JP6431346B2 - Face recognition device - Google Patents

Description

  Embodiments described herein relate generally to an obstacle-friendly face recognition apparatus.

  Face recognition applications in the security field, functions for detecting smiles with a digital camera, and vending machines that determine whether or not to purchase based on the age estimated from the face are being actively developed.

  In face recognition, if there is a physical obstacle in front of the camera, recognition using the camera becomes impossible. In such a case, it is necessary to present the user to avoid the shielding from the system side and to request the user to remove the shielding.

  However, such a request to the user is accompanied by the user's action, and has become a factor that makes the user aware of bothersomeness and complexity. And it cannot be said that the application which presents the information adapted for every user using the information which recognized the face is provided sufficiently.

JP 2008-283379 A

  The problem to be solved by the present invention is to provide a face recognition device that can perform a recognition operation smoothly without making the user aware of recognition by a camera in a natural flow as much as possible.

The face recognition device according to the embodiment is provided outside, an image acquisition unit that acquires a human image, and a face part detection unit that inputs image information from the image acquisition unit and detects each part constituting the face A person detection sensor that is arranged in the vicinity of the image acquisition unit and detects the presence of a person, a person position determination unit that takes in the sensor output of the person detection sensor and determines the position of the person, and the face part detection The detection signal is received from the unit, the determination signal is acquired from the person position determination unit, and the determination result of the shield determination unit that determines the presence or absence of the shield and the determination result of the shield determination unit is received, and it is determined that there is no shield A face recognition unit that generates a face recognition image for face authentication, and a display that displays display contents that guide the user to remove the shielding object when the shielding object determination unit determines that the shielding object is present and a section, wherein Object position determination unit comprises a person detection section that detects the presence or absence of a person, if the person is detected, to track a person can be estimated as the same person, consists person moving vector calculation unit for calculating a movement vector of the person The

It is a figure which shows the structural example of the face recognition apparatus which concerns on embodiment of this invention. It is a flowchart which shows the flow of a person position determination process. It is a flowchart which shows the flow of a face recognition process.

  Hereinafter, an embodiment of the present invention will be described with reference to the drawings. In the drawings, the same portions are denoted by the same reference numerals, and redundant description is omitted.

  FIG. 1 is a diagram illustrating a configuration example of a face recognition apparatus according to an embodiment of the present invention.

  As shown in FIG. 1, the face recognition device 100 mainly includes an image acquisition unit 10, a face part detection unit 20, a person detection sensor 30, a person position determination unit 40, an obstruction determination unit 50, a face recognition unit 60, a display. The unit 70 is configured. In this apparatus, the face part detection unit 20, the person position determination unit 40, the shielding object determination unit 50, and the face recognition unit 60 can be realized as a general-purpose CPU and software that operates on the same CPU. The present embodiment can also be implemented as a program that causes such a CPU to execute a series of procedures relating to face part detection, person position determination processing, and face recognition processing.

  The image acquisition unit 10 is provided outside and acquires image information, and is a face recognition camera (face camera) arranged so that a human face is reflected. For example, a wide-angle camera is preferable as the face camera, and a fish-eye camera that is a super-wide-angle camera is preferable.

  The face part detection unit 20 receives image information from the image acquisition unit 10 and detects each part constituting the face. For example, the image information is scanned in a predetermined direction, and feature portions (face parts) corresponding to eyes, nose, mouth and the like are specified.

  The person detection sensor 30 is disposed in the vicinity of the image acquisition unit 10 and detects the presence of a person. As the person detection sensor 30, for example, an image sensor, an ultrasonic sensor, or a pyroelectric infrared sensor is suitable. The ultrasonic sensor transmits ultrasonic waves to a target object (person) by a transmitter and receives the reflected waves by a receiver to detect the presence or absence of the target object (person) and the target object (person). The distance is detected. The pyroelectric infrared sensor detects light including infrared rays by the pyroelectric effect, and an output is obtained when the amount of light is biased due to the movement of the human body.

  The person position determination unit 40 takes in the sensor output of the person detection sensor 30 and determines the position of the person. The person position determination unit 40 can be configured by, for example, a person detection unit 41 and a person movement vector calculation unit 42. The person detection unit 41 acquires distance information output from the person detection sensor 30. When an image sensor is used as the person detection sensor 30, for example, a person is detected using background difference or CoHOG (Co-occurrence HOG). Human detection based on background difference is a technique for obtaining an object region based on a difference between a background image without a target and an input image. CoHOG is a technique for capturing the structure of a person's shape by expressing the relationship between gradients in two local regions.

  The person movement vector calculation unit 42, for example, tracks a person who can be estimated as the same person from the positions of the persons detected between frames, and calculates the movement vector. The person position determination unit 40 determines whether or not the person is close to the face recognition device 100 from the movement vector and the size of the person shown in the frame. It is possible to appropriately determine how close the distance is.

  The shielding object determination unit 50 receives the detection signal from the face part detection unit 20 and the determination signal from the person position determination unit 40, and determines the presence or absence of the shielding object. The determination of the presence / absence of the shielding object is, for example, when it is not possible to detect any part of the face and the brightness of the entire camera image is extremely low, it is determined that there is an shielding object in front of the camera. If there is a shield, the user is guided to remove the shield. A user's hand is most likely as a shield. Therefore, first, an OK button or the like is displayed on the display unit 70, and the display unit 70 is touched with the left and right hands. If that doesn't work, ask the user to check the object in front of the camera. If it still fails, it is determined that the camera is faulty. The determination result is sent to the face recognition unit 60 or the display unit 70.

