WO2018192448A1 - 一种人证比对的认证方法、***及相机 - Google Patents

一种人证比对的认证方法、***及相机 Download PDF

Info

Publication number
WO2018192448A1
WO2018192448A1 PCT/CN2018/083217 CN2018083217W WO2018192448A1 WO 2018192448 A1 WO2018192448 A1 WO 2018192448A1 CN 2018083217 W CN2018083217 W CN 2018083217W WO 2018192448 A1 WO2018192448 A1 WO 2018192448A1
Authority
WO
WIPO (PCT)
Prior art keywords
image
target face
preset
similarity
authentication
Prior art date
Application number
PCT/CN2018/083217
Other languages
English (en)
French (fr)
Inventor
冯仁光
Original Assignee
杭州海康威视数字技术股份有限公司
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 杭州海康威视数字技术股份有限公司 filed Critical 杭州海康威视数字技术股份有限公司
Priority to US16/606,435 priority Critical patent/US11256902B2/en
Priority to EP18788454.9A priority patent/EP3614300A4/en
Publication of WO2018192448A1 publication Critical patent/WO2018192448A1/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
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/16—Human faces, e.g. facial parts, sketches or expressions
    • G06V40/172—Classification, e.g. identification
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/22—Matching criteria, e.g. proximity measures
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/16—Human faces, e.g. facial parts, sketches or expressions
    • G06V40/161—Detection; Localisation; Normalisation
    • G06V40/166—Detection; Localisation; Normalisation using acquisition arrangements
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/16—Human faces, e.g. facial parts, sketches or expressions
    • G06V40/168—Feature extraction; Face representation
    • G06V40/171—Local features and components; Facial parts ; Occluding parts, e.g. glasses; Geometrical relationships

