WO2020258304A1 - 门锁及识别方法 - Google Patents

门锁及识别方法 Download PDF

Info

Publication number
WO2020258304A1
WO2020258304A1 PCT/CN2019/093876 CN2019093876W WO2020258304A1 WO 2020258304 A1 WO2020258304 A1 WO 2020258304A1 CN 2019093876 W CN2019093876 W CN 2019093876W WO 2020258304 A1 WO2020258304 A1 WO 2020258304A1
Authority
WO
WIPO (PCT)
Prior art keywords
view
field
angle
dimensional
image
Prior art date
Application number
PCT/CN2019/093876
Other languages
English (en)
French (fr)
Inventor
王玉栋
Original Assignee
深圳市汇顶科技股份有限公司
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 深圳市汇顶科技股份有限公司 filed Critical 深圳市汇顶科技股份有限公司
Priority to CN201980001121.9A priority Critical patent/CN110462693B/zh
Priority to PCT/CN2019/093876 priority patent/WO2020258304A1/zh
Publication of WO2020258304A1 publication Critical patent/WO2020258304A1/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
    • G06V40/166—Detection; Localisation; Normalisation using acquisition arrangements
    • G—PHYSICS
    • G07—CHECKING-DEVICES
    • G07C—TIME OR ATTENDANCE REGISTERS; REGISTERING OR INDICATING THE WORKING OF MACHINES; GENERATING RANDOM NUMBERS; VOTING OR LOTTERY APPARATUS; ARRANGEMENTS, SYSTEMS OR APPARATUS FOR CHECKING NOT PROVIDED FOR ELSEWHERE
    • G07C9/00—Individual registration on entry or exit
    • G07C9/00174—Electronically operated locks; Circuits therefor; Nonmechanical keys therefor, e.g. passive or active electrical keys or other data carriers without mechanical keys
    • G07C9/00563—Electronically operated locks; Circuits therefor; Nonmechanical keys therefor, e.g. passive or active electrical keys or other data carriers without mechanical keys using personal physical data of the operator, e.g. finger prints, retinal images, voicepatterns

