详细披露仅仅作为示例，而并不构成对本说明书的限定。虽然此处并没有明确说明，本领域技术人员可能会对本说明书进行各种修改、改进和修正。该类修改、改进和修正在本说明书中被建议，所以该类修改、改进、修正仍属于本说明书示范实施例的精神和范围。
同时，本说明书使用了特定词语来描述本说明书的实施例。如“一个实施例”、“一实施例”、和/或“一些实施例”意指与本说明书至少一个实施例相关的某一特征、结构或特点。因此，应强调并注意的是，本说明书中在不同位置两次或多次提及的“一实施例”或“一个实施例”或“一个替代性实施例”并不一定是指同一实施例。此外，本说明书的一个或多个实施例中的某些特征、结构或特点可以进行适当的组合。
此外，本领域技术人员可以理解，本说明书的各方面可以通过若干具有可专利性的种类或情况进行说明和描述，包括任何新的和有用的工序、机器、产品或物质的组合，或对他们的任何新的和有用的改进。相应地，本说明书的各个方面可以完全由硬件执行、可以完全由软件(包括固件、常驻软件、微码等)执行、也可以由硬件和软件组合执行。以上硬件或软件均可被称为“数据块”、“模块”、“引擎”、“单元”、“组件”或“***”。此外，本说明书的各方面可能表现为位于一个或多个计算机可读介质中的计算机产品，该产品包括计算机可读程序编码。
计算机存储介质可能包含一个内含有计算机程序编码的传播数据信号，例如在基带上或作为载波的一部分。该传播信号可能有多种表现形式，包括电磁形式、光形式等，或合适的组合形式。计算机存储介质可以是除计算机可读存储介质之外的任何计算机可读介质，该介质可以通过连接至一个指令执行***、装置或设备以实现通讯、传播或传输供使用的程序。位于计算机存储介质上的程序编码可以通过任何合适的介质进行传播，包括无线电、电缆、光纤电缆、RF、或类似介质，或任何上述介质的组合。
本说明书各部分操作所需的计算机程序编码可以用任意一种或多种程序语言编写，包括面向对象编程语言如Java、Scala、Smalltalk、Eiffel、JADE、Emerald、C++、C#、VB.NET、Python等，常规程序化编程语言如C语言、Visual Basic、Fortran 2003、Perl、COBOL 2002、PHP、ABAP，动态编程语言如Python、Ruby和Groovy，或其他编程语言等。该程序编码可以完全在用户计算机上运行、或作为独立的软件包在用户计算机上运行、或部分在用户计算机上运行部分在远程计算机运行、或完全在远程计算机或服务器上运行。在后种情况下，远程计算机可以通过任何网络形式与用户计算机连接，比如局域网(LAN)或广域网(WAN)，或连接至外部计算机(例如通过因特网)， 或在云计算环境中，或作为服务使用如软件即服务(SaaS)。
此外，除非权利要求中明确说明，本说明书所述处理元素和序列的顺序、数字字母的使用、或其他名称的使用，并非用于限定本说明书流程和方法的顺序。尽管上述披露中通过各种示例讨论了一些目前认为有用的发明实施例，但应当理解的是，该类细节仅起到说明的目的，附加的权利要求并不仅限于披露的实施例，相反，权利要求旨在覆盖所有符合本说明书实施例实质和范围的修正和等价组合。例如，虽然以上所描述的***组件可以通过硬件设备实现，但是也可以只通过软件的解决方案得以实现，如在现有的服务器或移动设备上安装所描述的***。
同理，为了简化本说明书披露的表述，从而帮助对一个或多个发明实施例的理解，前文对本说明书实施例的描述中，有时会将多种特征归并至一个实施例、附图或对其的描述中。但是，这种披露方法并不意味着本说明书对象所需要的特征比权利要求中提及的特征多。实际上，实施例的特征要少于上述披露的单个实施例的全部特征。
一些实施例中使用了描述成分、属性数量的数字，应当理解的是，此类用于实施例描述的数字，在一些示例中使用了修饰词“大约”、“近似”或“大体上”来修饰。除非另外说明，“大约”、“近似”或“大体上”表明所述数字允许有±20％的变化。相应地，在一些实施例中，说明书和权利要求中使用的数值参数均为近似值，该近似值根据个别实施例所需特点可以发生改变。在一些实施例中，数值参数应考虑规定的有效数位并采用一般位数保留的方法。尽管本说明书一些实施例中用于确认其范围广度的数值域和参数为近似值，在具体实施例中，此类数值的设定在可行范围内尽可能精确。
针对本说明书引用的每个专利、专利申请、专利申请公开物和其他材料，如文章、书籍、说明书、出版物、文档等，特此将其全部内容并入本说明书作为参考。与本说明书内容不一致或产生冲突的申请历史文件除外，对本说明书权利要求最广范围有限制的文件(当前或之后附加于本说明书中的)也除外。需要说明的是，如果本说明书附属材料中的描述、定义、和/或术语的使用与本说明书所述内容有不一致或冲突的地方，以本说明书的描述、定义和/或术语的使用为准。
最后，应当理解的是，本说明书中所述实施例仅用以说明本说明书实施例的原则。其他的变形也可能属于本说明书的范围。因此，作为示例而非限制，本说明书实施例的替代配置可视为与本说明书的教导一致。相应地，本说明书的实施例不仅限于本说明书明确介绍和描述的实施例。