Definitions

  • the present application relates to the field of image recognition technologies, and in particular, to a method, system and camera for authenticating a human witness.
  • the human witness comparison system in the related art takes a picture of the person through the camera, and sends the taken picture to the computer; the identity card information carried in the document is read by the certificate reader, and the identity information is sent. To the computer; the computer then authenticates based on the captured pictures and personal identification information.
  • the purpose of the embodiments of the present application is to provide a method, a system, and a camera for authenticating a person's license to achieve high real-time performance of personnel authentication through personal authentication.
  • the specific technical solutions are as follows:
  • an embodiment of the present application provides a method for authenticating a human witness, which is applied to a camera, and the method includes:
  • the number of target face images whose degree of similarity is higher than the preset similarity threshold reaches the first preset number, it is determined that the authentication is successful.
  • an embodiment of the present application provides a camera, where the camera includes:
  • a processor configured to acquire at least one target face image that meets a preset condition in the plurality of face images; receive a document image obtained by the accessory reading device; and separate the document image from each target face image The comparison is performed to obtain the similarity degree of the human face image of each target face image; if the number of the target face images whose human card similarity is higher than the preset similarity threshold reaches the first preset number, the authentication is determined to be successful.
  • an embodiment of the present application provides an authentication system for a human witness comparison, where the system includes:
  • a camera for acquiring a plurality of face images; acquiring at least one target face image of the plurality of face images that meets a preset condition; receiving a document image obtained by the accessory reading device; and displaying the document image
  • Each target face image is separately compared to obtain a similarity degree of the human face of each target face image; if the number of target face images whose human card similarity is higher than the preset similarity threshold reaches the first preset number, then it is determined Successful certification;
  • the document reading accessory device is configured to obtain a document image and send the document image to the camera.
  • the embodiment of the present application provides a storage medium for storing executable code, where the executable code is executed at runtime: the authentication of the human witness comparison provided by the first aspect of the embodiment of the present application method.
  • the embodiment of the present application provides an application program for performing the authentication method of the human witness comparison provided by the first aspect of the embodiment of the present application.
  • the authentication method, system and camera of the human witness comparison provided by the embodiment of the present invention compares at least one target face image and the document image of the collected plurality of face images by the camera to obtain the target face images.
  • the similarity of the person's card is determined according to the similarity of the person's card.
  • the certificate image is sent by the document reading accessory device.
  • the face image collection and the human witness comparison process are independently performed by the camera, which reduces the delay of the comparison, so that the real-time comparison between the human witness and the personnel authentication is higher.
  • the document image is often a frontal image
  • the target face image by selecting the target face image, and searching for the target face image with high similarity of human witnesses from the target face images, it is judged whether the person authentication is successful, and the judgment result is more To be accurate; when the comparison is performed, it is determined that the number of target face images whose human witness similarity is higher than the preset similarity threshold reaches the first preset number, and then the authentication is successful, and the authentication of the human witness is improved. Accuracy.
  • FIG. 1 is a schematic flow chart of a method for authenticating a person's license according to an embodiment of the present application
  • FIG. 2 is a schematic flowchart of a method for authenticating a person's license according to another embodiment of the present application
  • FIG. 3 is a schematic flowchart of a method for authenticating a person's license according to still another embodiment of the present application.
  • FIG. 4 is a side view of a camera installation of an example of an authentication method of a human witness comparison according to an embodiment of the present application
  • FIG. 5 is a schematic structural diagram of a camera according to an embodiment of the present application.
  • FIG. 6 is a schematic structural diagram of an authentication system for a human witness comparison according to an embodiment of the present application.
  • FIG. 7 is a schematic structural diagram of an authentication system for a human witness comparison according to another embodiment of the present application.
  • the embodiment of the present application provides a method, system, and camera for authenticating a person's license.
  • the execution body of the authentication method of the human witness authentication method provided by the embodiment of the present application may be a camera for collecting a face image, and may be a camera, a camera, a mobile phone, and the like having a photographing function.
  • the camera may include a processing chip with processing capability for performing human identification, and may include, for example, a DSP (Digital Signal Processor) and an ARM (Advanced Reduced Instruction Set Computer Machines). Processor processor) or core processing chip such as FPGA (Field-Programmable Gate Array).
  • the manner of implementing the authentication method of the human witness comparison provided by the embodiment of the present application may be at least one of software, hardware circuits, and logic circuits disposed in the execution body.
  • an authentication method for a human witness comparison may include the following steps:
  • S101 Acquire a plurality of face images, and acquire at least one target face image that meets a preset condition in the plurality of face images.
  • This embodiment is applicable to scenes such as access control and security check.
  • a camera is usually assumed at a specified position. Since the camera can be fixed or rotated, the camera can capture the fixed area or the moving area in real time. The image is then determined by the portrait recognition technology for the face image in all the images; the camera may also capture the face image in the fixed area or the moving area after detecting the human image information in the shooting area.
  • the camera can be set to be fixed and not rotated.
  • the camera collects a plurality of face images, and compares each face image and the document image one by one, so that a face image with a higher degree of similarity may exist a lot. Zhang, comprehensively determines whether the face image and the document image match.
  • the camera can be a camera or a camera
  • the camera can take a picture or a video. If a picture is taken, multiple face images are determined from the captured picture. For video, a multi-frame with face information is determined from the captured video as a face image.
  • the step of collecting multiple face images may include:
  • the region of interest of the face collection is determined according to the preset erection position and the preset shooting range.
  • the preset erection position is a erection position that satisfies the preset height range and the preset depression angle range.
  • the camera's erection position determines the shooting range, which directly affects the image analysis process.
  • the camera's erection position is related to the height and depression angle of the erection. For example, if the camera's erection height is set to 3m and the depression angle is set to 45°, the camera's shooting range is In a certain range of 3 m from the vertical position of the camera, the certain range mentioned here may be based on making the captured face image have higher definition, by adjusting the shooting parameters such as the focal length and aperture of the camera.
  • the door height is usually about 3 m, and the height of the person usually does not exceed 2 m. Therefore, the preset height range is usually set to 2 m to 4 m; in view of safety and It is convenient to follow the speed and effect of the comparison process. It is often necessary to make the shooting area farther from the camera.
  • the size of the collected face is similar to the size of the ID photo. Therefore, the preset depression angle range is usually set to 20° to 45°. .
  • the area of interest detection function ensures that there is only a single face in the camera's field of view.
  • the camera supports face detection in the region of interest, only the face inside the region of interest is photographed, and the person behind is prevented from entering the field of view.
  • the region of interest is related to the preset shooting range. In general, the preset shooting range can be selected as the camera.
  • the middle of the field of view accounts for 1/2 of the full screen. Of course, it can be 1/3 or 1/4 of the full screen. The smaller the ratio is, the smaller the chance of multiple people appearing in the region of interest, but if you set The ratio is too small and there may be situations where a complete face cannot be detected.
  • images in the region of interest are continuously acquired, and face images in all images are determined by the portrait recognition technique.
  • the camera may be a shooting video, and a plurality of frames having face features are determined as a face image from a region of interest of all frames of the video by a portrait recognition technology, or multiple photos may be continuously taken, and all photos are determined by the portrait recognition technology. Multiple face images with facial features in the region of interest.
  • the portrait recognition technology may be any existing human face recognition technology, for example, a convolutional neural network technology, a deep learning technology, etc., which are all protected by the embodiments of the present application, and will not be further described.
  • the collected plurality of face images may be first screened, and at least one target face image satisfying the preset condition is selected, wherein the preset condition may be a frontal, unoccluded, clear, and the like.
  • the step of acquiring at least one target facial image that meets a preset condition in the plurality of facial images may include:
  • the target face images satisfying the second preset number are acquired from the plurality of face images in descending order of the score.
  • the face scoring strategy refers to the strategy of judging the quality of the captured face image according to the rules of face angle and occlusion.
  • the pros and cons of the face scoring strategy directly affects the selection of the captured face image, which indirectly affects the result of the human face comparison. If the selected face image is not properly selected, for example, the face angle is relatively large or occluded. The face of the face will seriously affect the similarity of the person's identity. Therefore, it is usually possible to use the positive face, unobstructed, clear and other factors as the default face scoring strategy, that is, the face is more positive, unobstructed and the image is clear. The higher the face score.
  • a preset score threshold may be set, for example, 95 points, and all face images whose scores are greater than the preset score threshold are selected as the target face image; or a second preset number is set, for example Ten images were selected from the highest-rated face images in order of highest score to the target face image.
  • the document reading supporting device is used to obtain the ID image, and the document reading supporting device may be a second-generation identity card reader, a work permit card reader, etc., the image of the document in the document or the image of the certificate corresponding to the document information.
  • the document reading accessory device obtains the image of the document by directly capturing the image of the certificate on the document, or by extracting the document information, and extracting the image of the certificate corresponding to the document information from the information base.
  • the scope of protection of the embodiments of the present application is not specifically limited herein.
  • the ID reading accessory device can be directly connected to the camera through the transmission network, and the ID reading accessory device can also be connected to the client for display and connected to the camera through the client, wherein the transmission network can be a wired network. It can also be a wireless network.
  • the specific connection form of the wired network is not limited. It can be a USB connection, a serial connection, etc.
  • the client here has only a display function and does not have a control function.
  • the camera After receiving the ID image, the camera compares the ID image and each target facial image one by one, and compares the feature value or feature information of the feature point of the image, for example, the characteristics of the portrait feature such as glasses, nose, mouth, and face. Points can be scored according to the percentage system or percentage, which represents the similarity of the image. The higher the score, the closer the face image is to the document image.
  • the specific comparison process may be performed by using the similarity calculation of the feature points. Of course, other ways of achieving image comparison are also included in the protection scope of the embodiments of the present application, and are not described again.
  • the step of comparing the ID image with each target facial image to obtain the similarity of the human witness image of each target facial image may include:
  • face feature extraction is performed on the document image and each target face image, and the face image information of the document image and each target face image is obtained.
  • the facial feature information of the document image and the facial feature information of each target facial image are matched one by one, and the similarity of the human face of each target facial image is obtained.
  • the ID photo is usually a frontal photo
  • the facial feature is extracted from the target face image, and then the face feature is used to authenticate the image.
  • the image of each target face is compared and matched one by one, and the similarity of the person's face of each target face image is obtained.
  • the first preset number may be at least one. If the preset similarity threshold is set very high, for example, set to 95%, then as long as there is a target face image with a similarity of the person's card is higher than 95%, it can be determined that the person is successfully authenticated. However, if the preset similarity threshold is set to a relatively low level, for example, set to 85%, in order to avoid the effect of false detection on the result, you can set the preset number to multiple, if there are multiple (for example, 5) If the similarity of the human face of the target face image is greater than 85%, it can be determined that the person is successfully authenticated. After the personnel authentication is successful, the camera can send an authentication success indication to the user, or directly control the security device such as a gate.
  • the security device such as a gate.
  • the step of determining the success of the authentication may include:
  • the human witness similarity degree and the preset similarity threshold value of each target facial image are compared in sequence, and the number of target facial images whose statistical person similarity is greater than the preset similarity threshold is counted.
  • the number reaches the first preset number, it is determined that the authentication is successful.
  • the first preset number may be at least one. If the preset similarity threshold is set very high, for example, set to 95%, then as long as there is a target face image with a similarity of the person's card is higher than 95%, it can be determined that the person is successfully authenticated. However, if the preset similarity threshold is set to a relatively low level, for example, set to 85%, in order to avoid the effect of false detection on the result, you can set the preset number to multiple, if there are multiple (for example, 5) If the similarity of the human face of the target face image is greater than 85%, it can be determined that the person is successfully authenticated.
  • the document image is often a frontal image
  • the target face image by selecting the target face image, and searching for the target face image with high similarity of human witnesses from the target face images, it is judged whether the person authentication is successful, and the judgment result is more
  • the comparison when the comparison is performed, it is determined that the number of target face images in which all the human face similarities in the target face image are higher than the preset similarity threshold reaches the first preset number, and the authentication is determined to be successful, and the authentication is improved.
  • the authentication accuracy of the person's card comparison is performed.
  • the embodiment of the present application further provides a method for authenticating a person's license. After S104, the following steps may be further included:
  • S201 Send at least one of a target face image with the highest similarity of the person's identity, each target face image, a document image, and a result of successful authentication to the client, so that the client displays at least one of the following result information. :
  • the detection result of the camera may cause an error in detection
  • the detection result and the authentication result are transmitted to the client, and the client can display, for example, on the display screen.