Definitions

  • the invention relates to the technical field of face recognition, in particular to a door lock and a recognition method.
  • FIG. 1 is an application scene diagram of an existing door lock
  • FIG. 2 is a field of view diagram of an existing door lock.
  • the installation height of the door lock is about 1.2 meters relative to the ground.
  • the height of 9-year-old children is generally about 1.2 meters, and the height of adults is mostly between 1.4 meters and 2 meters.
  • the standing distance relative to the door may be in the range of 0.3 m to 1 m.
  • the length of a human face is generally within 0.2 meters. Therefore, if you want all the above people to be able to use the door lock normally, the field of view of the structured light module needs to be at least 100°.
  • the field of view of existing door locks generally does not exceed 90°.
  • the field of view of the existing door lock cannot reach 100°, and the pictures taken when recognizing users of different ages cannot meet the recognition requirements.
  • the existing structured light module cannot 55,000 speckles are projected on the face, which in turn causes the captured pictures to fail to meet the recognition requirements.
  • This application provides a door lock and identification method, which aims to solve the technical problem that the pictures taken by the existing door locks when identifying users of different ages cannot meet the identification requirements, and when the user is far away, the existing The structured light module cannot project 55,000 speckles on the face, which leads to the technical problem that the captured pictures cannot meet the recognition requirements.
  • the present invention provides a door lock, including:
  • Three-dimensional recognition structure for taking the first image of a human face
  • a processor configured to generate a driving signal when the first image does not meet a preset recognition condition
  • the drive structure is used to drive the three-dimensional recognition structure to move under the drive of the drive signal to change the angle of view of the three-dimensional recognition structure so that the first image meets the preset recognition conditions.
  • the first image of the human face does not meet the preset recognition conditions specifically includes:
  • the relationship between the human face and the field of view of the three-dimensional recognition structure for recognition processing is that the face is outside the field of view of the three-dimensional recognition structure for recognition processing.
  • the fact that the human face is outside the field of view specifically includes:
  • the ratio of the portion of the human face in the field of view to the human face is less than a preset value
  • the key part of the face is in the field of view.
  • the first image of the human face does not meet the preset recognition conditions specifically includes:
  • the total number of speckles on the face is less than the preset number.
  • the number of speckles on the face of the plurality of first images is counted to obtain the total number of speckles.
  • the door lock further includes: a first housing; and the three-dimensional identification structure is hinged to the first housing.
  • the driving structure includes: a driver and an electromagnet; an adsorption block is provided on the three-dimensional recognition structure;
  • the driver energizes the electromagnet to generate an electromagnetic force between the electromagnet and the adsorption block to drive the movement of the three-dimensional recognition structure.
  • the electromagnet includes: an upper electromagnet and a lower electromagnet;
  • the adsorption block is located between the upper electromagnet and the lower electromagnet.
  • the driving structure includes a gear set and a motor; the three-dimensional identification structure is hinged with the first housing through a hinge shaft;
  • the gear at the end of the gear set is mounted on the articulated shaft, and the motor drives the gear set to move, so that the gear set drives the three-dimensional recognition structure to move.
  • the present invention provides an identification method, which is applied to a door lock, the door lock includes: a three-dimensional identification structure and a driving structure; the method includes:
  • determining whether the first image meets a preset recognition condition specifically includes:
  • the fact that the human face is outside the field of view specifically includes:
  • the ratio of the portion of the human face in the field of view to the human face is less than a preset value
  • the key part of the face is in the field of view.
  • moving the driving structure to drive the three-dimensional recognition structure specifically includes:
  • the current angle is the first angle or the second angle
  • the target angle is the first angle or the second angle
  • the target angle and the current angle are not the same.
  • the sum of the first field angle of the three-dimensional recognition structure and the second field angle of the three-dimensional recognition structure is greater than 100°;
  • the first field of view angle is the field of view angle when the three-dimensional recognition structure is located at the first angle
  • the second field of view angle is the field of view angle when the three-dimensional recognition structure is located at the second angle
  • the present application provides a door lock and an identification method.
  • the door lock includes a three-dimensional identification structure, a driving structure and a processor.
  • the processor determines whether the first image taken by the three-dimensional recognition structure meets the preset recognition condition, and when the first image does not meet the preset recognition condition, the angle of the three-dimensional recognition structure is changed so that the first image meets the preset recognition condition.
  • the door lock provided in this embodiment can adjust the angle of the three-dimensional recognition structure, so that the three-dimensional recognition structure can take pictures that meet the preset recognition conditions.
  • Figure 1 is an application scenario diagram of an existing door lock
  • Figure 2 is a view of a conventional door lock
  • Fig. 3 is a schematic structural diagram of a door lock according to an exemplary embodiment of the present invention.
  • FIG. 4 is a schematic view of a field of view of the door lock shown in the embodiment shown in FIG. 3;
  • FIG. 5 is another schematic view of the door lock based on the embodiment shown in FIG. 3;
  • Fig. 6 is a schematic structural diagram of a door lock according to another exemplary embodiment of the present invention.
  • FIG. 7 is a working principle diagram of the door lock shown in the embodiment shown in FIG. 6;
  • Fig. 8 is a schematic structural diagram of a door lock according to another exemplary embodiment of the present invention.
  • Fig. 9 is a schematic structural diagram of a door lock according to still another exemplary embodiment of the present invention.
  • FIG. 10 is a working principle diagram of the door lock shown in the embodiment shown in FIG. 9;
  • Fig. 11 is a schematic structural diagram of a door lock according to yet another exemplary embodiment of the present invention.
  • FIG. 12 is a working principle diagram of the door lock shown based on the embodiment shown in FIG. 11;
  • Fig. 13 is a schematic diagram showing an identification method according to an exemplary embodiment of the present invention.
  • Fig. 14 is a schematic diagram showing an identification method according to another exemplary embodiment of the present invention.
  • This application provides a door lock and identification method, which aims to solve the technical problem that the pictures taken by the existing door locks when identifying users of different ages cannot meet the identification requirements, and when the user is far away, the existing The structured light module cannot project 55,000 speckles on the face, which leads to the technical problem that the captured pictures cannot meet the recognition requirements.
  • Fig. 3 is a schematic structural diagram of a door lock according to an exemplary embodiment of the present invention.
  • the door lock 100 provided in this embodiment includes: a three-dimensional recognition structure 101, a driving structure 102 and a processor 103.
  • the above-mentioned three-dimensional recognition structure 101 is used to capture a first image of a human face.
  • the processor 103 is configured to determine whether the first image meets the preset recognition condition, and if the first image does not meet the preset recognition condition, generate a driving signal.
  • the drive structure drives the three-dimensional recognition structure to move under the drive of the drive signal to change the angle of view of the three-dimensional recognition structure so that the first image meets the preset recognition conditions.
  • the processor determines whether the first image taken by the three-dimensional recognition structure meets the preset recognition conditions, and when the first image does not meet the preset recognition conditions, the angle of the field of view of the three-dimensional recognition structure is changed so that The first image meets the preset recognition condition.
  • the door lock provided in this embodiment can adjust the angle of view of the three-dimensional recognition structure, so that the three-dimensional recognition structure can take pictures that meet preset recognition conditions.
  • FIG. 1 Another exemplary embodiment of the present invention shows a door lock including: a three-dimensional recognition structure, a driving structure, and a processor.
  • the above-mentioned door lock is installed on the door, the human face stands in front of the door, the three-dimensional recognition structure takes a first image of the human face, and the processor extracts the field of view relationship between the human face and the three-dimensional recognition structure for recognition processing according to the first image. If the face is within the field of view of the three-dimensional recognition structure for recognition processing, no drive signal is generated. If the face is outside the field of view of the three-dimensional recognition structure for recognition processing, a drive signal is generated.
  • the driving structure drives the three-dimensional recognition structure to move under the driving of the driving signal to change the angle of the field of view of the three-dimensional recognition structure, so that the human face is in the field of view of the three-dimensional recognition structure for recognition processing.
  • FIG. 4 is a schematic view of the door lock based on the embodiment shown in FIG. 3.
  • the angle between the three-dimensional recognition structure and the horizontal plane is ⁇ 1
  • the field of view angle of the three-dimensional recognition structure is ⁇
  • the installation height of the door lock is 1.2
  • the height of the face fluctuates from 1m to 2m
  • the face stands The distance between the position and the door lock fluctuates between 0.3m and 1m.
  • the three-dimensional recognition structure When the angle between the three-dimensional recognition structure and the horizontal plane is ⁇ 1 , assuming that the length of the face is 0.2m, the three-dimensional recognition structure can only recognize faces whose height exceeds 1.4m. When the standing position of the face is only 0.3m from the door lock, It may not be able to recognize faces of 1.4m in height. When the three-dimensional recognition structure detects that the face cannot be recognized, the angle of the field of view of the three-dimensional recognition structure is adjusted.
  • FIG. 5 is another schematic view of the door lock based on the embodiment shown in FIG. 3.
  • the driving structure in the door lock drives the three-dimensional recognition structure so that the angle between the three-dimensional recognition structure and the horizontal plane is ⁇ 2 .
  • the field of view angle of the three-dimensional recognition structure is still ⁇ , but due to the change of the angle between the three-dimensional recognition structure and the horizontal plane, the three-dimensional recognition structure can recognize the face of 1m in height and expand the field of view angle of the door lock.
  • the three-dimensional recognition structure is used to recognize the face.
  • the driving structure drives the movement of the three-dimensional recognition structure to change the field of view of the three-dimensional recognition structure Angle, so that the face is in the field of view of the three-dimensional recognition structure for recognition processing, thereby expanding the field of view of the door lock, and then the face of different ages can be recognized.
  • the structure of the door lock shown in another exemplary embodiment provided by the present invention is the same as the structure of the door lock in the embodiment shown in FIG. 3.
  • the three-dimensional recognition structure takes the first image of the first face, and the processor extracts the number of speckles on the face in the first image. If the number of speckles is greater than or equal to the preset number, no driving signal is generated. If the number of speckles is less than the preset number, a drive signal is generated.