  The face recognition unit 60 receives the determination result of the shielding object determination unit 50, and generates a face recognition image for face authentication when it is determined that there is no shielding object. The generated face recognition image signal is output to the display unit 70.

  The display unit 70 displays the face recognition image received from the face recognition unit 60. Further, in the present embodiment, when the shielding object determination unit 50 determines that there is an obstacle, the above-described “guidance to guide the obstacle” is displayed.

  The face recognition device 100 configured as described above will be described in detail.

  In the present embodiment, it is possible to detect whether there is an obstruction in front of the camera and perform authentication work smoothly without making the user aware of the recognition by the camera in a natural flow as much as possible. Thus, it is possible to acquire face recognition image information.

<Person position determination>
FIG. 2 is a flowchart showing the flow of the person position determination process.

  First, the person detection sensor is activated and initialized (step S201).

  Next, person detection is started (step S202). Specifically, the distance information output from the person detection sensor 30 is acquired.

  Next, the person is tracked (step S203). In tracking a person, for example, a person who can be estimated as the same person is tracked from the position of the person detected between frames, and the movement vector is calculated.

  Subsequently, the person position is determined (step S204). Specifically, the person position determination unit 40 determines whether the person is close to the face recognition device 100 from the movement vector and the size of the person shown in the frame.

  Next, it is determined whether or not the power of the person detection sensor is OFF (step S205).

  If the power is not off (No in step S205), the processing from step S202 to step S205 is repeated. If the power is off (Yes in step S205), the person position determination process is terminated.

<Face recognition processing>
FIG. 3 is a flowchart showing the flow of face recognition processing.

  First, the face recognition camera, which is the image acquisition unit 10, is activated and initialized (step S301). For initialization, for example, camera calibration or the like is performed.

  Next, information on the person position is acquired (step S302).

  Subsequently, it is determined whether or not a person is close (step S303).

  If it is not close (No in Step S303), the process returns to Step S302, and if it is close (Yes in Step S303), a facial part is detected (Step S304). The face parts to be detected may be all of eyes, nose, or mouth.

  Next, the presence / absence of a shielding object is determined (step S305). For example, when none of the facial parts can be detected and the brightness of the entire camera image is extremely low, it is determined that there is a shield in front of the face camera.

  If there is a shielding object (Yes in step S305), after guiding the shielding object to be removed (step S306), the process proceeds to step S304.

  On the other hand, if there is no obstruction (No in step S305), face authentication is started (step S307). In the face recognition device 100 according to the present embodiment, face recognition image information for performing face authentication work may be acquired, and the authentication work may be performed in a separately prepared authentication system.

  Next, it is determined whether or not the face recognition camera is turned off (step S308).

  If the power is not turned off (No in step S308), the processing from step S302 to step S307 is repeated. If the power is turned off (Yes in step S308), the face recognition process is terminated.

  According to this embodiment, it is possible to prevent the face recognition system from not operating due to a human error. Moreover, there is no possibility that the user who uses the face recognition system will feel uncomfortable.

  Although several embodiments of the present invention have been described, these embodiments are presented by way of example and are not intended to limit the scope of the invention. These novel embodiments can be implemented in various other forms, and various omissions, replacements, and changes can be made without departing from the scope of the invention. These embodiments and modifications thereof are included in the scope and gist of the invention, and are included in the invention described in the claims and the equivalents thereof.

DESCRIPTION OF SYMBOLS 100 ... Face recognition apparatus 10 ... Image acquisition part 20 ... Face parts detection part 30 ... Person detection sensor 40 ... Person position determination part 50 ... Shielding object determination part 60 ... Face Recognizing unit 70 ... display unit

Claims (8)

An image acquisition unit arranged outside and acquiring an image of a person;
A face part detection unit that inputs image information from the image acquisition unit and detects each part constituting the face;
A person detection sensor arranged in the vicinity of the image acquisition unit for detecting the presence of a person;
A person position determination unit that takes in the sensor output of the person detection sensor and determines the position of the person;
Detecting a detection signal from the face part detection unit, capturing a determination signal from the person position determination unit,
A face recognition unit that receives a determination result of the shielding object determination unit and generates a face recognition image for face authentication when it is determined that there is no shielding object;
When it is determined that there is a shielding object in the shielding object determination unit, the display unit displays a display content that guides to remove the shielding object ,
The person position determination unit
A person detection unit for detecting the presence or absence of a person;
If the person is detected, to track a person can be estimated as the same person, adult Ru facial recognition device from the person moving vector calculation unit for calculating a movement vector of the person.   The face recognition apparatus according to claim 1, wherein the display unit displays a face recognition image received from the face recognition unit when the shielding object determination unit determines that there is no shielding object. The face recognition apparatus according to claim 1, wherein the person detection sensor is an image sensor that detects the whole body of a person. The face recognition apparatus according to any one of claims 1 to 3 , wherein the image acquisition unit is a face recognition camera arranged so that a human face is captured. The face part detecting section, the image information from the image acquisition unit by scanning in a predetermined direction, eyes, nose, any one of claims 1 to 4 will identify the feature portion corresponding to the mouth The face recognition device described in 1. The person detecting unit, a face recognition apparatus according to any one of claims 1 to 5 perform a person detected by background subtraction. The face recognition device according to any one of claims 1 to 6 , wherein the person position determination unit determines whether or not the person is nearby based on a movement vector and the size of the person shown in the frame. The shielding object determination unit determines that there is a shielding object when the facial part detection unit cannot detect any part of the face and the brightness of the entire image acquired by the image acquisition unit is extremely low. The face recognition device according to any one of claims 1 to 7 .

Images