  • the detection result of the target face image, the target face image, the document image and the like with the highest similarity of the person's card and the result of the successful authentication are provided for the user to make a second judgment. In this way, the user can determine whether the result of the successful authentication is correct by observing the display on the client, and whether the operation of the gate can be turned on for the person.
  • the client is a client that includes at least a display device.
  • Steps other than S201 shown in this embodiment may be performed with reference to S101 to S104 of the embodiment shown in FIG. 1, and details are not described herein again.
  • the document image is often a frontal image
  • the target face image by selecting the target face image, and searching for the target face image with high similarity of human witnesses from the target face images, it is judged whether the person authentication is successful, and the judgment result is more
  • the comparison when the comparison is performed, it is determined that the number of target face images in which all the human face similarities in the target face image are higher than the preset similarity threshold reaches the first preset number, and the authentication is determined to be successful, and the authentication is improved.
  • the client displays the detection result and the authentication result, which increases the human-computer interaction. When the machine judges that the authentication succeeds in error, the user can also make a second judgment to ensure the high accuracy of the personnel authentication.
  • the embodiment of the present application further provides a method for authenticating a person's license. After S103, the following steps may be further included:
  • the third preset number may be equal to the first preset number, or may be smaller than the first preset number.
  • the third preset number is one, if the target face image has no target face image whose degree of similarity is greater than or equal to the preset similarity threshold, the person authentication fails; the third preset number is more If the number of people in the target face image is greater than or equal to the preset similarity threshold does not reach the third preset number, the person authentication fails.
  • the camera may output an authentication failure information, and the authentication failure information may be an alarm information, a display information, or the like, and is used to prompt the client to fail the authentication.
  • the step of outputting the authentication failure information may include:
  • the client can display, on the display screen, a detection result such as a target face image with the highest similarity of the person, a target face image, a document image, and the result of the authentication failure, for the user to make a second judgment. In this way, the user can determine whether the result of the authentication failure is correct by observing the display on the client.
  • Steps other than S301 shown in this embodiment may be performed with reference to S101 to S104 of the embodiment shown in FIG. 1, and details are not described herein again.
  • the document image is often a frontal image
  • the target face image by selecting the target face image, and searching for the target face image with high similarity of human witnesses from the target face images, it is judged whether the person authentication is successful, and the judgment result is more
  • the comparison when the comparison is performed, it is determined that the number of target face images in which all the human face similarities in the target face image are higher than the preset similarity threshold reaches the first preset number, and the authentication is determined to be successful, and the authentication is improved.
  • the client displays the detection result and the authentication result, and the human-computer interaction is increased. When the machine determines that the authentication failure has an error, the user can perform the second judgment to ensure the high accuracy of the personnel authentication.
  • the embodiment of the present application further provides an embodiment of the authentication method of the human witness comparison, which may include all the steps of the embodiment shown in FIG. 2 and FIG. 3, wherein S101 to S104 and S201 may refer to FIG.
  • S101 to S104 and S201 may refer to FIG.
  • S301 can be performed with reference to the embodiment shown in FIG. 3, and details are not described herein again.
  • the authentication method of the human witness comparison provided by the embodiment of the present application is introduced in the following with reference to a specific application example.
  • the installation height h of the camera 410 is set to 3 m, and the depression angle ⁇ of the camera is 30°, and a camera that cannot be rotated is selected.
  • the person applying the punching scene to the office area sets the region of interest to be within the field of view of the camera, and is in the middle of the field of view, occupying 1/2 of the total field of view.
  • the employee scans his work permit on the work permit reader, the work permit reader obtains the corresponding ID image, the employee stands in the photo area and takes photos by the camera, the camera collects multiple face images, and adopts the preset face scoring strategy.
  • the camera compares each target face image with the document image one by one, and obtains the similarity of the human face of each target face image, and sets the similarity threshold to 90.
  • the target face image has 5 target face images whose degree of similarity is greater than the preset similarity threshold, and is equal to the first preset number of 5, and it is determined that the employee authentication is successful, that is, the employee punches successfully.
  • the solution compares at least one target face image and the document image in the collected plurality of face images by the camera, and obtains the similarity degree of the human face of each target face image, according to the similarity of the person witness. Determine whether the personnel authentication is successful, and the ID image is sent by the document reading accessory device.
  • the face image collection and the human witness comparison process are independently performed by the camera, which reduces the delay of the comparison, so that the real-time comparison between the human witness and the personnel authentication is higher.
  • the document image is often a frontal image
  • the target face image by selecting the target face image, and searching for the target face image with high similarity of human witnesses from the target face images, it is judged whether the person authentication is successful, and the judgment result is more
  • the comparison when the comparison is performed, it is determined that the number of target face images in which all the human face similarities in the target face image are higher than the preset similarity threshold reaches the first preset number, and the authentication is determined to be successful, and the authentication is improved.
  • the authentication accuracy of the person's card comparison is performed.
  • the embodiment of the present application provides a camera.
  • the camera may include:
  • a camera 510 configured to collect multiple face images
  • the processor 520 is configured to acquire at least one target facial image that meets a preset condition in the plurality of facial images; receive a document image obtained by the ID reading accessory device; and display the document image and each target facial image The comparison is performed separately, and the similarity degree of the human face image of each target face image is obtained; if the number of the target face images whose human card similarity is higher than the preset similarity threshold reaches the first preset number, the authentication is determined to be successful.
  • the document image is often a frontal image
  • the target face image by selecting the target face image, and searching for the target face image with high similarity of human witnesses from the target face images, it is judged whether the person authentication is successful, and the judgment result is more
  • the comparison when the comparison is performed, it is determined that the number of target face images in which all the human face similarities in the target face image are higher than the preset similarity threshold reaches the first preset number, and the authentication is determined to be successful, and the authentication is improved.
  • the authentication accuracy of the person's card comparison is performed.
  • the camera 510 is specifically configured to:
  • the preset erection position is a erection position that satisfies a preset height range and a preset depression angle range;
  • the images in the region of interest are continuously acquired, and the face images in all images are determined by the portrait recognition technique.
  • the processor 520 is specifically configured to:
  • the target face images satisfying the second preset number are acquired from the plurality of face images in descending order of the score.
  • the processor 520 is specifically configured to:
  • the target face image with the highest degree of similarity the target face image, the document image, and the result of the successful authentication.
  • the processor 520 is specifically configured to:
  • the face feature information of the document image and the face feature information of each target face image are feature-matched one by one, and the similarity of the person's identity of each target face image is obtained.
  • the processor 520 is specifically configured to:
  • the number reaches the first preset number, it is determined that the authentication is successful.
  • the processor 520 is specifically configured to:
  • the authentication failure information is output.
  • the processor 520 is specifically configured to:
  • the target face image with the highest degree of similarity the target face image, the document image, and the result of the authentication failure.
  • the camera of the embodiment of the present application is a camera applying the authentication method of the above-mentioned person authentication, and all the embodiments of the authentication method of the above-mentioned person authentication are applicable to the camera, and both can achieve the same or similar.
  • the benefits are not limited to the above-mentioned person authentication, and all the embodiments of the authentication method of the above-mentioned person authentication are applicable to the camera, and both can achieve the same or similar.
  • the embodiment of the present application provides an authentication system for comparing people's cards.
  • the authentication system for comparing people's cards may include:
  • a camera 610 configured to capture a face image; acquire at least one target face image that meets a preset condition in the plurality of face images; receive a document image obtained by the accessory reading device; and display the certificate image Comparing with each target face image to obtain the similarity degree of the human face image of each target face image; if the number of target face images whose human card similarity is higher than the preset similarity threshold reaches the first preset number, then Determine the success of the certification;
  • the document reading accessory device 620 is configured to obtain a document image and send the certificate image to the camera.
  • the document image is often a frontal image
  • the target face image by selecting the target face image, and searching for the target face image with high similarity of human witnesses from the target face images, it is judged whether the person authentication is successful, and the judgment result is more
  • the comparison when the comparison is performed, it is determined that the number of target face images in which all the human face similarities in the target face image are higher than the preset similarity threshold reaches the first preset number, and the authentication is determined to be successful, and the authentication is improved.
  • the authentication accuracy of the person's card comparison is performed.
  • the camera 610 is connected to the ID reading accessory device 620 through a connection port.
  • the camera 610 is specifically configured to:
  • the preset erection position is a erection position that satisfies a preset height range and a preset depression angle range;
  • the images in the region of interest are continuously acquired, and the face images in all images are determined by the portrait recognition technique.
  • the camera 610 is specifically configured to:
  • the target face images satisfying the second preset number are acquired from the plurality of face images in descending order of the score.
  • the camera 610 is specifically configured to:
  • the face feature information of the document image and the face feature information of each target face image are feature-matched one by one, and the similarity of the person's identity of each target face image is obtained.
  • the camera 610 is specifically configured to:
  • the number reaches the first preset number, it is determined that the authentication is successful.
  • the camera 610 is specifically configured to:
  • the authentication failure information is output.
  • the document reading accessory device 620 is specifically configured to:
  • the authentication method and the camera system using the above-mentioned personal authentication are applicable to The system can achieve the same or similar benefits.
  • the authentication system of the human witness comparison provided by the embodiment of the present application may further include:
  • the client 710 is configured to receive and display at least one of a face image sent by the camera 610, a result of the authentication, and a document image sent by the document reading accessory device 620.
  • the document image is often a frontal image
  • the target face image by selecting the target face image, and searching for the target face image with high similarity of human witnesses from the target face images, it is judged whether the person authentication is successful, and the judgment result is more
  • the comparison when the comparison is performed, it is determined that the number of target face images in which all the human face similarities in the target face image are higher than the preset similarity threshold reaches the first preset number, and the authentication is determined to be successful, and the authentication is improved.
  • the client displays the detection result and the authentication result, which increases the human-computer interaction. When the machine judges that an error occurs, the user can perform the second judgment to ensure the high accuracy of the personnel authentication.
  • the camera 610 is specifically configured to:
  • the client 710 can also be specifically used to:
  • At least one of the following result information is displayed: the target face image with the highest similarity of the person card, each target face image, the document image, and the result of the successful authentication.
  • the camera 610 is specifically configured to:
  • the client 710 can also be specifically used to:
  • At least one of the following result information is displayed: the target face image having the highest similarity of the person's card, each target face image, the document image, and the result of the authentication failure.
  • the camera 610 sends the target face image and the result of the authentication to the client 710 by using a USB transfer network card;
  • the credential reading accessory device 620 sends a credential image to the client 710 via a USB transfer network card.
  • the authentication method and the camera system using the above-mentioned personal authentication are applicable to The system can achieve the same or similar benefits.
  • the embodiment of the present application provides a storage medium for storing executable code, which is used to execute at runtime:
  • the number of target face images whose degree of similarity is higher than the preset similarity threshold reaches the first preset number, it is determined that the authentication is successful.
  • the storage medium stores an application that performs the authentication method of the human witness comparison provided by the embodiment of the present application at runtime, and thus can realize that the face image collection and the human witness comparison process are independently completed by the camera. The delay of the comparison is reduced, so that the real-time comparison between the human witness and the personnel authentication is higher.
  • the document image is often a frontal image
  • the target face image by selecting the target face image, and searching for the target face image with high similarity of human witnesses from the target face images, it is judged whether the person authentication is successful, and the judgment result is more To be accurate; when the comparison is performed, it is determined that the number of target face images whose human witness similarity is higher than the preset similarity threshold reaches the first preset number, and then the authentication is successful, and the authentication of the human witness is improved. Accuracy.
  • the embodiment of the present application provides an application program, which is executed at runtime: the authentication method of the human witness comparison provided by the embodiment of the present application Specifically, the authentication method for the human witness comparison provided by the embodiment of the present application includes:
  • the number of target face images whose degree of similarity is higher than the preset similarity threshold reaches the first preset number, it is determined that the authentication is successful.
  • the application performs the authentication method of the human witness comparison provided by the embodiment of the present application at runtime, so that the face image collection and the human witness comparison process are independently performed by the camera, and the comparison is reduced.
  • the delay makes the verification of personnel and the real-time of personnel certification higher.
  • the document image is often a frontal image
  • the target face image by selecting the target face image, and searching for the target face image with high similarity of human witnesses from the target face images, it is judged whether the person authentication is successful, and the judgment result is more To be accurate; when the comparison is performed, it is determined that the number of target face images whose human witness similarity is higher than the preset similarity threshold reaches the first preset number, and then the authentication is successful, and the authentication of the human witness is improved. Accuracy.