  • the driving structure drives the 3D recognition structure to move under the drive of the driving signal to change the angle of view of the 3D recognition structure.
  • the 3D recognition structure takes the first image again under the new angle of view, and the processor extracts the face in the first image And count the number of speckles on the face in the multiple first images to obtain the total number of speckles on the face.
  • no driving signal is generated.
  • a drive signal is generated.
  • the driving structure drives the three-dimensional recognition structure to move under the driving of the driving signal to change the angle of view of the three-dimensional recognition structure until the total number of speckles is greater than or equal to the preset number.
  • the field of view angle of the three-dimensional recognition structure is changed by the driving structure, and multiple first pictures are taken, and then the number of speckles on the face in the multiple first pictures is counted, and the first pictures
  • the superposition method of the number of speckles on the face enables the number of speckles on the face to meet the preset recognition conditions, and finally multiple first pictures are used for face recognition.
  • Fig. 6 is a schematic structural diagram of a door lock according to another exemplary embodiment of the present invention.
  • the door lock 200 provided in this embodiment includes: a three-dimensional identification structure 201, a driving structure 202, a processor 203, a memory 204, and a first housing.
  • the driving structure 202 includes a driver 205 and an actuator 206.
  • the three-dimensional recognition structure 201 includes a second housing, an infrared sensor, an RGB sensor, a structured light projector, and a fill light.
  • the processor 203 is respectively connected with the three-dimensional recognition structure 201, the driver 205, and the memory 204, the driver 205 is connected with the actuator 206, and the actuator 206 is connected with the three-dimensional recognition structure 201.
  • the three-dimensional identification structure 201 is hinged to the first housing.
  • the infrared sensor, RGB sensor, structured light projector and fill light are all located inside the second housing.
  • the preset recognition condition is that the face is within the field of view of the three-dimensional recognition structure.
  • the infrared sensor or the RGB sensor in the three-dimensional recognition structure 201 captures the first image of the human face, and sends the first image to the processor 203. After receiving the first image, the processor 203 extracts the human face and the first image in the first image.
  • the positional relationship of the image. The positional relationship may be: the human face is at the middle position of the first image, and the human face is at the edge position of the first image. According to the position relationship, the relationship between the face and the field of view of the three-dimensional recognition structure is determined.
  • the relationship between the field of view of the face and the three-dimensional recognition structure for recognition processing can be divided into: the face is outside the field of view of the three-dimensional recognition structure or the face is within the field of view of the three-dimensional recognition structure.
  • Judging whether the face is within the field of view of the three-dimensional recognition structure can be: judging the size of the face area in the first image, or judging whether the key part of the face is in the first image, or judging the size and size of the face area at the same time Whether the key part is in the first image.
  • the first judgment method is: there are more than 2/3 of the face area in the first image, and then the first image can be directly used for face identification.
  • the second judgment method is that the facial features such as eyebrows, eyes, nose, and mouth are in the second image, and the second image can be directly used for face identification.
  • the third judgment method is: there are more than 2/3 of the face area in the second image, and the facial features of eyebrows, eyes, nose, and mouth are in the second image, you can directly use the second image Face recognition.
  • the processor obtains the three-dimensional recognition according to the first image
  • the current field of view angle of the structure and determine the target field of view angle of the three-dimensional recognition structure according to the current field of view angle and position relationship; to generate a drive signal according to the target field of view angle, the drive signal is used to control the drive structure to drive the three-dimensional recognition structure relative to the first
  • the housing rotates so that the three-dimensional recognition structure rotates from the current field of view angle to the target field of view angle, so that the human face enters the field of view of the three-dimensional recognition structure.
  • the door lock when the height of the face fluctuates between 1m and 2m, the distance between the face standing position and the door lock fluctuates between 0.3m and 1m, and the height of the door lock is 1.2m, the door lock The field of view of the door lock needs to reach 100°, and the field of view of the door lock is generally 60° to 90°. Therefore, only need to change the three-dimensional recognition structure and the horizontal plane angle once to expand the field of view angle of the door lock to 100 °.
  • the three-dimensional recognition structure changes only at two angles, namely the first angle and the second angle.
  • the sum of the first field of view angle of the three-dimensional recognition structure and the second field of view angle of the three-dimensional recognition structure is greater than 100°; wherein the first field of view angle is the field of view angle when the three-dimensional recognition structure is located at the first angle, and the second field of view angle
  • the field angle is the field of view angle when the three-dimensional recognition structure is located at the second angle.
  • Fig. 7 is a working principle diagram of the door lock shown based on the embodiment shown in Fig. 6.
  • the three-dimensional recognition structure takes a first image
  • the processor extracts the position relationship between the face in the first image and the first image, and determines the current field of view angle of the three-dimensional recognition structure according to the position relationship.
  • the face is located below the first image
  • it can be determined that the current field of view angle of the three-dimensional recognition structure is the first angle ⁇ 1
  • the current field of view angle of the three-dimensional recognition structure is the second angle ⁇ 2 .
  • the processor extracts the positional relationship between the face in the first image and the first image. And according to the position relationship, the relationship between the face and the three-dimensional recognition structure for recognition processing is determined. If the face is outside the field of view of the three-dimensional recognition structure, the processor obtains the current field of view angle of the three-dimensional recognition structure, and determines the target field of view angle of the three-dimensional recognition structure according to the current field of view angle and position relationship. If the current field of view angle of the three-dimensional recognition structure is the first angle ⁇ 1 and the face is located below the first image, the target field of view angle is the second angle ⁇ 2 , and the actuator drives the three-dimensional recognition structure to rotate to the second angle ⁇ 2 .
  • the actuator drives the three-dimensional recognition structure to rotate to the first angle ⁇ 1 .
  • the three-dimensional recognition structure first takes a first image, and the processor obtains the relationship between the human face and the three-dimensional recognition structure for recognition processing based on the first image.
  • the processor controls the driver, and the driver generates a drive signal to drive the actuator to move, so as to drive the three-dimensional recognition structure to rotate relative to the first housing and change the field of view angle of the three-dimensional recognition structure.
  • the field of view of the door lock is expanded, and the human faces of different ages can be recognized.
  • the structure of the door lock shown in another exemplary embodiment provided by the present invention is the same as the structure of the door lock in the embodiment shown in FIG. 6.
  • the infrared sensor or RGB sensor in the three-dimensional recognition structure takes the first image of the human face and sends the first image to the processor. After receiving the first image, the processor extracts the human face in the first image. The number of upper speckles. If the number of speckles is greater than or equal to the preset number, no driving signal is generated. If the number of speckles is less than the preset number, a drive signal is generated.
  • the drive structure drives the three-dimensional recognition structure to move under the drive of the drive signal to change the angle of view of the three-dimensional recognition structure, so that the total number of speckles on the face is greater than or equal to the preset number.
  • the processor controls the drive structure, and the drive structure generates a drive signal to drive the actuator to move, so as to drive the three-dimensional recognition structure to rotate relative to the first housing, and to change the angle of view of the three-dimensional recognition structure to make the face
  • the total number of speckles is greater than or equal to the preset number, and finally multiple first pictures are used for face recognition.
  • Fig. 8 is a schematic structural diagram of a door lock according to another exemplary embodiment of the present invention.
  • the door lock 300 provided in this embodiment includes: a three-dimensional recognition structure 301, a driving structure 302, a processor 303, a memory 304, and a first housing.
  • the driving structure 302 includes a driver 305 and an electromagnet 306.
  • the three-dimensional recognition structure 301 includes a second housing, an infrared sensor, an RGB sensor, a structured light projector, and a fill light.
  • the three-dimensional identification structure is hinged to the first housing, and the three-dimensional identification structure 301 is provided with an adsorption block.
  • the adsorption block may be located inside or outside the second housing.
  • the material of the adsorption block is a magnet or a material that can be adsorbed by a magnet.
  • the processor 303 is respectively connected to the three-dimensional recognition structure 301, the driver 305 and the memory 304, and the driver 305 is connected to the electromagnet 306.
  • the electromagnet 306 generates electromagnetic force with the adsorption block to drive the three-dimensional recognition structure 301 to move. .
  • the three-dimensional recognition structure takes a first picture, and the processor obtains the relationship between the human face and the three-dimensional recognition structure for recognition processing according to the first image. If the face is out of the field of view of the three-dimensional recognition structure, the processor controls the driver, and the driver energizes the electromagnet to generate electromagnetic force between the electromagnet and the adsorption block to drive the three-dimensional recognition structure to rotate relative to the first housing.
  • the angle of the field of view of the three-dimensional recognition structure makes the face in the field of view of the three-dimensional recognition structure.
  • the actuator is an electromagnet, which uses the electromagnetic force generated when the electromagnet is energized to drive the movement of the three-dimensional recognition structure, change the field of view angle of the three-dimensional recognition structure, and expand the field of view of the three-dimensional recognition structure.
  • the electromagnet reaction block can quickly switch the angle, saving time.
  • the structure of the door lock shown in another exemplary embodiment provided by the present invention is the same as the structure of the door lock in the embodiment shown in FIG. 8.
  • the infrared sensor or RGB sensor in the three-dimensional recognition structure takes the first image of the human face and sends the first image to the processor. After receiving the first image, the processor extracts the human face in the first image. The number of upper speckles. If the number of speckles is greater than or equal to the preset number, no driving signal is generated. If the number of speckles is less than the preset number, a drive signal is generated.
  • the driver energizes the electromagnet under the control of the drive signal to generate electromagnetic force between the electromagnet and the adsorption block to drive the three-dimensional recognition structure to rotate relative to the first housing, and change the angle of view of the three-dimensional recognition structure to make the face
  • the number of speckles is greater than or equal to the preset number.
  • the actuator is an electromagnet.