Claims (20)

  1. 一种身份识别方法，包括：
    获取由第一成像设备采集的第一图像，所述第一图像中包括一个或多个候选目标；
    获取由第二成像设备采集的第二图像，所述第二图像包括所述一个或多个候选目标中的至少一个候选目标的深度信息；
    基于所述第一图像与所述第二图像，提取所述一个或多个候选目标的深度信息；
    基于所述一个或多个候选目标的所述深度信息，从所述一个或多个候选目标中确定至少一个候选目标为待识别目标；
    基于所述待识别目标的至少一部分的深度信息获取由第三成像设备采集的第三图像，所述第三图像包括所述待识别目标的所述至少一部分；以及
    基于所述第三图像，对所述待识别目标进行身份识别。
  2. 根据权利要求1所述的方法，所述待识别目标包括面部，所述待识别目标的所述至少一部分包括虹膜、眼纹和眼周中的至少一个。
  3. 根据权利要求1或2所述的方法，所述第二成像设备包括一个或多个深度成像设备，所述深度成像设备包括结构光深度摄像头和飞行时间TOF深度摄像头中的至少一种。
  4. 根据权利要求1所述的方法，所述第三成像设备包括一个或多个图像传感器，所述一个或多个图像传感器安装于所述第三成像设备中使得第三成像设备的垂直视角或水平视角中的至少一个大于阈值。
  5. 根据权利要求1所述的方法，所述第三成像设备包括一个或多个图像传感器，所述一个或多个图像传感器可沿一个或多个自由度旋转，所述第三成像设备通过所述一个或多个图像传感器的旋转改变所述垂直视角或所述水平视角中的至少一个。
  6. 根据权利要求1所述的方法，基于所述待识别目标的至少一部分的深度信息，获取第三图像包括：
    判断所述待识别目标的所述至少一部分的所述深度信息是否满足条件；以及
    响应于所述待识别目标的所述至少一部分的所述深度信息满足所述条件，启动所述第三成像设备采集所述待识别目标的所述至少一部分的所述第三图像。
  7. 根据权利要求1所述的方法，基于所述待识别目标的至少一部分的深度信息，获取第三图像包括：
    基于所述待识别目标的所述至少一部分的所述深度信息确定所述待识别目标的所述至少一部分相对于所述第三成像设备的空间位置信息；以及
    基于所述空间位置信息，使所述第三成像设备对所述待识别目标的所述至少一部分进行对焦。
  8. 根据权利要求1所述的方法，基于所述待识别目标的至少一部分的深度信息，获取第三图像包括：
    获取所述第二成像设备与所述第三成像设备之间的几何映射关系；
    获取所述第三成像设备采集的一个或多个第四图像；以及
    基于所述几何映射关系以及所述待识别目标的所述至少一部分的所述深度信息，从所述一个或多个第四图像中确定所述第三图像。
  9. 根据权利要求1所述的方法，基于所述一个或多个候选目标的所述深度信息，从所述一个或多个候选目标中确定至少一个候选目标为待识别目标包括：
    基于所述一个或多个候选目标的所述深度信息确定所述一个或多个候选目标的空间位置关系；以及
    基于所述一个或多个候选目标的所述空间位置关系确定所述待识别目标。
  10. 一种身份识别装置，包括：
    候选目标图像获取模块，用于获取第一图像，所述第一图像中包括一个或多个候选目标；以及获取由第二成像设备采集的第二图像，所述第二图像包括所述一个或多个候选目标中的至少一个候选目标的深度信息；
    深度信息提取模块，用于基于所述第一图像与所述第二图像，从所述第二图像中提取所述一个或多个候选目标的深度信息；
    待识别目标确定模块，用于基于所述一个或多个候选目标的所述深度信息，从所述一个或多个候选目标中确定至少一个候选目标为待识别目标；以及
    待识别目标图像获取模块，用于基于所述待识别目标的至少一部分的深度信息获取由第三成像设备采集的第三图像，所述第三图像包括所述待识别目标的所述至少一部；以及
    识别模块，用于基于所述第三图像，对所述待识别目标进行身份识别。
  11. 根据权利要求10所述的装置，所述待识别目标包括面部，所述待识别目标的所述至少一部分包括虹膜、眼纹和眼周中的至少一个。