Landscapes

  • Engineering & Computer Science (AREA)
  • Oral & Maxillofacial Surgery (AREA)
  • Theoretical Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • General Health & Medical Sciences (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Human Computer Interaction (AREA)
  • Multimedia (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Data Mining & Analysis (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Artificial Intelligence (AREA)
  • Evolutionary Biology (AREA)
  • Evolutionary Computation (AREA)
  • General Engineering & Computer Science (AREA)
  • Collating Specific Patterns (AREA)
  • Image Analysis (AREA)

Abstract

一种人证比对的认证方法、***及相机，其中，人证比对的认证方法包括：采集多张人脸图像；并获取多张人脸图像中满足预设条件的至少一张目标人脸图像（S101）；接收证件读取配套设备获取的证件图像（S102）；将证件图像与各目标人脸图像分别进行比对，得到各目标人脸图像的人证相似度（S103）；如果人证相似度高于预设相似度阈值的目标人脸图像的数量达到第一预设数量，则确定认证成功（S104）。该方法通过人证比对实现人员认证的高实时性。

Description

一种人证比对的认证方法、***及相机
本申请要求于2017年04月20日提交中国专利局、申请号为201710263608.7发明名称为“一种人证比对的认证方法、***及相机”的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
技术领域
本申请涉及图像识别技术领域，特别是涉及一种人证比对的认证方法、***及相机。
背景技术
近年来，随着社会经济的快速发展，并且由于缺乏统一的信息共享平台和现代化的监管手段，各种身份证造假技术不断涌现，使得访客来访验证、出行人员验证等辨别人员与身份证上是否为同一个人的难度大大增加，给企业和社会治安带来很大的安全隐患。因此，通过对人证进行比对，确定人员身份尤为重要，传统的人证比对方法通过人肉眼对人员和身份证进行比对，肉眼判断极有可能判断错误，并且需要反复确认，效率极低。
针对上述问题，相关技术中的人证比对***通过摄像机对人员进行拍照，并将拍摄的图片发送至计算机；通过证件读卡器读取证件中携带的人员身份信息，并将人员身份信息发送至计算机；再由计算机根据拍摄的图片和人员身份信息，进行认证。
上述方案中，为了保证人证比对的准确率，需要拍摄高分辨率的图片，然后将该高分辨率的图片发送至计算机，由于图片的分辨率越高，则图片所占的内存就越大，使得在发送的过程中增加传输的时延，影响人证比对的效率，从而导致人员认证具有较低的实时性。
发明内容
本申请实施例的目的在于提供一种人证比对的认证方法、***及相机，以通过人证比对实现人员认证的高实时性。具体技术方案如下：
第一方面，本申请实施例提供了一种人证比对的认证方法，应用于相机，所述方法包括：
采集多张人脸图像；
获取所述多张人脸图像中满足预设条件的至少一张目标人脸图像；
接收证件读取配套设备获取的证件图像；
将所述证件图像与各目标人脸图像分别进行比对，得到各目标人脸图像的人证相似度；
如果人证相似度高于预设相似度阈值的目标人脸图像的数量达到第一预设数量，则确定认证成功。
第二方面，本申请实施例提供了一种相机，所述相机包括：
摄像头，用于采集多张人脸图像；
处理器，用于获取所述多张人脸图像中满足预设条件的至少一张目标人脸图像；接收证件读取配套设备获取的证件图像；将所述证件图像与各目标人脸图像分别进行比对，得到各目标人脸图像的人证相似度；如果人证相似度高于预设相似度阈值的目标人脸图像的数量达到第一预设数量，则确定认证成功。
第三方面，本申请实施例提供了一种人证比对的认证***，所述***包括：
相机，用于采集多张人脸图像；获取所述多张人脸图像中满足预设条件的至少一张目标人脸图像；接收证件读取配套设备获取的证件图像；将所述证件图像与各目标人脸图像分别进行比对，得到各目标人脸图像的人证相似度；如果人证相似度高于预设相似度阈值的目标人脸图像的数量达到第一预设数量，则确定认证成功；
证件读取配套设备，用于获取证件图像，并发送所述证件图像至所述相机。
第四方面，本申请实施例提供了一种存储介质，用于存储可执行代码，所述可执行代码用于在运行时执行：本申请实施例第一方面所提供的人证比对的认证方法。
第五方面，本申请实施例提供了一种应用程序，用于在运行时执行：本申请实施例第一方面所提供的人证比对的认证方法。
本申请实施例提供的人证比对的认证方法、***及相机，通过相机对采集的多张人脸图像中的至少一张目标人脸图像与证件图像进行比对，得到各目标人脸图像的人证相似度，根据人证相似度确定人员认证是否成功，其中，证件图像是由证件读取配套设备采集后发送来的。人脸图像的采集和人证比对流程由相机独立完成，减少了比对的延时，使得人证比对、人员认证的实时性更高。并且，基于证件图像往往为正面照的经验，通过选择目标人脸图像，再从这些目标人脸图像中查找人证相似度高的目标人脸图像，进而判断人员认证是否成功，使得判断结果更为精确；在进行比对时，需要确定人证相似度高于预设相似度阈值的目标人脸图像的数量达到第一预设数量时，才确定认证成功，提高了人证比对的认证准确率。
附图说明
为了更清楚地说明本申请实施例和现有技术的技术方案，下面对实施例和现有技术中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本申请的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其他的附图。
图1为本申请一实施例的人证比对的认证方法的流程示意图；
图2为本申请另一实施例的人证比对的认证方法的流程示意图；
图3为本申请再一实施例的人证比对的认证方法的流程示意图；
图4为本申请实施例的人证比对的认证方法的实例的相机安装侧视图；
图5为本申请实施例的相机的结构示意图；
图6为本申请一实施例的人证比对的认证***的结构示意图；
图7为本申请另一实施例的人证比对的认证***的结构示意图。
具体实施方式
为使本申请的目的、技术方案、及优点更加清楚明白，以下参照附图并举实施例，对本申请进一步详细说明。显然，所描述的实施例仅仅是本申请一部分实施例，而不是全部的实施例。基于本申请中的实施例，本领域普通技术人员在没有作出创造性劳动前提下所获得的所有其他实施例，都属于本 申请保护的范围。
为了实现人员认证的高实时性，本申请实施例提供了一种人证比对的认证方法、***及相机。
下面首先对本申请实施例所提供的一种人证比对的认证方法进行介绍。
需要说明的是，本申请实施例所提供的一种人证比对的认证方法的执行主体可以为一种相机，用于采集人脸图像，可以为摄像机、照相机、手机等具有拍摄功能的设备，其中，相机可以包括具有处理能力的处理芯片，用于进行人证比对，可以包括例如DSP(Digital Signal Processor，数字信号处理器)、ARM(Advanced Reduced Instruction Set Computer Machines，精简指令集计算机微处理器)或者FPGA(Field-Programmable Gate Array，现场可编程门阵列)等核心处理芯片。实现本申请实施例所提供的一种人证比对的认证方法的方式可以为设置于执行主体中的软件、硬件电路和逻辑电路中的至少一种。
如图1所示，本申请实施例所提供的一种人证比对的认证方法，可以包括如下步骤：
S101，采集多张人脸图像，并获取多张人脸图像中满足预设条件的至少一张目标人脸图像。
本实施例适用于门禁、安检等场景，在这些场景下，通常在指定位置处假设有相机，由于相机可以是固定的、也可以是转动的，则相机可以实时拍摄固定区域或移动区域内的图像，然后通过人像识别技术确定所有图像中的人脸图像；相机也可以是在检测到拍摄区域中有人像信息后，拍摄固定区域或移动区域内的人脸图像。这些情况均属于本申请实施例的保护范围。为了保证人证比对的效果，可以将相机设置为固定不旋转的。由于需要对人员和证件图像进行比对，如果只采集一张人脸图像，在人脸图像不清晰等无法准确判断人脸图像是否与证件图像相符时，只采集一张人脸图像会严重影响判断的结果。因此，为了保证判断结果的精确度，本实施例中，相机采集多张人脸图像，逐一比对每一张人脸图像和证件图像，这样，相似程度较高的人脸图像可能会存在很多张，综合确定人脸图像和证件图像是否相符。
由于相机可以为照相机，也可以为摄像机，则相机拍摄的可以是图片， 也可以是视频，如果拍摄到的是图片，则从拍摄到的图片中确定多张人脸图像，如果拍摄到的是视频，则从拍摄到的视频中确定具有人脸信息的多帧作为人脸图像。
可选的，采集多张人脸图像的步骤，可以包括：
第一步，根据预设架设位置及预设拍摄范围，确定人脸采集的感兴趣区域。
其中，预设架设位置为满足预设高度范围及预设俯角范围的架设位置。相机的架设位置决定着拍摄范围，直接影响了图像的分析过程，相机的架设位置与架设的高度和俯角有关，例如，相机的架设高度设置为3m，俯角设置为45°，则相机的拍摄范围为离相机的垂直位置外3m的一定范围内，这里提到的一定范围可以是基于使拍摄到的人脸图像具有较高的清晰度，通过调整相机的焦距、光圈等拍摄参数得到的。由于本实施例主要适用于门禁、安检等场景，门高往往为3m左右、且人的身高通常不会超过2m，因此，通常将预设高度范围设置为2m至4m；鉴于安全性、且为了便于后续人证比对过程的速度和效果，往往需要令拍摄区域距相机较远，采集到的人脸尺寸近似于证件照的尺寸，因此，通常将预设俯角范围设置为20°至45°。除了相机架设角度外，由于相机的视野内可能会存在多个人的情况，导致相机无法确定待比对人脸，因此需要保证相机的视野中只有单个人脸，因此，可以通过ROI(Region of Interest感兴趣区域)检测功能保证相机视野中只有单个人脸。例如相机支持感兴趣区域人脸检测，只拍摄感兴趣区域内部的人脸，避免后面的人进入视野，感兴趣区域与预设拍摄范围相关，一般情况下，预设拍摄范围可以选择为位于相机视野的正中间占全屏的1/2的范围，当然，占全屏1/3或1/4也可以，设置的比例越小，出现感兴趣区域中出现多个人的几率越小，但是，如果设置的比例太小，可能会出现无法检测到一个完整的人脸的情况。
第二步，连续采集感兴趣区域内的图像，并通过人像识别技术确定所有图像中的人脸图像。
相机可以是拍摄视频，通过人像识别技术从视频的所有帧的感兴趣区域中确定具有人脸特征的多个帧作为人脸图像，也可以是连续拍摄多张照片， 通过人像识别技术确定所有照片的感兴趣区域中具有人脸特征的多张人脸图像。其中，人像识别技术可以是现有的任一人像识别技术，例如，卷积神经网络技术、深度学习技术等，均属于本申请实施例的保护范围，不再一一赘述。
由于采集到的多张人脸图像中可能有正面的图像，也有可能有侧面的图像，还可能有遮挡严重的图像，这些图像中，例如遮挡严重的图像或者背面的图像对于人证比对的作用不大，因此可以先对采集到的多张人脸图像进行筛选，从中选择满足预设条件的至少一张目标人脸图像，其中，预设条件可以是正面、无遮挡、清晰等条件。