  • the electromagnetic force generated when the electromagnet is energized drives the movement of the three-dimensional recognition structure, changes the angle of view of the three-dimensional recognition structure, and increases the number of speckles on the face.
  • the iron reaction block can quickly switch the angle of view, saving time.
  • Fig. 9 is a schematic structural diagram of a door lock according to still another exemplary embodiment of the present invention.
  • the door lock 400 provided in this embodiment includes: a three-dimensional recognition structure 401, a driving structure 402, a processor 403, a memory 404, and a first housing.
  • the driving structure 402 includes a driver 405, an upper electromagnet 406, and a lower electromagnet 407.
  • the three-dimensional identification structure is hinged to the first housing, the three-dimensional identification structure 401 is provided with an adsorption block, and the adsorption block is located between the upper electromagnet 406 and the lower electromagnet 407.
  • the processor 403 is electrically connected to the three-dimensional identification structure 401, the driver 405, and the memory 404, and the driver 405 is electrically connected to the upper electromagnet 406 and the lower electromagnet 407, and the upper electromagnet 406 and the lower electromagnet 307 are both electrically connected.
  • the three-dimensional recognition structure 301 is driven to move by electromagnetic force.
  • Fig. 10 is a working principle diagram of the door lock shown based on the embodiment shown in Fig. 9.
  • the three-dimensional recognition structure takes a first picture, and the processor obtains the relationship between the human face and the three-dimensional recognition structure according to the first image to perform recognition processing. If the face is outside the field of view of the three-dimensional recognition structure, the processor obtains the current field of view angle of the three-dimensional recognition structure.
  • the target field of view angle of the three-dimensional recognition structure is determined according to the current field of view angle and position relationship, and the drive signal is generated according to the target field of view angle.
  • the driver energizes the lower electromagnet, so that the lower electromagnet generates electromagnetic force to drive the three-dimensional recognition structure to move from the first angle ⁇ 1 to the second angle ⁇ 2 . Furthermore, the angle of the field of view of the three-dimensional recognition structure is changed so that the face is within the range of the field of view of the three-dimensional recognition structure.
  • the processor controls the driver, and the driver energizes the upper electromagnet so that the upper electromagnet generates electromagnetic force to drive the three-dimensional recognition structure to move from the second angle ⁇ 2 to the first angle ⁇ 1 . Furthermore, the angle of the field of view of the three-dimensional recognition structure is changed so that the face is within the range of the field of view of the three-dimensional recognition structure.
  • the upper electromagnet and the lower electromagnet drive the door lock to switch between two angles.
  • the human face can be changed.
  • the control difficulty of the processor is reduced, and the door lock structure is simple.
  • Fig. 11 is a schematic structural diagram of a door lock according to yet another exemplary embodiment of the present invention.
  • the door lock 500 provided in this embodiment includes: a three-dimensional recognition structure 501, a driving structure 502, a processor 503, and a memory 504.
  • the driving structure 502 includes a motor 505 and a gear set 506.
  • the processor 503 is respectively connected to the three-dimensional recognition structure 501, the motor 502, and the memory 504.
  • the motor 505 is mechanically connected to the driving wheel of the gear set 506, and the driven wheel of the gear set 506 drives the three-dimensional recognition structure 501 to move.
  • Fig. 12 is a working principle diagram of the door lock shown based on the embodiment shown in Fig. 11.
  • the three-dimensional recognition structure takes a first picture
  • the processor obtains the relationship between the human face and the three-dimensional recognition structure for recognition processing according to the first image. If the face is outside the field of view of the three-dimensional recognition structure, the processor controls the motor, the motor rotates to drive the driving wheel to rotate, the driving wheel drives the driven wheel to rotate, and the driven wheel drives the three-dimensional recognition structure to move, changing the angle of view of the three-dimensional recognition structure Make the face in the field of view of the three-dimensional recognition structure.
  • the rotation angle of the driving wheel can be controlled by controlling the rotation time and speed of the motor, and then the rotation angle of the three-dimensional recognition structure.
  • the processor detects that no human face is in the field of view of the door lock, The processor obtains the current field of view angle of the three-dimensional recognition structure, determines the target field of view angle according to the current field of view angle and position relationship, and generates a drive signal according to the target field of view angle to control the rotation time and speed of the motor.
  • the actuator is a gear set.
  • the field of view angle of the three-dimensional recognition module can be accurately rotated from the current field of view to the target field of view, so that the face is in three-dimensional recognition.
  • Within the field of view of the module to adapt to face recognition of different age groups.
  • the structure of the door lock shown in another exemplary embodiment provided by the present invention is the same as the structure of the door lock in the embodiment shown in FIG. 9. The only difference is that after receiving the first image, the processor extracts the number of speckles on the human face in the first image. If the number of speckles is greater than or equal to the preset number, no driving signal is generated. If the number of speckles is less than the preset number, a drive signal is generated.
  • the drive signal controls the motor to rotate, and then the motor drives the driving wheel to rotate, the driving wheel drives the driven wheel to rotate, and the driven wheel drives the movement of the three-dimensional recognition structure, changing the angle of view of the three-dimensional recognition structure, so that the total number of speckles on the face is greater than or equal to
  • the preset number is finally used for face recognition using multiple first pictures.
  • Fig. 13 is a schematic diagram showing an identification method according to an exemplary embodiment of the present invention. As shown in FIG. 13, the identification method provided in this embodiment is applied to a door lock, and the door lock includes a three-dimensional identification structure. The identification method includes the following steps:
  • S602 Determine whether the first image meets the preset recognition condition; if the determination result is yes, go to S603, otherwise, go to S604.
  • the preset recognition condition is that the human face is within the field of view of the three-dimensional recognition structure for recognition processing.
  • the following method is used to obtain the relationship between the acquired face and the field of view of the three-dimensional recognition structure for recognition processing:
  • the first image of the face taken by the three-dimensional recognition structure and extract the area size of the face in the first image. If 2/3 of the face is in the first image, it can be judged that the face is in the field of view of the door lock Inside. It can also further extract the facial features of the face to determine whether the facial features are all in the first image. If the facial features are all in the first image, it can be determined that the features are within the field of view of the door lock. It can also extract the area size of the face and the facial features of the face, and determine that the facial features are in the first image, and if 2/3 of the faces are in the first image, then it can be determined that the features are in the door lock Within the field of view.
  • the preset recognition condition is whether the total number of speckles on the face meets the preset number. Extract the number of speckles on the face in each first image, and count the number of speckles on the face in the first images to obtain the total number of speckles on the face.
  • S603 Move the driving structure to drive the three-dimensional recognition structure to move, so that the first image meets the preset recognition condition.
  • the three-dimensional recognition structure is moved to change the angle of the field of view of the three-dimensional recognition structure, so that the human face is in the field of view of the three-dimensional recognition structure.
  • S604 Perform recognition processing using the first image.
  • the processor determines whether the first image taken by the three-dimensional recognition structure meets the preset recognition conditions, and when the first image does not meet the preset recognition conditions, it changes The angle of the field of view of the three-dimensional recognition structure makes the first image meet the preset recognition condition.
  • the door lock provided in this embodiment can adjust the angle of view of the three-dimensional recognition structure, so that the three-dimensional recognition structure can take pictures that meet preset recognition conditions.
  • Fig. 14 is a schematic diagram showing an identification method according to another exemplary embodiment of the present invention.
  • the height of the face fluctuates between 1m and 2m
  • the distance between the face's standing position and the door lock fluctuates between 0.2m and 1m
  • the height of the door lock is 1.2m
  • the field of view of the door lock needs to reach 100 °
  • the field of view of the door lock is between 60° and 90°. Therefore, it only needs to change the angle of the door lock and the horizontal plane once to expand the field of view angle of the door lock to 100°.
  • the identification method provided in this embodiment includes the following steps:
  • S702 Determine whether the first image meets the preset recognition condition; if the judgment result is yes, go to S707, otherwise, go to S703.
  • the current field of view angle of the three-dimensional recognition structure is determined according to the second position relationship.
  • the current field of view angle is the second angle.
  • the current field of view angle is the first angle.
  • the three-dimensional recognition structure has only two angles: the first angle or the second angle.
  • the first field of view angle is the field of view angle when the three-dimensional recognition structure is located at the first angle
  • the second field of view angle is the field of view angle when the three-dimensional recognition structure is located at the second angle.
  • the sum of the first field angle of the three-dimensional recognition structure and the second field angle of the three-dimensional recognition structure is greater than 100°.
  • the positional relationship includes the position of the face in the first image and the area size of the face in the first image.
  • the position of the face in the first image can be divided into a middle position and an edge position.
  • the size of the area of the face in the first image may be: the size of the key area of the face in the first image. Or the size of the entire face in the first image. For example: key areas such as eyes, nose, and mouth.
  • S705 Determine the target field of view angle of the three-dimensional recognition structure according to the current field of view angle and the position relationship.
  • the target field of view angle is the second angle. If the current field of view angle is the second angle and the face is located at the upper edge of the first image, the target field of view angle is the first angle.
  • S707 Perform face recognition using the first image.
  • the three-dimensional structure is switched between two angles.
  • the face can be within the field of view of the three-dimensional recognition structure, reducing the recognition method
  • the difficulty makes the door lock structure simple.
  • door locks in the text can also refer to various door lock systems installed in public places.
  • a person of ordinary skill in the art can understand that all or part of the steps in the foregoing method embodiments can be implemented by a program instructing relevant hardware.
  • the aforementioned program can be stored in a computer readable storage medium. When the program is executed, it executes the steps including the foregoing method embodiments; and the foregoing storage medium includes: ROM, RAM, magnetic disk, or optical disk and other media that can store program codes.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Health & Medical Sciences (AREA)
  • General Health & Medical Sciences (AREA)
  • Oral & Maxillofacial Surgery (AREA)
  • Human Computer Interaction (AREA)
  • Multimedia (AREA)
  • Theoretical Computer Science (AREA)
  • Lock And Its Accessories (AREA)