  12. 根据权利要求10或11所述的装置，所述第二成像设备包括一个或多个深度成像设备，所述深度成像设备包括结构光深度摄像头和飞行时间TOF深度摄像头中的至少一种。
  13. 根据权利要求10所述的装置，所述第三成像设备包括一个或多个图像传感器， 所述一个或多个图像传感器安装于所述第三成像设备中使得第三成像设备的垂直视角或水平视角中的至少一个大于阈值。
  14. 根据权利要求10所述的装置，所述第三成像设备包括一个或多个图像传感器，所述一个或多个图像传感器可沿一个或多个自由度旋转，所述第三成像设备通过所述一个或多个图像传感器的旋转改变所述垂直视角或所述水平视角中的至少一个。
  15. 根据权利要求10所述的装置，所述待识别目标图像获取模块进一步包括启动单元，用于：
    判断所述待识别目标的所述至少一部分的所述深度信息是否满足条件；以及
    响应于所述待识别目标的所述至少一部分的所述深度信息满足所述条件，启动所述第三成像设备采集所述待识别目标的所述至少一部分的所述第三图像。
  16. 根据权利要求10所述的装置，所述待识别目标图像获取模块进一步包括对焦单元，用于：
    基于所述待识别目标的所述至少一部分的所述深度信息确定所述待识别目标的所述至少一部分相对于所述第三成像设备的空间位置信息；以及
    基于所述空间位置信息，使所述第三成像设备对所述待识别目标的所述至少一部分进行对焦。
  17. 根据权利要求10所述的装置，所述待识别目标图像获取模块进一步包括筛选单元，用于：
    获取所述第二成像设备与所述第三成像设备之间的几何映射关系；
    获取所述第三成像设备采集的一个或多个第四图像；以及
    基于所述几何映射关系以及所述待识别目标的所述至少一部分的所述深度信息，从所述一个或多个第四图像中确定所述第三图像。
  18. 根据权利要求10所述的装置，所述待识别目标确定模块进一步用于：
    基于所述一个或多个候选目标的所述深度信息确定所述一个或多个候选目标的空间位置关系；以及
    基于所述一个或多个候选目标的所述空间位置关系确定所述待识别目标。
  19. 一种身份识别***，包括：
    第一成像设备，用于采集第一图像，所述第一图像中包括一个或多个候选目标；
    第二成像设备，用于采集第二图像，所述第二图像包括所述一个或多个候选目标中的至少一个候选目标的深度信息；
    第三成像设备，用于采集第三图像，所述第三图像包括所述一个或多个候选目标中 的至少一个候选目标的至少一部分；
    至少一个处理器；以及
    可执行的指令，所述可执行的指令由所述至少一个处理器执行，使所述***实现如权利要求1-9中任意一项所述的身份识别方法。
  20. 一种计算机可读存储介质，所述存储介质存储计算机指令，当所述计算机指令被处理器执行时实现如权利要求1至9中任意一项所述的身份识别方法。
PCT/CN2021/086266 2020-04-10 2021-04-09 身份识别 WO2021204267A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202010277300.X 2020-04-10
CN202010277300.XA CN111191644B (zh) 2020-04-10 2020-04-10 身份识别方法、***及装置

Publications (1)

Publication Number Publication Date
WO2021204267A1 true WO2021204267A1 (zh) 2021-10-14

Family

ID=70708731

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/086266 WO2021204267A1 (zh) 2020-04-10 2021-04-09 身份识别

Country Status (2)

Country Link
CN (1) CN111191644B (zh)
WO (1) WO2021204267A1 (zh)