可选的，获取多张人脸图像中满足预设条件的至少一张目标人脸图像的步骤，可以包括：
根据预设人脸评分策略，获取多张人脸图像中至少一张评分高于预设评分阈值的目标人脸图像；
或者，
根据预设人脸评分策略，按照评分从高到低的顺序，从多张人脸图像中获取满足第二预设数量的目标人脸图像。
其中，人脸评分策略是指根据人脸角度、遮挡等规则评判拍摄得到的人脸图像的质量好坏的策略。人脸评分策略的优劣直接影响到拍摄的人脸图像的选取，间接影响到人证比对的结果，如果拍摄的人脸图像的选取不合适，比如选取到人脸角度比较大或者有遮挡的人脸，会严重影响到比对的人证相似度，因此，通常可以以正脸、无遮挡、清晰等因素作为预设人脸评分策略，即脸越正、无遮挡及图像清晰的人脸评分越高。为了提高比对的准确度，可以设置一个预设评分阈值，例如95分，选取所有评分大于该预设评分阈值的人脸图像作为目标人脸图像；或者设定一个第二预设数量，例如10张图像，按照评分从高到低的顺序，从评分最高的人脸图像开始选取10张人脸图像作为目标人脸图像。
S102，接收证件读取配套设备获取的证件图像。
证件读取配套设备用于获取证件图像，证件读取配套设备可以为二代身 份证读卡器、工作证读卡器等，证件图像指证件中或者是证件信息所对应的证件照的图像。证件读取配套设备对证件图像的获取，可以是直接拍摄证件上的证件图像，也可以是通过读取证件信息，从信息库中提取证件信息所对应的证件图像。这里不做具体限定，都属于本申请实施例的保护范围。可选地，证件读取配套设备可以直接通过传输网络与相机相连，证件读取配套设备也可以与用于显示的客户机相连再通过客户机与相机相连，其中，传输网络可以是有线网络，也可以是无线网络，有线网络的具体连接形式不作限定，可以为USB连接、串口连接等，这里的客户机只有显示功能，不具有控制功能。
S103，将证件图像与各目标人脸图像分别进行比对，得到各目标人脸图像的人证相似度。
相机在接收到证件图像后，对证件图像和各目标人脸图像一一进行比对，比较图像的特征点的特征值或特征信息，例如，眼镜、鼻子、嘴、脸型等具有人像特征的特征点，可以按照百分制或者百分比给比对的结果进行评分，代表图像的相似度，分值越高说明该张人脸图像与证件图像越相近。具体的比对过程可以采用特征点的相似度计算等方式，当然，其他能够实现图像比对的方式也属于本申请实施例的保护范围，不再赘述。
可选的，将证件图像与各目标人脸图像分别进行比对，得到每张目标人脸图像的人证相似度的步骤，可以包括：
第一步，对证件图像及各目标人脸图像进行人脸特征提取，得到证件图像及各目标人脸图像的人脸特征信息。
第二步，对证件图像的人脸特征信息与各目标人脸图像的人脸特征信息逐一进行特征匹配，得到各目标人脸图像的人证相似度。
由于证件照通常是正面照，则人脸图像中越是正脸的人脸图像与证件图像的相似度就越高，因此，对目标人脸图像进行人脸特征提取，再通过人脸特征对证件图像和各目标人脸图像一一比对、匹配，得到各目标人脸图像的人证相似度。
S104，如果人证相似度高于预设相似度阈值的目标人脸图像的数量达到 第一预设数量，则确定认证成功。
其中，第一预设数量可以为至少1张。如果预设相似度阈值设置的很高，例如设置为95％，则只要有一张目标人脸图像的人证相似度高于95％，则可确定该人员认证成功。但是，如果预设相似度阈值设置的相对低一点，例如设置为85％，为了避免误检测给结果带来的影响，可以设置预设数量为多张，则如果有多张(例如5张)目标人脸图像的人证相似度均大于85％，则可以确定该人员认证成功。人员认证成功后，相机可以发送认证成功指示给用户，也可以直接控制例如闸机等安防设备开启。
可选的，确定认证成功的步骤，可以包括：
依次比较各目标人脸图像的人证相似度与预设相似度阈值的大小，统计人证相似度大于预设相似度阈值的目标人脸图像的数量。
若该数量达到第一预设数量，则确定认证成功。
其中，第一预设数量可以为至少1张。如果预设相似度阈值设置的很高，例如设置为95％，则只要有一张目标人脸图像的人证相似度高于95％，则可确定该人员认证成功。但是，如果预设相似度阈值设置的相对低一点，例如设置为85％，为了避免误检测给结果带来的影响，可以设置预设数量为多张，则如果有多张(例如5张)目标人脸图像的人证相似度均大于85％，则可以确定该人员认证成功。
应用本实施例，通过相机对采集的多张人脸图像中的至少一张目标人脸图像与证件图像进行比对，得到各目标人脸图像的人证相似度，根据人证相似度确定人员认证是否成功，其中，证件图像是由证件读取配套设备采集后发送来的。人脸图像的采集和人证比对流程由相机独立完成，减少了比对的延时，使得人证比对、人员认证的实时性更高。并且，基于证件图像往往为正面照的经验，通过选择目标人脸图像，再从这些目标人脸图像中查找人证相似度高的目标人脸图像，进而判断人员认证是否成功，使得判断结果更为精确；在进行比对时，需要确定所有目标人脸图像中人证相似度高于预设相似度阈值的目标人脸图像的数量达到第一预设数量时，才确定认证成功，提高了人证比对的认证准确率。
基于图1所示实施例，如图2所示，本申请实施例还提供了一种人证比对的认证方法，在S104之后，还可以包括如下步骤：
S201，将人证相似度最高的目标人脸图像、各目标人脸图像、证件图像和认证成功的结果中的至少一项发送至客户机，以使客户机显示下列结果信息中的至少一项：
人证相似度最高的目标人脸图像、各目标人脸图像、证件图像、认证成功的结果。
由于相机的检测结果可能会出现检测出错的情况，为了提高控制的准确性，在相机确定检测结果和认证结果后，将检测结果和认证结果发送至客户机，客户机可以在显示屏上显示例如人证相似度最高的目标人脸图像、各目标人脸图像、证件图像等检测结果和认证成功的结果，供用户进行二次判断。这样，用户可以通过观察客户机上的显示，决定认证成功的结果是否正确，是否可以对该人员开启闸机等操作。其中，客户机为至少包含有显示设备的客户端。
本实施例中除所示的S201之外的步骤，可以参照图1所示实施例的S101至S104执行，这里不再一一赘述。
应用本实施例，通过相机对采集的多张人脸图像中的至少一张目标人脸图像与证件图像进行比对，得到各目标人脸图像的人证相似度，根据人证相似度确定人员认证是否成功，其中，证件图像是由证件读取配套设备采集后发送来的。人脸图像的采集和人证比对流程由相机独立完成，减少了比对的延时，使得人证比对、人员认证的实时性更高。并且，基于证件图像往往为正面照的经验，通过选择目标人脸图像，再从这些目标人脸图像中查找人证相似度高的目标人脸图像，进而判断人员认证是否成功，使得判断结果更为精确；在进行比对时，需要确定所有目标人脸图像中人证相似度高于预设相似度阈值的目标人脸图像的数量达到第一预设数量时，才确定认证成功，提高了人证比对的认证准确率。并且利用客户机显示检测结果和认证结果，增加了人机交互，在机器判断认证成功出现错误时，还可以由用户进行二次判 断，保证了人员认证的高准确度。
基于图1所示实施例，如图3所示，本申请实施例还提供了一种人证比对的认证方法，在S103之后，还可以包括如下步骤：
S301，如果人证相似度低于预设相似度阈值的目标人脸图像的数量少于第三预设数量，则确定认证失败，并输出认证失败信息。
其中，第三预设数量可以等于第一预设数量，也可以小于第一预设数量。在第三预设数量为1张时，如果目标人脸图像中没有人证相似度大于或等于预设相似度阈值的目标人脸图像，则该人员认证失败；在第三预设数量为多张时，如果目标人脸图像中人证相似度大于或等于预设相似度阈值的数量没有达到第三预设数量，则该人员认证失败。相机在确定认证失败后，可以输出认证失败信息，该认证失败信息可以为报警信息、显示信息等，用于提示客户端该人员认证失败。
可选的，输出认证失败信息的步骤，可以包括：
将人证相似度最高的目标人脸图像、各目标人脸图像、证件图像和认证失败的结果中的至少一项发送至客户机，以使客户机显示下列结果信息中的至少一项：
人证相似度最高的目标人脸图像、各目标人脸图像、证件图像、认证失败的结果。
客户机可以在显示屏上显示例如人证相似度最高的目标人脸图像、各目标人脸图像、证件图像等检测结果和认证失败的结果，供用户进行二次判断。这样，用户可以通过观察客户机上的显示，决定认证失败的结果是否正确。
本实施例中所示的S301之外的步骤，可以参照图1所示实施例的S101至S104执行，这里不再一一赘述。
应用本实施例，通过相机对采集的多张人脸图像中的至少一张目标人脸图像与证件图像进行比对，得到各目标人脸图像的人证相似度，根据人证相似度确定人员认证是否成功，其中，证件图像是由证件读取配套设备采集后 发送来的。人脸图像的采集和人证比对流程由相机独立完成，减少了比对的延时，使得人证比对、人员认证的实时性更高。并且，基于证件图像往往为正面照的经验，通过选择目标人脸图像，再从这些目标人脸图像中查找人证相似度高的目标人脸图像，进而判断人员认证是否成功，使得判断结果更为精确；在进行比对时，需要确定所有目标人脸图像中人证相似度高于预设相似度阈值的目标人脸图像的数量达到第一预设数量时，才确定认证成功，提高了人证比对的认证准确率。并且利用客户机显示检测结果和认证结果，增加了人机交互，在机器判断认证失败出现错误时，还可以由用户进行二次判断，保证了人员认证的高准确度。
需要说明的是，本申请实施例还提供一种人证比对的认证方法的实施例，可以包括图2及图3所示实施例的所有步骤，其中，S101至S104及S201可以参照图2所示实施例执行，S301可以参照图3所示实施例执行，这里不再一一赘述。
下面结合具体的应用实例，对本申请实施例所提供的人证比对的认证方法进行介绍。
如图4所示，为本申请实施例的应用实例中相机安装的侧视图，设定相机410的安装高度h为3m、相机的俯角α为30°，选取无法旋转的相机。
假定应用于办公区的人员打卡场景，设定感兴趣区域为相机视野范围内，处于视野范围正中间占总视野面积的1/2区域。员工在工作证读卡器上扫描自己的工作证，工作证读卡器获取到对应的证件图像，员工站在拍照区域内由相机拍照，相机采集到多张人脸图像，并通过预设人脸评分策略，得到评分高于95分的8张目标人脸图像，相机将各目标人脸图像与证件图像一一进行比对，得到各目标人脸图像的人证相似度，设定相似度阈值为90％，目标人脸图像中有5张人证相似度大于预设相似度阈值的目标人脸图像，且等于第一预设数量5张，则确定该员工认证成功，即该员工打卡成功。
相较于相关技术，本方案通过相机对采集的多张人脸图像中的至少一张目标人脸图像与证件图像进行比对，得到各目标人脸图像的人证相似度，根 据人证相似度确定人员认证是否成功，其中，证件图像是由证件读取配套设备采集后发送来的。人脸图像的采集和人证比对流程由相机独立完成，减少了比对的延时，使得人证比对、人员认证的实时性更高。并且，基于证件图像往往为正面照的经验，通过选择目标人脸图像，再从这些目标人脸图像中查找人证相似度高的目标人脸图像，进而判断人员认证是否成功，使得判断结果更为精确；在进行比对时，需要确定所有目标人脸图像中人证相似度高于预设相似度阈值的目标人脸图像的数量达到第一预设数量时，才确定认证成功，提高了人证比对的认证准确率。
相应于上述实施例，本申请实施例提供了一种相机，如图5所示，所述相机可以包括：
摄像头510，用于采集多张人脸图像；
处理器520，用于获取所述多张人脸图像中满足预设条件的至少一张目标人脸图像；接收证件读取配套设备获取的证件图像；将所述证件图像与各目标人脸图像分别进行比对，得到各目标人脸图像的人证相似度；如果人证相似度高于预设相似度阈值的目标人脸图像的数量达到第一预设数量，则确定认证成功。