Abstract

一种门锁及方法，门锁包括：三维识别结构（101），用于拍摄人脸的第一图像；处理器（103），用于当所述第一图像不符合预设识别条件时生成驱动信号；驱动结构（102），用于在所述驱动信号驱动下带动所述三维识别结构（101）运动，以改变所述三维识别结构（101）的角度，以使所述第一图像符合预设识别条件。处理器（103）判断三维识别结构（101）拍摄的第一图像是否满足预设识别条件，当第一图像不满足预设识别条件时，则改变三维识别结构（101）的角度，使得第一图像满足预设识别条件。能够调整三维识别结构（101）的角度，使得三维识别结构（101）能够拍摄出满足预设识别条件的图片。

Description

门锁及识别方法 技术领域
本发明涉及人脸识别技术领域，尤其涉及一种门锁及识别方法。
背景技术
随着3D人脸识别方案的兴起，越来越多产品开始增加2D或3D人脸识别功能，例如智能门锁。
图1为现有的门锁的应用场景图，图2为现有的门锁的视场图。如图1和图2所示，门锁的安装高度相对与地面约1.2米，9岁小朋友的身高普遍在1.2m左右，成年人的身高大多数在1.4米到2米之间。相对与门的站立距离可能分布在0.3米到1米的范围。人脸的长度一般在0.2米以内。所以，想要所有以上人群能够正常使用门锁，则结构光模块的视场至少需要在100°。而现有的门锁的视场一般不超过90°。另外，通常情况下，人脸上至少需要300个散斑才能有效重建3D人脸。人脸的大小一般在0.025～0.04平方米。所以在1米位置处时，人脸仅占整个市场面积的0.4％左右。所以在1米位置处至少需要55000个散斑才能有效重建3D人脸。
然而，现有的门锁的视场无法达到100°，在对不同年龄层的用户进行识别时所拍摄的图片不能满足识别要求，另外，当用户距离较远时，现有的结构光模块无法在人脸上投射55000个散斑，进而导致所拍摄图片不能满足识别要求。
发明内容
本申请提供一种门锁及识别方法，旨在解决现有门锁在对不同年龄层的用户进行识别时所拍摄的图片不能满足识别要求的技术问题，以及当用户距离较远时，现有的结构光模块无法在人脸上投射55000个散斑，进而导致所拍摄图片不能满足识别要求的技术问题。
第一方面，本发明提供一种门锁，包括：
三维识别结构，用于拍摄人脸的第一图像；
处理器，用于当第一图像不符合预设识别条件时生成驱动信号；
驱动结构，用于在驱动信号驱动下带动三维识别结构运动，以改变三维识别结构的视场角度，以使第一图像符合预设识别条件。
可选地，人脸的第一图像不符合预设识别条件具体包括：
人脸与三维识别结构进行识别处理的视场的关系为人脸处于三维识别结构进行识别处理的视场之外。
可选地，所述人脸处于所述视场之外具体包括：
处于所述视场内的人脸部分与所述人脸的比值小于预设数值；和/或
所述人脸的关键部分处于所述视场内。
可选地，人脸的第一图像不符合预设识别条件具体包括：
人脸上的总散斑数量少于预设数量。
可选地，统计多张第一图像中人脸上的散斑数量，以获得总散斑数量。
可选地，门锁还包括：第一外壳；三维识别结构与第一外壳铰接。
可选地，驱动结构包括：驱动器和电磁铁；三维识别结构上设有吸附块；
驱动器使电磁铁通电，以使电磁铁与吸附块之间产生电磁力，以带动三维识别结构运动。
可选地，电磁铁包括：上电磁铁和下电磁铁；
吸附块位于上电磁铁和下电磁铁之间。
可选地，驱动结构包括齿轮组和电机；三维识别结构通过铰接轴与第一外壳铰接；
齿轮组中处于末端的齿轮安装于铰接轴上，电机驱动齿轮组运动，以使齿轮组带动三维识别结构运动。
第二方面，本发明提供一种识别方法，应用于门锁，门锁包括：三维识别结构和驱动结构；方法包括：
获取三维识别结构拍摄的人脸的第一图像；
判断第一图像是否满足预设识别条件；
若判断结果为否，使驱动结构运动，以带动三维识别结构，以使第一图像符合预设识别条件；
利用第一图像进行识别处理。
可选地，判断第一图像是否满足预设识别条件，具体包括：
判断人脸与三维识别结构进行识别处理的视场的关系是否为人脸处于三维识别结构进行识别处理的视场之内。
可选地，所述人脸处于所述视场之外具体包括：
处于所述视场内的人脸部分与所述人脸的比值小于预设数值；和/或
所述人脸的关键部分处于所述视场内。
可选地，使驱动结构运动，以带动三维识别结构，具体包括：
根据第一图像获取三维识别结构的当前角度；
提取第一图像中人脸与第一图像的位置关系；
根据当前角度和位置关系确定三维识别结构的目标角度；
使动三维识别结构运动，以使三维识别结构由当前角度运动至目标角度。
可选地，当前角度为第一角度或者第二角度；目标角度为第一角度或者第二角度；且目标角度和当前角度不相同。
可选地，三维识别结构的第一视场角度和三维识别结构的第二视场角度之和大于100°；
其中，第一视场角度为三维识别结构位于第一角度时的视场角度，第二视场角度为三维识别结构位于第二角度时的视场角度。
本申请提供一种门锁及识别方法，该门锁包括三维识别结构、驱动结构以及处理器。处理器判断三维识别结构拍摄的第一图像是否满足预设识别条件，当第一图像不满足预设识别条件时，则改变三维识别结构的角度，使得第一图像满足预设识别条件。相较于现有的门锁，本实施例提供的门锁能够调整三维识别结构的角度，使得三维识别结构能够拍摄出满足预设识别条件的图片。
附图说明
为了更清楚地说明本申请实施例或现有技术中的技术方案，下面将对实施例或现有技术描述中所需要使用的附图作一简单地介绍，显而易见地，下面描述中的附图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性的前提下，还可以根据这些附图获得其他的附图。
图1为现有的门锁的应用场景图；
图2为现有的门锁的视场图；
图3为本发明根据一示例性实施例示出的门锁的结构示意图；
图4为基于图3所示实施例示出的门锁的一视场示意图；
图5为基于图3所示实施例示出的门锁的另一视场示意图；
图6为本发明根据另一示例性实施例示出的门锁的结构示意图；
图7为基于图6所示实施例示出的门锁的工作原理图；
图8为本发明根据又一示例性实施例示出的门锁的结构示意图；
图9为本发明根据再一示例性实施例示出的门锁的结构示意图；
图10为基于图9所示实施例示出的门锁的工作原理图；
图11为本发明根据再又一示例性实施例示出的门锁的结构示意图；
图12为基于图11所示实施例示出的门锁的工作原理图；
图13为本发明根据一示例性实施例示出的识别方法的示意图；
图14为本发明根据另一示例性实施例示出的识别方法的示意图。
具体实施方式
为使本申请实施例的目的、技术方案和优点更加清楚，下面将结合本申请实施例中的附图，对本申请实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例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。
本申请的说明书和权利要求书及上述附图中的术语“第一”、“第二”、“第三”、“第四”等(如果存在)是用于区别类似的对象，而不必用于描述特定的顺序或先后次序。应该理解这样使用的数据在适当情况下可以互换，以便这里描述的本申请的实施例，例如能够以除了在这里图示或描述的那些以外的顺序实施。此外，术语“包括”和“具有”以及他们的任何变形，意图在于覆盖不排他的包含，例如，包含了一系列步骤或单元的过程、方法、系统、产品或设备不必限于清楚地列出的那些步骤或单元，而是可包括没有清楚地列出的或对于这些过程、方法、产品或设备固有的其它步骤或单元。
本申请提供一种门锁及识别方法，旨在解决现有门锁在对不同年龄层的 用户进行识别时所拍摄的图片不能满足识别要求的技术问题，以及当用户距离较远时，现有的结构光模块无法在人脸上投射55000个散斑，进而导致所拍摄图片不能满足识别要求的技术问题。
图3为本发明根据一示例性实施例示出的门锁的结构示意图。如图3所示，本实施例提供的门锁100，包括：三维识别结构101、驱动结构102以及处理器103。
上述三维识别结构101用于拍摄人脸的第一图像。处理器103用于判断第一图像是否符合预设识别条件，若第一图像不符合预设识别条件时，则生成驱动信号。驱动结构在驱动信号驱动下带动三维识别结构运动，以改变三维识别结构的视场角度，以使第一图像符合预设识别条件。
在本实施例提供门锁中，处理器判断三维识别结构拍摄的第一图像是否满足预设识别条件，当第一图像不满足预设识别条件时，则改变三维识别结构的视场角度，使得第一图像满足预设识别条件。相较于现有的门锁，本实施例提供的门锁能够调整三维识别结构的视场角度，使得三维识别结构能够拍摄出满足预设识别条件的图片。
本发明另一示例性实施例示出的门锁包括：三维识别结构、驱动结构以及处理器。
上述门锁安装于门上，人脸站立于门前，三维识别结构通过拍摄人脸的第一图像，处理器根据第一图像提取人脸与三维识别结构进行识别处理的视场关系。若人脸在三维识别结构进行识别处理的视场之内，不生成驱动信号。若人脸在三维识别结构进行识别处理的视场之外，则生成驱动信号。驱动结构在驱动信号的驱动下带动三维识别结构运动，以改变三维识别结构的视场角度，以使人脸处于三维识别结构进行识别处理的视场内。
下面详细介绍本发明提供的门锁的工作原理：