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111191644B (zh) * 2020-04-10 2020-10-20 支付宝(杭州)信息技术有限公司 身份识别方法、***及装置
CN112200073B (zh) * 2020-06-19 2022-05-06 支付宝(杭州)信息技术有限公司 一种刷脸测温及核身的方法、装置和设备
CN113722692B (zh) * 2021-09-07 2022-09-02 墨奇科技（北京）有限公司 身份识别的装置及其方法
CN116843731A (zh) * 2022-03-23 2023-10-03 腾讯科技（深圳）有限公司 对象识别方法以及相关设备

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104732210A (zh) * 2015-03-17 2015-06-24 深圳超多维光电子有限公司 目标人脸跟踪方法及电子设备
US9934436B2 (en) * 2014-05-30 2018-04-03 Leidos Innovations Technology, Inc. System and method for 3D iris recognition
CN109753926A (zh) * 2018-12-29 2019-05-14 深圳三人行在线科技有限公司 一种虹膜识别的方法和设备
CN110472582A (zh) * 2019-08-16 2019-11-19 腾讯科技（深圳）有限公司 基于眼部识别的3d人脸识别方法、装置和终端
CN111191644A (zh) * 2020-04-10 2020-05-22 支付宝(杭州)信息技术有限公司 身份识别方法、***及装置

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20050084137A1 (en) * 2002-01-16 2005-04-21 Kim Dae-Hoon System and method for iris identification using stereoscopic face recognition
CN102855471B (zh) * 2012-08-01 2014-11-26 中国科学院自动化研究所 远距离虹膜智能成像装置及方法
CN105574525B (zh) * 2015-12-18 2019-04-26 天津中科虹星科技有限公司 一种复杂场景多模态生物特征图像获取方法及其装置

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9934436B2 (en) * 2014-05-30 2018-04-03 Leidos Innovations Technology, Inc. System and method for 3D iris recognition
CN104732210A (zh) * 2015-03-17 2015-06-24 深圳超多维光电子有限公司 目标人脸跟踪方法及电子设备
CN109753926A (zh) * 2018-12-29 2019-05-14 深圳三人行在线科技有限公司 一种虹膜识别的方法和设备
CN110472582A (zh) * 2019-08-16 2019-11-19 腾讯科技（深圳）有限公司 基于眼部识别的3d人脸识别方法、装置和终端
CN111191644A (zh) * 2020-04-10 2020-05-22 支付宝(杭州)信息技术有限公司 身份识别方法、***及装置

Also Published As

Publication number Publication date
CN111191644A (zh) 2020-05-22
CN111191644B (zh) 2020-10-20

Similar Documents

Publication Publication Date Title
WO2021204267A1 (zh) 身份识别
CN106446873B (zh) 人脸检测方法及装置
US11048953B2 (en) Systems and methods for facial liveness detection
TWI766201B (zh) 活體檢測方法、裝置以及儲存介質
KR101831973B1 (ko) 홍채 이미지 형성 장치 및 홍채 이미지 형성 장치의 구성 방법
US11227149B2 (en) Method and apparatus with liveness detection and object recognition
US9626553B2 (en) Object identification apparatus and object identification method
WO2016086343A1 (en) System and method for personal identification based on multimodal biometric information
US20150363670A1 (en) Image recognition method and camera system
CN112052831B (zh) 人脸检测的方法、装置和计算机存储介质
WO2016010721A1 (en) Multispectral eye analysis for identity authentication
WO2016010720A1 (en) Multispectral eye analysis for identity authentication
WO2016010724A1 (en) Multispectral eye analysis for identity authentication
CN109583304A (zh) 一种基于结构光模组的快速3d人脸点云生成方法及装置
CN104680128B (zh) 一种基于四维分析的生物特征识别方法和***
US9449217B1 (en) Image authentication
US11080557B2 (en) Image authentication apparatus, method, and storage medium using registered image
WO2016086341A1 (en) System and method for acquiring multimodal biometric information
US10915739B2 (en) Face recognition device, face recognition method, and computer readable storage medium
CN112052830B (zh) 人脸检测的方法、装置和计算机存储介质
JP6157165B2 (ja) 視線検出装置及び撮像装置
TWI509466B (zh) 物件辨識方法與裝置
KR20160037305A (ko) 지문 인식을 이용한 사용자 인증 방법
WO2018147059A1 (ja) 画像処理装置、および画像処理方法、並びにプログラム
US9282237B2 (en) Multifocal iris recognition device

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: 21783886

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: 21783886

Country of ref document: EP

Kind code of ref document: A1