应用本实施例，通过相机对采集的多张人脸图像中的至少一张目标人脸图像与证件图像进行比对，得到各目标人脸图像的人证相似度，根据人证相似度确定人员认证是否成功，其中，证件图像是由证件读取配套设备采集后发送来的。人脸图像的采集和人证比对流程由相机独立完成，减少了比对的延时，使得人证比对、人员认证的实时性更高。并且，基于证件图像往往为正面照的经验，通过选择目标人脸图像，再从这些目标人脸图像中查找人证相似度高的目标人脸图像，进而判断人员认证是否成功，使得判断结果更为精确；在进行比对时，需要确定所有目标人脸图像中人证相似度高于预设相似度阈值的目标人脸图像的数量达到第一预设数量时，才确定认证成功，提高了人证比对的认证准确率。
可选的，所述摄像头510，具体可以用于：
根据预设架设位置及预设拍摄范围，确定人脸采集的感兴趣区域，其中，所述预设架设位置为满足预设高度范围及预设俯角范围的架设位置；
连续采集所述感兴趣区域内的图像，并通过人像识别技术确定所有图像中的人脸图像。
可选的，所述处理器520，具体可以用于：
根据预设人脸评分策略，获取所述多张人脸图像中至少一张评分高于预设评分阈值的目标人脸图像；
或者，
根据预设人脸评分策略，按照评分从高到低的顺序，从所述多张人脸图像中获取满足第二预设数量的目标人脸图像。
可选的，所述处理器520，具体还可以用于：
将人证相似度最高的目标人脸图像、各目标人脸图像、所述证件图像和认证成功的结果中的至少一项发送至客户机，以使所述客户机显示下列结果信息中的至少一项：
所述人证相似度最高的目标人脸图像、各目标人脸图像、所述证件图像、所述认证成功的结果。
可选的，所述处理器520，具体还可以用于：
对所述证件图像及所有目标人脸图像进行人脸特征提取，得到所述证件图像及各目标人脸图像的人脸特征信息；
对所述证件图像的人脸特征信息与各目标人脸图像的人脸特征信息逐一进行特征匹配，得到各目标人脸图像的人证相似度。
可选的，所述处理器520，具体还可以用于：
依次比较各目标人脸图像的人证相似度与预设相似度阈值的大小，统计认证相似度大于所述预设相似度阈值的目标人脸图像的数量；
若所述数量达到第一预设数量，则确定认证成功。
可选的，所述处理器520，具体还可以用于：
如果人证相似度低于预设相似度阈值的目标人脸图像的数量少于第三预设数量，则确定认证失败，并输出认证失败信息。
可选的，所述处理器520，具体还用于：
将人证相似度最高的目标人脸图像、各目标人脸图像、所述证件图像和认证失败的结果中的至少一项发送至客户机，以使所述客户机显示下列结果信息中的至少一项：
所述人证相似度最高的目标人脸图像、各目标人脸图像、所述证件图像、所述认证失败的结果。
需要说明的是，本申请实施例的相机为应用上述人证比对的认证方法的相机，则上述人证比对的认证方法的所有实施例均适用于该相机，且均能达到相同或相似的有益效果。
相应于上述实施例，本申请实施例提供了一种人证比对的认证***，如图6所示，人证比对的认证***可以包括：
相机610，用于采集的张人脸图像；获取所述多张人脸图像中满足预设条件的至少一张目标人脸图像；接收证件读取配套设备获取的证件图像；将所述证件图像与各目标人脸图像分别进行比对，得到各目标人脸图像的人证相似度；如果人证相似度高于预设相似度阈值的目标人脸图像的数量达到第一预设数量，则确定认证成功；
证件读取配套设备620，用于获取证件图像，并发送所述证件图像至所述相机。
应用本实施例，通过相机对采集的多张人脸图像中的至少一张目标人脸图像与证件图像进行比对，得到各目标人脸图像的人证相似度，根据人证相似度确定人员认证是否成功，其中，证件图像是由证件读取配套设备采集后发送来的。人脸图像的采集和人证比对流程由相机独立完成，减少了比对的延时，使得人证比对、人员认证的实时性更高。并且，基于证件图像往往为 正面照的经验，通过选择目标人脸图像，再从这些目标人脸图像中查找人证相似度高的目标人脸图像，进而判断人员认证是否成功，使得判断结果更为精确；在进行比对时，需要确定所有目标人脸图像中人证相似度高于预设相似度阈值的目标人脸图像的数量达到第一预设数量时，才确定认证成功，提高了人证比对的认证准确率。
可选的，所述相机610与所述证件读取配套设备620通过连接端口相连。
可选的，所述相机610，具体可以用于：
根据预设架设位置及预设拍摄范围，确定人脸采集的感兴趣区域，其中，所述预设架设位置为满足预设高度范围及预设俯角范围的架设位置；
连续采集所述感兴趣区域内的图像，并通过人像识别技术确定所有图像中的人脸图像。
可选的，所述相机610，具体还可以用于：
根据预设人脸评分策略，获取所述多张人脸图像中至少一张评分高于预设评分阈值的目标人脸图像；
或者，
根据预设人脸评分策略，按照评分从高到低的顺序，从所述多张人脸图像中获取满足第二预设数量的目标人脸图像。
可选的，所述相机610，具体还可以用于：
对所述证件图像及各目标人脸图像进行人脸特征提取，得到所述证件图像及各目标人脸图像的人脸特征信息；
对所述证件图像的人脸特征信息与各目标人脸图像的人脸特征信息逐一进行特征匹配，得到各目标人脸图像的人证相似度。
可选的，所述相机610，具体还可以用于：
依次比较各目标人脸图像的人证相似度与预设相似度阈值的大小，统计人证相似度大于所述预设相似度阈值的目标人脸图像的数量；
若所述数量达到第一预设数量，则确定认证成功。
可选的，所述相机610，具体还可以用于：
如果人证相似度低于预设相似度阈值的目标人脸图像的数量少于第三预设数量，则确定认证失败，并输出认证失败信息。
可选的，所述证件读取配套设备620，具体可以用于：
采集并根据证件信息，从信息库中读取所述证件信息对应的人脸图像作为证件图像；
或者，
通过拍摄、提取所述证件中的人脸图像作为证件图像。
需要说明的是，本申请实施例的人证比对的认证***为应用上述人证比对的认证方法及相机的***，则上述人证比对的认证方法及相机的所有实施例均适用于该***，且均能达到相同或相似的有益效果。
更进一步的，在包含相机610、证件读取配套设备620的基础上，如图7所示，本申请实施例所提供的一种人证比对的认证***，还可以包括：
客户机710，用于接收并显示所述相机610发送的人脸图像、认证的结果和所述证件读取配套设备620发送的证件图像中的至少一项。
应用本实施例，通过相机对采集的多张人脸图像中的至少一张目标人脸图像与证件图像进行比对，得到各目标人脸图像的人证相似度，根据人证相似度确定人员认证是否成功，其中，证件图像是由证件读取配套设备采集后发送来的。人脸图像的采集和人证比对流程由相机独立完成，减少了比对的延时，使得人证比对、人员认证的实时性更高。并且，基于证件图像往往为正面照的经验，通过选择目标人脸图像，再从这些目标人脸图像中查找人证相似度高的目标人脸图像，进而判断人员认证是否成功，使得判断结果更为精确；在进行比对时，需要确定所有目标人脸图像中人证相似度高于预设相似度阈值的目标人脸图像的数量达到第一预设数量时，才确定认证成功，提高了人证比对的认证准确率。并且利用客户机显示检测结果和认证结果，增加了人机交互，在机器判断出现错误时，还可以由用户进行二次判断，保证 了人员认证的高准确度。
可选的，所述相机610，具体还可以用于：
将人证相似度最高的目标人脸图像、各目标人脸图像、所述证件图像和认证成功的结果中的至少一项发送至客户机710；
所述客户机710，具体还可以用于：
显示下列结果信息中的至少一项：所述人证相似度最高的目标人脸图像、各目标人脸图像、所述证件图像、所述认证成功的结果。
可选的，所述相机610，具体还可以用于：
将人证相似度最高的目标人脸图像、各目标人脸图像、所述证件图像和认证失败的结果中的至少一项发送至客户机710；
所述客户机710，具体还可以用于：
显示下列结果信息中的至少一项：所述人证相似度最高的目标人脸图像、各目标人脸图像、所述证件图像、所述认证失败的结果。
可选的，所述相机610通过USB转网卡向所述客户机710发送目标人脸图像及认证的结果；
所述证件读取配套设备620通过USB转网卡向所述客户机710发送证件图像。
需要说明的是，本申请实施例的人证比对的认证***为应用上述人证比对的认证方法及相机的***，则上述人证比对的认证方法及相机的所有实施例均适用于该***，且均能达到相同或相似的有益效果。
另外，相应于上述实施例所提供的人证比对的认证方法，本申请实施例提供了一种存储介质，用于存储可执行代码，所述可执行代码用于在运行时执行：本申请实施例所提供的人证比对的认证方法；具体的，所述人证比对的认证方法，包括：
采集多张人脸图像；
获取所述多张人脸图像中满足预设条件的至少一张目标人脸图像；
接收证件读取配套设备获取的证件图像；
将所述证件图像与各目标人脸图像分别进行比对，得到各目标人脸图像的人证相似度；
如果人证相似度高于预设相似度阈值的目标人脸图像的数量达到第一预设数量，则确定认证成功。
本实施例中，存储介质存储有在运行时执行本申请实施例所提供的人证比对的认证方法的应用程序，因此能够实现：人脸图像的采集和人证比对流程由相机独立完成，减少了比对的延时，使得人证比对、人员认证的实时性更高。并且，基于证件图像往往为正面照的经验，通过选择目标人脸图像，再从这些目标人脸图像中查找人证相似度高的目标人脸图像，进而判断人员认证是否成功，使得判断结果更为精确；在进行比对时，需要确定人证相似度高于预设相似度阈值的目标人脸图像的数量达到第一预设数量时，才确定认证成功，提高了人证比对的认证准确率。
另外，相应于上述实施例所提供的人证比对的认证方法，本申请实施例提供了一种应用程序，用于在运行时执行：本申请实施例所提供的人证比对的认证方法；具体的，本申请实施例所提供的人证比对的认证方法，包括：
采集多张人脸图像；
获取所述多张人脸图像中满足预设条件的至少一张目标人脸图像；
接收证件读取配套设备获取的证件图像；
将所述证件图像与各目标人脸图像分别进行比对，得到各目标人脸图像的人证相似度；
如果人证相似度高于预设相似度阈值的目标人脸图像的数量达到第一预设数量，则确定认证成功。
本实施例中，应用程序在运行时执行本申请实施例所提供的人证比对的认证方法，因此能够实现：人脸图像的采集和人证比对流程由相机独立完成， 减少了比对的延时，使得人证比对、人员认证的实时性更高。并且，基于证件图像往往为正面照的经验，通过选择目标人脸图像，再从这些目标人脸图像中查找人证相似度高的目标人脸图像，进而判断人员认证是否成功，使得判断结果更为精确；在进行比对时，需要确定人证相似度高于预设相似度阈值的目标人脸图像的数量达到第一预设数量时，才确定认证成功，提高了人证比对的认证准确率。
对于应用程序以及存储介质实施例而言，由于其所涉及的方法内容基本相似于前述的方法实施例，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。
需要说明的是，在本文中，诸如第一和第二等之类的关系术语仅仅用来将一个实体或者操作与另一个实体或操作区分开来，而不一定要求或者暗示这些实体或操作之间存在任何这种实际的关系或者顺序。而且，术语“包括”、“包含”或者其任何其他变体意在涵盖非排他性的包含，从而使得包括一系列要素的过程、方法、物品或者设备不仅包括那些要素，而且还包括没有明确列出的其他要素，或者是还包括为这种过程、方法、物品或者设备所固有的要素。在没有更多限制的情况下，由语句“包括一个……”限定的要素，并不排除在包括所述要素的过程、方法、物品或者设备中还存在另外的相同要素。
本说明书中的各个实施例均采用相关的方式描述，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***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。
以上所述仅为本申请的较佳实施例而已，并不用以限制本申请，凡在本申请的精神和原则之内，所做的任何修改、等同替换、改进等，均应包含在本申请保护的范围之内。