图4为基于图3所示实施例示出的门锁的一视场示意图。如图4所示，三维识别结构与水平面的角度为α 1，三维识别结构的视场角度为θ，门锁的安装高度为1.2，人脸的身高由1m到2m之间波动，人脸站立位置离门锁的距离在0.3m到1m之间波动。
三维识别结构与水平面的角度为α 1时，假设人脸的长度为0.2m，三 维识别结构仅能身高超过1.4m的人脸进行识别，当人脸站立位置离门锁仅有0.3m时，可能无法对身高1.4m的人脸进行识别。当三维识别结构检测到无法对人脸进行识别时，调整三维识别结构的视场角度。
图5为基于图3所示实施例示出的门锁的另一视场示意图。如图5所示，当处于角度α 1的三维识别结构检测到无法对人脸进行识别时，门锁中驱动结构驱动三维识别结构，使三维识别结构与水平面的角度为α 2。三维识别结构的视场角度仍为θ，但由于三维识别结构与水平面的角度变化，三维识别结构能够实现对身高1m的人脸进行识别，扩大了门锁的视场角度。
本实施例提供的门锁，三维识别结构用于对人脸进行识别处理，当人脸处于三维识别结构的视场之外时，驱动结构带动三维识别结构运动，以改变三维识别结构的视场角度，以使人脸处于三维识别结构进行识别处理的视场内，进而扩大门锁的视场范围，进而可以对处于不同年龄层的人脸进行识别处理。
本发明提供的又一示例性实施例示出的门锁的结构与图3所示实施例中门锁的结构相同。区别仅在于：三维识别结构拍摄第一张人脸的第一图像，处理器提取上述第一图像中人脸上的散斑数量。若散斑数量大于等于预设数量，不生成驱动信号。若散斑数量小于预设数量，生成驱动信号。驱动结构在驱动信号的驱动下带动三维识别结构运动，以改变三维识别结构的视场角度，三维识别结构在新的视场角度下再次拍摄第一图像，处理器提取第一图像中人脸上的散斑数量，并统计多张第一图像中人脸上的散斑数量，以获得人脸上总散斑数量，当总散斑数量大于等于预设数量时，不生成驱动信号。若总散斑数量小于预设数量，生成驱动信号。驱动结构在驱动信号的驱动下带动三维识别结构运动，以改变三维识别结构的视场角度，直至总散斑数量大于等于预设数量。
在本实施例提供的门锁中，通过驱动结构改变三维识别结构的视场角度，拍摄多张第一图片，再统计多张第一图片中人脸的散斑数量，通过多张第一图片的散斑数量叠加方式，使得人脸上散斑数量满足预设识别条件，最终利用多张第一图片进行人脸识别。
图6为本发明根据另一示例性实施例示出的门锁的结构示意图。如图6所示，本实施例提供的门锁200包括：三维识别结构201、驱动结构202、处理器203、存储器204以及第一外壳。其中，驱动结构202包括驱动器205和执行机构206。三维识别结构201包括第二外壳、红外传感器、RGB传感器、结构光投射器以及补光灯。
在上述门锁中，处理器203分别与三维识别结构201、驱动器205以及存储器204连接，驱动器205与执行机构206连接，执行机构206与三维识别结构201连接。三维识别结构201与第一外壳铰接。红外传感器、RGB传感器、结构光投射器以及补光灯均位于第二外壳内部。
在上述门锁中，预设识别条件为人脸处于三维识别结构的视场之内。三维识别结构201中红外传感器或者RGB传感器拍摄人脸的第一图像，并将第一图像发送至处理器203，处理器203在接收到第一图像后，提取第一图像中人脸与第一图像的位置关系。其中，位置关系可以为：人脸处于第一图像的中间位置，人脸处于第一图像的边缘位置。根据位置关系确定人脸与三维识别结构进行识别处理视场的关系。其中，人脸与三维识别结构进行识别处理视场的关系可以分为：人脸处于三维识别结构的视场之外或人脸处于三维识别结构的视场内。
判断人脸是否处于三维识别结构的视场之内可以为：判断第一图像中人脸区域的大小，或者判断人脸关键部分是否有在第一图像中，或者同时判断人脸区域的大小和关键部分是否在第一图像中。
例如：第一种判断方式为：在第一图像中存在2/3以上的人脸区域，则可以直接利用该第一图像进行人脸身份识别。第二种判断方式为：眉毛、眼睛、鼻子以及嘴巴这些人脸特征在第二图像中，则可以直接利用该第二图像进行人脸身份识别。第三种判断方式为：在第二图像中存在2/3以上的人脸区域，且眉毛、眼睛、鼻子以及嘴巴这些人脸特征在第二图像中，则可以直接利用该第二图像进行人脸身份识别。
当人脸处于第二图像的边缘位置，且上述判断结果为否，表明人脸处于三维识别结构的视角外，三维识别结构所处视场角度位置不合适，处理器根据第一图像获取三维识别结构的当前视场角度，并根据当前视场角度 和位置关系确定三维识别结构的目标视场角度；以根据目标视场角度生成驱动信号，驱动信号用于控制驱动结构带动三维识别结构相对第一外壳转动，以使三维识别结构由当前视场角度转动至目标视场角度，进而使得人脸进入三维识别结构的视场内。
作为一种具体实施方式，当人脸的身高在1m到2m之间波动，人脸站立的位置离门锁的距离在0.3m到1m之间波动，门锁的高度在1.2m时，门锁的视场范围需要达到100°，而门锁的视场范围一般在60°到90°，因此，只需要改变1次三维识别结构与水平面角度，即可使门锁的视场角度扩大到100°。
下面描述只改变1次三维识别结构与水平面的角度的方案：在该方案中三维识别结构仅在两个角度变化，即第一角度和第二角度。且三维识别结构的第一视场角度和三维识别结构的第二视场角度之和大于100°；其中，第一视场角度为三维识别结构位于第一角度时的视场角度，第二视场角度为三维识别结构位于第二角度时的视场角度。
图7为基于图6所示实施例示出的门锁的工作原理图。如图7所示，三维识别结构拍摄第一图像，处理器提取第一图像中人脸与第一图像的位置关系，并根据位置关系判断三维识别结构的当前视场角度。当人脸位于第一图像的下方时，可以判断三维识别结构的当前视场角度为第一角度α 1。当人脸位于第一图像的上方时，可以判断三维识别结构的当前视场角度为第二角度α 2。
处理器提取第一图像中人脸与第一图像的位置关系。并根据位置关系确定人脸与三维识别结构进行识别处理视场的关系。若人脸处于三维识别结构的视场之外时，处理器获取三维识别结构的当前视场角度，并根据当前视场角度和位置关系确定三维识别结构的目标视场角度。若三维识别结构的当前视场角度为第一角度α 1，且人脸位于第一图像的下方，则目标视场角度为第二角度α 2，执行机构带动三维识别结构转动至第二角度α 2。若三维识别结构为的当前视场角度为第二角度α 2，人脸位于第一图像的上方，则目标视场角度为第一角度α 1，执行机构带动三维识别结构转动至第一角度α 1。
在本实施例提供的门锁中，三维识别结构先拍摄第一图像，处理器根据 第一图像获得人脸与三维识别结构进行识别处理视场的关系。当人脸处于三维识别结构的视场之外时，则由处理器控制驱动器，驱动器生成驱动信号驱动执行机构运动，以带动三维识别结构相对第一外壳转动，改变三维识别结构的视场角度，以使人脸处于三维识别结构进行识别处理的视场内，进而扩大门锁的视场范围，进而可以对处于不同年龄层的人脸进行识别处理。
本发明提供的又一示例性实施例示出的门锁的结构与图6所示实施例中门锁的结构相同。区别仅在于：三维识别结构中红外传感器或者RGB传感器拍摄人脸的第一图像，并将第一图像发送至处理器，处理器在接收到第一图像后，并提取上述第一图像中人脸上散斑数量。若散斑数量大于等于预设数量，不生成驱动信号。若散斑数量小于预设数量，生成驱动信号。驱动结构在驱动信号的驱动下带动三维识别结构运动，以改变三维识别结构的视场角度，以使人脸上总散斑数量大于等于预设数量。
本实施例提供的门锁，由处理器控制驱动结构，驱动结构生成驱动信号驱动执行机构运动，以带动三维识别结构相对第一外壳转动，改变三维识别结构的视场角度，以使人脸上总散斑数量大于等于预设数量，最终利用多张第一图片进行人脸识别。
图8为本发明根据又一示例性实施例示出的门锁的结构示意图。如图8所示，本实施例提供的门锁300包括：三维识别结构301、驱动结构302、处理器303、存储器304以及第一外壳。其中，驱动结构302包括驱动器305和电磁铁306。三维识别结构301包括第二外壳、红外传感器、RGB传感器、结构光投射器以及补光灯。三维识别结构与第一外壳铰接，三维识别结构301上设有吸附块。吸附块可位于第二外壳内部或者第二外壳外部。吸附块的材料为磁铁或者可以被磁铁吸附的材质。
在上述门锁中，处理器303分别与三维识别结构301、驱动器305以及存储器304连接，驱动器305与电磁铁306连接，电磁铁306通过与吸附块之间产生电磁力，带动三维识别结构301运动。
当上述门锁工作时，三维识别结构拍摄第一图片，处理器根据第一图像获得人脸与三维识别结构进行识别处理视场的关系。若人脸处于三维识别结 构的视场之外时，处理器控制驱动器，驱动器使电磁铁通电，以使电磁铁与吸附块之间产生电磁力，以带动三维识别结构相对第一外壳转动，改变三维识别结构的视场角度，使人脸处于三维识别结构的视场范围内。
在本实施例提供的门锁中，执行机构为电磁铁，利用电磁铁通电时产生的电磁力带动三维识别结构运动，改变三维识别结构的视场角度，进而扩大三维识别结构的视场角度，以实现对不同年龄层人脸的识别处理，且电磁铁反应块，可以迅速切换角度，节省时间。
本发明提供的又一示例性实施例示出的门锁的结构与图8所示实施例中门锁的结构相同。区别仅在于：三维识别结构中红外传感器或者RGB传感器拍摄人脸的第一图像，并将第一图像发送至处理器，处理器在接收到第一图像后，并提取上述第一图像中人脸上散斑数量。若散斑数量大于等于预设数量，不生成驱动信号。若散斑数量小于预设数量，生成驱动信号。驱动器在驱动信号控制下使电磁铁通电，以使电磁铁与吸附块之间产生电磁力，以带动三维识别结构相对第一外壳转动，改变三维识别结构的视场角度，以使人脸上总散斑数量大于等于预设数量。
在本实施例提供的门锁中，执行机构为电磁铁，利用电磁铁通电时产生的电磁力带动三维识别结构运动，改变三维识别结构的视场角度，提高人脸上散斑数量，且电磁铁反应块，可以迅速切换视场角度，节省时间。
图9为本发明根据再一示例性实施例示出的门锁的结构示意图。如图9所示，本实施例提供的门锁400包括：三维识别结构401、驱动结构402、处理器403、存储器404以及第一外壳。其中，驱动结构402包括驱动器405、上电磁铁406和下电磁铁407。三维识别结构与第一外壳铰接，三维识别结构401上设有吸附块，且吸附块位于上电磁铁406和下电磁铁407之间。
在上述门锁中，处理器403分别与三维识别结构401、驱动器405以及存储器404电连接，驱动器405分别与上电磁铁406和下电磁铁407电连接，上电磁铁406和下电磁铁307均通过电磁力带动三维识别结构301运动。
图10为基于图9所示实施例示出的门锁的工作原理图。如图10所示，三维识别结构拍摄第一图片，处理器根据第一图像获得人脸与三维识别结构 进行识别处理视场的关系。若人脸处于三维识别结构的视场之外时，处理器获取三维识别结构的当前视场角度。并根据当前视场角度和位置关系确定三维识别结构的目标视场角度，并根据目标视场角度生成驱动信号。
若三维识别结构的当前视场角度为第一角度α 1。驱动器使下电磁铁通电，以使下电磁铁产生电磁力，以带动三维识别结构从第一角度α 1运动至第二角度α 2。进而改变三维识别结构的视场角度，使人脸处于三维识别结构的视场范围内。
若三维识别结构的当前视场角度为第二角度α 2。处理器控制驱动器，驱动器使上电磁铁通电，以使上电磁铁产生电磁力，以带动三维识别结构从第二角度α 2运动至第一角度α 1。进而改变三维识别结构的视场角度，使人脸处于三维识别结构的视场范围内。
在本实施例提供的门锁中，由上电磁铁和下电磁铁带动门锁在两个角度中切换，通过改变一次门锁的角度，改变三维识别结构的视场角度，即可使人脸处于三维识别结构的视场范围内，降低处理器的控制难度，使得该门锁结构简单。
图11为本发明根据再又一示例性实施例示出的门锁的结构示意图。如图11所示，本实施例提供的门锁500包括：三维识别结构501、驱动结构502、处理器503以及存储器504。其中，驱动结构502包括电机505和齿轮组506。
在上述门锁中，处理器503分别与三维识别结构501、电机502以及存储器504连接，电机505与齿轮组506的主动轮机械连接，齿轮组506的从动轮带动三维识别结构501运动。
图12为基于图11所示实施例示出的门锁的工作原理图。如图12所示，当上述门锁工作时，三维识别结构拍摄第一图片，处理器根据第一图像获得人脸与三维识别结构进行识别处理视场的关系。若人脸处于三维识别结构的视场之外时，处理器控制电机，电机转动驱动主动轮转动，主动轮带动从动轮转动，从动轮带动三维识别结构运动，改变三维识别结构的视场角度，使人脸处于三维识别结构的视场范围内。
在上述门锁中，可以通过控制电机的转动时间和转速来控制主动轮的转 动角度，进而控制三维识别结构的转动角度，当处理器检测到并未人脸在门锁的视场内时，处理器获得三维识别结构的当前视场角度，并根据当前视场角度和位置关系确定目标视场角度，再根据目标视场角度生成驱动信号，以控制电机的电机的转动时间和转速。
在本实施例提供的门锁中，执行机构为齿轮组，通过驱动齿轮组可以实现精确地将三维识别模块的视场角度由当前视场角度转动至目标视场角度，使人脸处于三维识别模块的视场范围内，以适应不同年龄层的人脸识别。
本发明提供的又一示例性实施例示出的门锁的结构与图9所示实施例中门锁的结构相同。区别仅在于：处理器在接收到第一图像后，并提取上述第一图像中人脸上散斑数量。若散斑数量大于等于预设数量，不生成驱动信号。若散斑数量小于预设数量，生成驱动信号。驱动信号控制使电机转动，进而通过电机驱动主动轮转动，主动轮带动从动轮转动，从动轮带动三维识别结构运动，改变三维识别结构的视场角度，以使人脸上总散斑数量大于等于预设数量，最终利用多张第一图片进行人脸识别。
图13为本发明根据一示例性实施例示出的识别方法的示意图。如图13所示，本实施例提供的识别方法应用于门锁，门锁包括：三维识别结构。该识别方法包括如下步骤：
S601、获取三维识别结构拍摄的人脸的第一图像。
S602、判断第一图像是否满足预设识别条件；若判断结果为是，进入S603，否则，进入S604。
更具体地，当预设识别条件为人脸处于三维识别结构进行识别处理的视场之内时。采用如下方式获得获取人脸与三维识别结构进行识别处理的视场的关系：
获取三维识别结构拍摄的人脸的第一图像
获取三维识别结构拍摄的人脸的第一图像，提取在第一图像中人脸的区域大小，若有2/3的人脸在第一图像中，则可以判断人脸在门锁的视场内。也可以为进一步提取人脸的五官特征，判断五官特征是否均在第一图像中，若五官特征均在第一图像中，则可以判断特征在门锁的视场内。也可以为提 取人脸的区域大小以及提取人脸的五官特征，并判断五官特征在第一图像中，并判断有2/3的人脸在第一图像中，则可以判断特征在门锁的视场内。
当预设识别条件为人脸上总散斑数量是否满足预设数量时。提取每张第一图像中人脸上散斑数量，并统计多张第一图像中人脸上散斑数量，以获得人脸上总散斑数量。
S603、使驱动结构运动，以带动三维识别结构运动，以使第一图像符合预设识别条件。
更具体地，使三维识别结构运动，以改变三维识别结构的视场角度，以使人脸处于三维识别结构的视场内。
S604、利用第一图像进行识别处理。
在本实施例提供的方法中，在本实施例提供门锁中，处理器判断三维识别结构拍摄的第一图像是否满足预设识别条件，当第一图像不满足预设识别条件时，则改变三维识别结构的视场角度，使得第一图像满足预设识别条件。相较于现有的门锁，本实施例提供的门锁能够调整三维识别结构的视场角度，使得三维识别结构能够拍摄出满足预设识别条件的图片。
图14为本发明根据另一示例性实施例示出的识别方法的示意图。当人脸的身高在1m到2m之间波动，人脸站立的位置离门锁的距离在0.2m到1m之间波动，门锁的高度在1.2m时，门锁的视场范围需要达到100°，而门锁的视场范围在60°到90°，因此，只需要改变1次门锁与水平面角度，即可使门锁的视场角度扩大到100°。
如图14所示，本实施例提供的识别方法包括如下步骤：
S701、获取三维识别结构拍摄的人脸的第一图像。
S702、判断第一图像是否满足预设识别条件；若判断结果为是，进入S707，否则，进入S703。
S703、获取三维识别结构的当前视场角度。
更具体地，在本实施例中，根据第二位置关系确定三维识别结构的当前视场角度。当人脸位于第一图像的上边缘时，当前视场角度为第二角度。当人脸位于第一图像的下边缘时，当前视场角度为第一角度。
三维识别结构的所处的角度只有两个：第一角度或第二角度。第一角度 和第二角度。第一视场角度为三维识别结构位于第一角度时的视场角度，第二视场角度为三维识别结构位于第二角度时的视场角度。三维识别结构的第一视场角度和三维识别结构的第二视场角度之和大于100°。
S704、提取第一图像中人脸与第一图像的位置关系。
更具体地，位置关系包括人脸在第一图像中位置和人脸在第一图像中区域大小。人脸在第一图像中位置可以分为中间位置和边缘位置。人脸在第一图像中区域大小可以为：人脸的关键区域在第一图像中区域大小。或者人脸整体在第一图像中区域大小。例如：眼睛、鼻子、嘴巴这些关键区域。
S705、根据当前视场角度和位置关系确定三维识别结构的目标视场角度。
更具体地，若当前视场角度为第一角度，人脸位于第一图像的下边缘的，目标视场角度为第二角度。若当前视场角度为第二角度，人脸位于第一图像上边缘，则目标视场角度为第一角度。
S706、带动三维识别结构运动，以使三维识别结构由当前视场角度运动至目标视场角度，以使第一图像符合预设识别条件。
S707、利用第一图像进行人脸识别。
在本实施例提供的识别方法中，三维是被结构在两个角度中切换，通过改变一次三维识别结构的视场角度，即可使人脸处于三维识别结构的视场范围内，降低识别方法的难度，使得该门锁结构简单。
本领域普通技术人员可以理解，文中的门锁除了普通家庭用户安装的门锁之外，也可以指公众场所安装的各类门禁门锁系统。
本领域普通技术人员可以理解：实现上述各方法实施例的全部或部分步骤可以通过程序指令相关的硬件来完成。前述的程序可以存储于一计算机可读取存储介质中。该程序在执行时，执行包括上述各方法实施例的步骤；而前述的存储介质包括：ROM、RAM、磁碟或者光盘等各种可以存储程序代码的介质。
最后应说明的是：以上各实施例仅用以说明本发明的技术方案，而非对其限制；尽管参照前述各实施例对本发明进行了详细的说明，本领域的普通技术人员应当理解：其依然可以对前述各实施例所记载的技术方案进行修改，或者对其中部分或者全部技术特征进行等同替换；而这些修改或者替换，并不使相应技术方案的本质脱离本发明各实施例技术方案的范围。