Claims (23)

  1. 一种人证比对的认证方法，其特征在于，应用于相机，所述方法包括：
    采集多张人脸图像；
    获取所述多张人脸图像中满足预设条件的至少一张目标人脸图像；
    接收证件读取配套设备获取的证件图像；
    将所述证件图像与各目标人脸图像分别进行比对，得到各目标人脸图像的人证相似度；
    如果人证相似度高于预设相似度阈值的目标人脸图像的数量达到第一预设数量，则确定认证成功。
  2. 根据权利要求1所述的方法，其特征在于，所述采集多张人脸图像，包括：
    根据预设架设位置及预设拍摄范围，确定人脸采集的感兴趣区域，其中，所述预设架设位置为满足预设高度范围及预设俯角范围的架设位置；
    连续采集所述感兴趣区域内的图像，并通过人像识别技术确定所有图像中的人脸图像。
  3. 根据权利要求1所述的方法，其特征在于，所述获取所述多张人脸图像中满足预设条件的至少一张目标人脸图像，包括：
    根据预设人脸评分策略，获取所述多张人脸图像中至少一张评分高于预设评分阈值的目标人脸图像；
    或者，
    根据预设人脸评分策略，按照评分从高到低的顺序，从所述多张人脸图像中获取满足第二预设数量的目标人脸图像。
  4. 根据权利要求1所述的方法，其特征在于，所述如果人证相似度高于预设相似度阈值的目标人脸图像的数量达到预设数量，则确定认证成功之后，所述方法还包括：
    将人证相似度最高的目标人脸图像、各目标人脸图像、所述证件图像和 认证成功的结果中的至少一项发送至客户机，以使所述客户机显示下列结果信息中的至少一项：
    所述人证相似度最高的目标人脸图像、各目标人脸图像、所述证件图像、所述认证成功的结果。
  5. 根据权利要求1所述的方法，其特征在于，所述将所述证件图像与各目标人脸图像分别进行比对，得到各目标人脸图像的人证相似度，包括：
    对所述证件图像及各目标人脸图像进行人脸特征提取，得到所述证件图像及各目标人脸图像的人脸特征信息；
    对所述证件图像的人脸特征信息与各目标人脸图像的人脸特征信息逐一进行特征匹配，得到各目标人脸图像的人证相似度。
  6. 根据权利要求1所述的方法，其特征在于，所述如果人证相似度高于预设相似度阈值的目标人脸图像的数量达到第一预设数量，则确定认证成功，包括：
    依次比较各目标人脸图像的人证相似度与预设相似度阈值的大小，统计人证相似度大于所述预设相似度阈值的目标人脸图像的数量；
    若所述数量达到第一预设数量，则确定认证成功。
  7. 根据权利要求1所述的方法，其特征在于，所述将所述证件图像与各目标人脸图像分别进行比对，得到各目标人脸图像的人证相似度之后，所述方法还包括：
    如果人证相似度低于预设相似度阈值的目标人脸图像的数量少于第三预设数量，则确定认证失败，并输出认证失败信息。
  8. 根据权利要求7所述的方法，其特征在于，所述输出认证失败信息，包括：
    将人证相似度最高的目标人脸图像、各目标人脸图像、所述证件图像和认证失败的结果中的至少一项发送至客户机，以使所述客户机显示下列结果信息中的至少一项：
    所述人证相似度最高的目标人脸图像、各目标人脸图像、所述证件图像、所述认证失败的结果。
  9. 一种相机，其特征在于，所述相机包括：
    摄像头，用于采集多张人脸图像；
    处理器，用于获取所述多张人脸图像中满足预设条件的至少一张目标人脸图像；接收证件读取配套设备获取的证件图像；将所述证件图像与各目标人脸图像分别进行比对，得到各目标人脸图像的人证相似度；如果人证相似度高于预设相似度阈值的目标人脸图像的数量达到第一预设数量，则确定认证成功。
  10. 根据权利要求9所述的相机，其特征在于，所述摄像头，具体用于：
    根据预设架设位置及预设拍摄范围，确定人脸采集的感兴趣区域，其中，所述预设架设位置为满足预设高度范围及预设俯角范围的架设位置；
    连续采集所述感兴趣区域内的图像，并通过人像识别技术确定所有图像中的人脸图像。
  11. 根据权利要求9所述的相机，其特征在于，所述处理器，具体用于：
    根据预设人脸评分策略，获取所述多张人脸图像中至少一张评分高于预设评分阈值的目标人脸图像；
    或者，
    根据预设人脸评分策略，按照评分从高到低的顺序，从所述多张人脸图像中获取满足第二预设数量的目标人脸图像。
  12. 根据权利要求9所述的相机，其特征在于，所述处理器，具体还用于：
    将人证相似度最高的目标人脸图像、各目标人脸图像、所述证件图像和认证成功的结果中的至少一项发送至客户机，以使所述客户机显示下列结果信息中的至少一项：
    所述人证相似度最高的目标人脸图像、各目标人脸图像、所述证件图像、 所述认证成功的结果。
  13. 根据权利要求9所述的相机，其特征在于，所述处理器，具体还用于：
    对所述证件图像及各目标人脸图像进行人脸特征提取，得到所述证件图像及各目标人脸图像的人脸特征信息；
    对所述证件图像的人脸特征信息与各目标人脸图像的人脸特征信息逐一进行特征匹配，得到各目标人脸图像的人证相似度。
  14. 根据权利要求9所述的相机，其特征在于，所述处理器，具体还用于：
    依次比较各目标人脸图像的人证相似度与预设相似度阈值的大小，统计认证相似度大于所述预设相似度阈值的目标人脸图像的数量；
    若所述数量达到第一预设数量，则确定认证成功。
  15. 根据权利要求9所述的相机，其特征在于，所述处理器，具体还用于：
    如果人证相似度低于预设相似度阈值的目标人脸图像的数量少于第三预设数量，则确定认证失败，并输出认证失败信息。
  16. 根据权利要求15所述的相机，其特征在于，所述处理器，具体还用于：
    将人证相似度最高的目标人脸图像、各目标人脸图像、所述证件图像和认证失败的结果中的至少一项发送至客户机，以使所述客户机显示下列结果信息中的至少一项：
    所述人证相似度最高的目标人脸图像、各目标人脸图像、所述证件图像、所述认证失败的结果。
  17. 一种人证比对的认证***，其特征在于，所述***包括：
    相机，用于采集多张人脸图像；获取所述多张人脸图像中满足预设条件的至少一张目标人脸图像；接收证件读取配套设备获取的证件图像；将所述 证件图像与各目标人脸图像分别进行比对，得到各目标人脸图像的人证相似度；如果人证相似度高于预设相似度阈值的目标人脸图像的数量达到第一预设数量，则确定认证成功；
    证件读取配套设备，用于获取证件图像，并发送所述证件图像至所述相机。
  18. 根据权利要求17所述的***，其特征在于，所述相机与所述证件读取配套设备通过连接端口相连。
  19. 根据权利要求17所述的***，其特征在于，所述***还包括：
    客户机，用于接收并显示所述相机发送的目标人脸图像、认证的结果和所述证件读取配套设备发送的证件图像中的至少一项。
  20. 根据权利要求19所述的***，其特征在于，所述相机通过USB转网卡向所述客户机发送目标人脸图像及认证的结果；
    所述证件读取配套设备通过USB转网卡向所述客户机发送证件图像。
  21. 根据权利要求17所述的***，其特征在于，所述证件读取配套设备，具体用于：
    采集并根据证件信息，从信息库中读取所述证件信息对应的人脸图像作为证件图像；
    或者，
    通过拍摄、提取所述证件中的人脸图像作为证件图像。
  22. 一种存储介质，其特征在于，用于存储可执行代码，所述可执行代码用于在运行时执行：权利要求1-8任一项所述的人证比对的认证方法。
  23. 一种应用程序，其特征在于，用于在运行时执行：权利要求1-8任一项所述的人证比对的认证方法。
PCT/CN2018/083217 2017-04-20 2018-04-16 一种人证比对的认证方法、***及相机 WO2018192448A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
US16/606,435 US11256902B2 (en) 2017-04-20 2018-04-16 People-credentials comparison authentication method, system and camera
EP18788454.9A EP3614300A4 (en) 2017-04-20 2018-04-16 METHOD, SYSTEM AND CAMERA FOR COMPARATIVE AUTHENTICATION OF PERSONS AND ACCESS DATA