Claims (15)

  1. 一种门锁，其特征在于，包括：
    三维识别结构，用于拍摄人脸的第一图像；
    处理器，用于当所述第一图像不符合预设识别条件时生成驱动信号；
    驱动结构，用于在所述驱动信号驱动下带动所述三维识别结构运动，以改变所述三维识别结构的视场角度，以使所述第一图像符合所述预设识别条件。
  2. 根据权利要求1所述的门锁，其特征在于，所述人脸的第一图像不符合预设识别条件具体包括：
    所述人脸与所述三维识别结构进行识别处理的视场的关系为所述人脸处于所述视场之外。
  3. 根据权利要求1所述的门锁，其特征在于，所述人脸处于所述视场之外具体包括：
    处于所述视场内的人脸部分与所述人脸的比值小于预设数值；和/或
    所述人脸的关键部分处于所述视场内。
  4. 根据权利要求1所述的门锁，其特征在于，所述人脸的第一图像不符合预设识别条件具体包括：
    所述人脸上的总散斑数量少于预设数量。
  5. 根据权利要求4所述的门锁，其特征在于，统计多张所述第一图像中所述人脸上的散斑数量，以获得所述总散斑数量。
  6. 根据权利要求1至5任一项所述的门锁，其特征在于，所述门锁还包括：第一外壳；所述三维识别结构与所述第一外壳铰接。
  7. 根据权利要求6所述的门锁，其特征在于，所述驱动结构包括驱动器和电磁铁，所述三维识别结构上设有吸附块；
    所述驱动器受控于所述驱动信号使所述电磁铁通电，以使所述电磁铁与所述吸附块之间产生电磁力，以带动所述三维识别结构运动。
  8. 根据权利要求7所述的门锁，其特征在于，所述电磁铁包括：上电磁铁和下电磁铁；
    所述吸附块位于所述上电磁铁和所述下电磁铁之间。
  9. 根据权利要求6所述的门锁，其特征在于，所述驱动结构包括齿轮组 和电机；所述三维识别结构通过铰接轴与所述第一外壳铰接；
    所述齿轮组中处于末端的齿轮安装于所述铰接轴上，所述电机驱动所述齿轮组运动，以使所述齿轮组带动所述三维识别结构运动。
  10. 一种识别方法，其特征在于，应用于门锁，所述门锁包括：三维识别结构和驱动结构；所述方法包括：
    获取所述三维识别结构拍摄的人脸的第一图像；
    判断所述第一图像是否满足预设识别条件；
    若判断结果为否，使所述驱动结构运动，以带动所述三维识别结构运动，以使所述第一图像符合所述预设识别条件；
    利用所述第一图像进行识别处理。
  11. 根据权利要求10所述的方法，其特征在于，所述判断所述第一图像是否满足预设识别条件，具体包括：
    判断所述人脸与所述三维识别结构进行识别处理的视场的关系是否为所述人脸处于所述三维识别结构进行识别处理的视场之内。
  12. 根据权利要求11所述的方法，其特征在于，所述人脸处于所述视场之外具体包括：
    处于所述视场内的人脸部分与所述人脸的比值小于预设数值；和/或
    所述人脸的关键部分处于所述视场内。
  13. 根据权利要求11或12所述的方法，其特征在于，所述使所述驱动结构运动，以带动所述三维识别结构运动，具体包括：
    根据所述第一图像获取所述三维识别结构所处的当前视场角度；
    提取所述第一图像中人脸与第一图像的位置关系；
    根据所述当前视场角度和所述位置关系确定所述三维识别结构的目标视场角度；
    所述驱动结构带动所述三维识别结构，以使所述三维识别结构由所述当前视场角度运动至所述目标视场角度。
  14. 根据权利要求13所述的方法，其特征在于，所述当前视场角度为第一角度或者第二角度；所述目标视场角度为所述第一角度或者所述第二角度；且所述目标视场角度和所述当前视场角度不相同。
  15. 根据权利要求14所述的方法，其特征在于，第一视场角度和第二视 场角度之和大于100°；
    其中，所述第一视场角度为所述三维识别结构位于所述第一角度时的视场角度，所述第二视场角度为所述三维识别结构位于所述第二角度时的视场角度。
PCT/CN2019/093876 2019-06-28 2019-06-28 门锁及识别方法 WO2020258304A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
CN201980001121.9A CN110462693B (zh) 2019-06-28 2019-06-28 门锁及识别方法
PCT/CN2019/093876 WO2020258304A1 (zh) 2019-06-28 2019-06-28 门锁及识别方法

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2019/093876 WO2020258304A1 (zh) 2019-06-28 2019-06-28 门锁及识别方法

Publications (1)

Publication Number Publication Date
WO2020258304A1 true WO2020258304A1 (zh) 2020-12-30

Family

ID=68492784

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/093876 WO2020258304A1 (zh) 2019-06-28 2019-06-28 门锁及识别方法

Country Status (2)

Country Link
CN (1) CN110462693B (zh)
WO (1) WO2020258304A1 (zh)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114046718B (zh) * 2022-01-14 2022-04-22 深圳市创腾智能电子有限公司 一种基于图像的摆闸摆动角度检测设备及检测方法

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2017122376A (ja) * 2016-01-08 2017-07-13 日本電産リード株式会社 扉用虹彩認証装置、及び虹彩認証システム
CN107146305A (zh) * 2017-05-02 2017-09-08 上海灵至科技有限公司 基于人脸识别的解锁方法及门锁
CN107590454A (zh) * 2017-09-04 2018-01-16 江苏非寻信息技术有限公司 一种具有虹膜识别技术的安检检测装置
CN109184358A (zh) * 2018-10-09 2019-01-11 芜湖市越泽机器人科技有限公司 具有人脸识别的智能门锁系统

Family Cites Families (14)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH08329213A (ja) * 1995-05-31 1996-12-13 Yazaki Corp 通過物体検知装置及び通過物体計数装置
CN104077804B (zh) * 2014-06-09 2017-03-01 广州嘉崎智能科技有限公司 一种基于多帧视频图像构建三维人脸模型的方法
WO2018133027A1 (zh) * 2017-01-20 2018-07-26 深圳大学 基于灰度约束的三维数字散斑的整像素搜索方法及装置
CN107093230A (zh) * 2017-03-28 2017-08-25 上海斐讯数据通信技术有限公司 一种智能摄像头的云管理方法及其系统
CN109501722A (zh) * 2017-09-15 2019-03-22 南京志超汽车零部件有限公司 一种车辆防盗用3d脸部识别系统的光学组件
CN207586943U (zh) * 2017-11-14 2018-07-06 南京甄视智能科技有限公司 用于银行的具有自动调节功能的3d人脸识别装置
CN109903328B (zh) * 2017-12-11 2021-12-21 宁波盈芯信息科技有限公司 一种应用于智能手机的物体体积测量的装置及方法
CN108052506B (zh) * 2017-12-28 2021-06-29 Oppo广东移动通信有限公司 自然语言处理方法、装置、存储介质及电子设备
CN111126146B (zh) * 2018-04-12 2024-03-05 Oppo广东移动通信有限公司 图像处理方法、装置、计算机可读存储介质和电子设备
CN208488784U (zh) * 2018-07-06 2019-02-12 北京驭光科技发展有限公司 进门控制系统
CN109086694B (zh) * 2018-07-17 2024-01-19 北京量子光影科技有限公司 一种人脸识别系统及方法
CN109087382A (zh) * 2018-08-01 2018-12-25 宁波发睿泰科智能科技有限公司 一种三维重构方法及三维成像系统
CN109284597A (zh) * 2018-11-22 2019-01-29 北京旷视科技有限公司 一种人脸解锁方法、装置、电子设备和计算机可读介质
CN109359648B (zh) * 2018-12-08 2021-06-25 广东伟邦科技股份有限公司 一种用于对不同身高的人进行人脸识别的系统

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2017122376A (ja) * 2016-01-08 2017-07-13 日本電産リード株式会社 扉用虹彩認証装置、及び虹彩認証システム
CN107146305A (zh) * 2017-05-02 2017-09-08 上海灵至科技有限公司 基于人脸识别的解锁方法及门锁
CN107590454A (zh) * 2017-09-04 2018-01-16 江苏非寻信息技术有限公司 一种具有虹膜识别技术的安检检测装置
CN109184358A (zh) * 2018-10-09 2019-01-11 芜湖市越泽机器人科技有限公司 具有人脸识别的智能门锁系统

Also Published As

Publication number Publication date
CN110462693A (zh) 2019-11-15
CN110462693B (zh) 2022-04-22

Similar Documents

Publication Publication Date Title
US7711155B1 (en) Method and system for enhancing three dimensional face modeling using demographic classification
US20130286161A1 (en) Three-dimensional face recognition for mobile devices
WO2018223469A1 (zh) 动向投影装置及其工作方法
US20020149613A1 (en) Automatic positioning of display depending upon the viewer's location
JP6222948B2 (ja) 特徴点抽出装置
JP5730095B2 (ja) 顔画像認証装置
CN103543828A (zh) 一种显示系统及自动调整显示屏位置的方法
WO2013074153A1 (en) Generating three dimensional models from range sensor data
KR101436290B1 (ko) 생체인식 유형의 접근제어시스템에 대한 부정 검출
KR102441974B1 (ko) Tof 카메라를 이용한 주차 관리장치 및 관리방법
WO2020258304A1 (zh) 门锁及识别方法
WO2012117901A1 (ja) データ処理装置、データ処理システム、及びプログラム
TW202040426A (zh) 資訊處理裝置、資訊處理系統、資訊處理方法以及非暫時性儲存媒體
US11244182B2 (en) Spoof detection by estimating subject motion from captured image frames
TWI445511B (zh) 化妝鏡調整系統、方法及具有該調整系統的化妝鏡
Hanna et al. A System for Non-Intrusive Human Iris Acquisition and Identification.
Zoidi et al. Appearance based object tracking in stereo sequences
JP2005135339A (ja) 通行人検出方法および装置ならびに通行人計数装置
Venugopalan et al. Unconstrained iris acquisition and recognition using cots ptz camera
TW201205449A (en) Video camera and a controlling method thereof
Murali et al. Autonomous exploration using rapid perception of low-resolution image information
JP2007172509A (ja) 顔検出照合装置
CN109993029A (zh) 注视点模型初始化方法
JP2019133275A (ja) 顔認証装置及び顔認証方法
Hussain et al. Accurate Face Recognition system based on ARM 9 processor using HAAR wavelets

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: 19934867

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: 19934867

Country of ref document: EP

Kind code of ref document: A1