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201710263608.7 2017-04-20
CN201710263608.7A CN108734067A (zh) 2017-04-20 2017-04-20 一种人证比对的认证方法、***及相机

Publications (1)

Publication Number Publication Date
WO2018192448A1 true WO2018192448A1 (zh) 2018-10-25

Family

ID=63855572

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2018/083217 WO2018192448A1 (zh) 2017-04-20 2018-04-16 一种人证比对的认证方法、***及相机

Country Status (4)

Country Link
US (1) US11256902B2 (zh)
EP (1) EP3614300A4 (zh)
CN (1) CN108734067A (zh)
WO (1) WO2018192448A1 (zh)

Families Citing this family (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US11157601B2 (en) * 2017-08-03 2021-10-26 Morphotrust Usa, Llc Electronic identity verification
CN108805005A (zh) * 2018-04-16 2018-11-13 深圳市商汤科技有限公司 身份验证方法和装置、电子设备、计算机程序和存储介质
US10817706B2 (en) * 2018-05-01 2020-10-27 Universal City Studios Llc System and method for facilitating throughput using facial recognition
EP3588364A1 (en) * 2018-06-21 2020-01-01 Onfido Ltd Within document face verification
CN110737789A (zh) * 2019-10-23 2020-01-31 中国工程物理研究院激光聚变研究中心 一种损伤排查方法、装置、电子设备及存储介质
CN111539252A (zh) * 2020-03-23 2020-08-14 深圳奇迹智慧网络有限公司 人员监测方法、装置、计算机设备和存储介质
CN113269120A (zh) * 2020-08-11 2021-08-17 支付宝实验室(新加坡)有限公司 用于人脸图像质量识别的方法、***和装置
CN112052780A (zh) * 2020-09-01 2020-12-08 北京嘀嘀无限科技发展有限公司 人脸验证方法、装置、***及存储介质
CN114554137A (zh) * 2020-11-24 2022-05-27 京东方科技集团股份有限公司 区域管控方法、装置、设备及存储介质
CN113158849A (zh) * 2021-04-07 2021-07-23 东华工程科技股份有限公司 一种基于5g的工业设备通信微处理器以及方法
CN113283359A (zh) * 2021-06-02 2021-08-20 万达信息股份有限公司 一种手持证件照的认证方法、***和电子设备
CN113627387A (zh) * 2021-08-30 2021-11-09 平安国际融资租赁有限公司 基于人脸识别的并行身份认证方法、装置、设备和介质

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20100329568A1 (en) * 2008-07-02 2010-12-30 C-True Ltd. Networked Face Recognition System
CN103309851A (zh) * 2013-05-10 2013-09-18 微梦创科网络科技（中国）有限公司 短文本的垃圾识别方法及***
CN103914686A (zh) * 2014-03-11 2014-07-09 辰通智能设备(深圳)有限公司 一种基于证件照与采集照的人脸比对认证方法及***
CN204155293U (zh) * 2014-09-01 2015-02-11 上海智达商投资管理合伙企业(有限合伙) 一种基于人脸识别的验证装置及验证***
CN105243357A (zh) * 2015-09-15 2016-01-13 深圳市环阳通信息技术有限公司 一种基于身份证件的人脸识别方法及人脸识别装置

Family Cites Families (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6128398A (en) * 1995-01-31 2000-10-03 Miros Inc. System, method and application for the recognition, verification and similarity ranking of facial or other object patterns
JP4374759B2 (ja) * 2000-10-13 2009-12-02 オムロン株式会社 画像比較システムおよび画像比較装置
US9082235B2 (en) 2011-07-12 2015-07-14 Microsoft Technology Licensing, Llc Using facial data for device authentication or subject identification
CN103473529B (zh) * 2013-08-26 2017-07-07 昆明学院 一种多视角取像人脸识别的方法及装置
JP6483485B2 (ja) 2015-03-13 2019-03-13 株式会社東芝 人物認証方法
US20160275518A1 (en) * 2015-03-19 2016-09-22 ecoATM, Inc. Device recycling systems with facial recognition
CN104794386A (zh) * 2015-04-08 2015-07-22 天脉聚源(北京)传媒科技有限公司 一种基于人脸识别的数据处理的方法及装置
CN104809794B (zh) * 2015-05-18 2017-04-12 苏州科达科技股份有限公司 门禁控制方法及***

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20100329568A1 (en) * 2008-07-02 2010-12-30 C-True Ltd. Networked Face Recognition System
CN103309851A (zh) * 2013-05-10 2013-09-18 微梦创科网络科技（中国）有限公司 短文本的垃圾识别方法及***
CN103914686A (zh) * 2014-03-11 2014-07-09 辰通智能设备(深圳)有限公司 一种基于证件照与采集照的人脸比对认证方法及***
CN204155293U (zh) * 2014-09-01 2015-02-11 上海智达商投资管理合伙企业(有限合伙) 一种基于人脸识别的验证装置及验证***
CN105243357A (zh) * 2015-09-15 2016-01-13 深圳市环阳通信息技术有限公司 一种基于身份证件的人脸识别方法及人脸识别装置

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP3614300A4

Also Published As

Publication number Publication date
EP3614300A4 (en) 2020-04-22
US11256902B2 (en) 2022-02-22
CN108734067A (zh) 2018-11-02
US20200042772A1 (en) 2020-02-06
EP3614300A1 (en) 2020-02-26

Similar Documents

Publication Publication Date Title
WO2018192448A1 (zh) 一种人证比对的认证方法、***及相机
US11256906B2 (en) Face verification method and apparatus
US10992666B2 (en) Identity verification method, terminal, and server
CN109948408B (zh) 活性测试方法和设备
CN106599772B (zh) 活体验证方法和装置及身份认证方法和装置
TWI706264B (zh) 證件的驗證方法、裝置、電子設備和儲存介質
WO2019149056A1 (zh) 一种身份验证方法、***、装置及设备
US20170262472A1 (en) Systems and methods for recognition of faces e.g. from mobile-device-generated images of faces
US8422746B2 (en) Face authentication system and authentication method thereof
CN105956572A (zh) 一种基于卷积神经网络的活体人脸检测方法
JP4760049B2 (ja) 顔認証装置、その顔認証方法、その顔認証装置を組み込んだ電子機器およびその顔認証プログラムを記録した記録媒体
WO2022062379A1 (zh) 图像检测方法和相关装置、设备、存储介质、计算机程序
WO2019200872A1 (zh) 身份验证方法和装置、电子设备、计算机程序和存储介质
WO2016172923A1 (zh) 视频检测方法、视频检测***以及计算机程序产品
KR101724971B1 (ko) 광각 카메라를 이용한 얼굴 인식 시스템 및 그를 이용한 얼굴 인식 방법
WO2019216091A1 (ja) 顔認証装置、顔認証方法および顔認証システム
US10970953B2 (en) Face authentication based smart access control system
CN109034029A (zh) 检测活体的人脸识别方法、可读存储介质和电子设备
CN106991448A (zh) 一种人像比对处理方法
KR20150069799A (ko) 얼굴 인증 방법 및 그 장치
WO2020007191A1 (zh) 活体识别检测方法、装置、介质及电子设备
CN113705428B (zh) 活体检测方法及装置、电子设备及计算机可读存储介质
JP6679291B2 (ja) 申請者認証装置、及び認証方法、並びにその方法を用いたセキュリティ認証システム
JP7424469B2 (ja) ゲートシステム、ゲート装置、その画像処理方法、およびプログラム、ならびに、ゲート装置の配置方法
WO2023095068A1 (en) A fool-proof registration system and methods thereof

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: 18788454

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

ENP Entry into the national phase

Ref document number: 2018788454

Country of ref document: EP

Effective date: 20191120