JP2006099486A - Entrance management system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To manage the staying time or in-room time of a person to be authenticated and a person not to be authenticated, centering on face authentication in one system. <P>SOLUTION: In the system for performing an entrance processing to the person to be subjected to the face authentication and the person not to be subjected to the face authentication, the authenticated face image at the time of entrance, entrance time and staying limit time of the person to be authenticated are stored, the face image at the time of the entrance, entrance time and staying limit time of the person not to be authenticated are stored, the face image of the person to be authenticated exceeding the staying limit time is displayed and the face image of the person not to be authenticated exceeding the staying limit time is displayed. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a room entry management system for managing the staying time of an authentication target person who performs authentication using a face image and an authentication non-target person who does not perform authentication using a face image.

  In the conventional visitor reception system, only the visitor's identity is confirmed and entry permission is granted, and the management is managed only by the reception ledger until the visitor comes out. It was a situation that could not respond to.

  Further, diversion of a card access control system (Patent Document 1) has been used.

  In the prior art, the following security holes exist.

1. When a card access control system is diverted, the card cannot be used to manage outsiders, or theft of the card can be a big problem (security hole).

2. In the card system, since the person is identified by the identification ID written in the card, the identity of the person in the room cannot be confirmed in real time.
Japanese Patent Laid-Open No. 4-237979

  It is an object of the present invention to provide an entrance management system capable of managing the stay time or occupancy time of a person to be authenticated and a non-authenticated person centered on face authentication.

  An entrance management system according to the present invention includes a registration unit that registers an image including at least a face of a person to be authenticated, an image capturing unit that captures an image including at least a face of a person entering from a desired entrance, and the image capturing unit captures the image. A determination means for determining whether the person is an authentication subject or a non-authentication person by collating an image including at least a face with an image including at least a face registered by the registration means; It has a management means which manages each stay time with respect to an authentication object person and an authentication non-object person.

  An entrance management system according to the present invention includes a registration unit that registers an image including at least a face of a person to be authenticated, an image capturing unit that captures an image including at least a face of a person entering from a desired entrance, and the image capturing unit captures the image. A determination unit that determines an authentication target person by collating an image including at least a face with an image including at least a face registered by the registration unit; and a setting unit that sets information indicating a non-authentication target person. And a managing means for managing respective stay times for the person to be authenticated determined by the determining means and the non-authenticated person set by the setting means.

  The room entry management system according to the present invention includes a registration unit that registers an image including at least a face of a person to be authenticated, an image capturing unit that captures an image including at least a face of a person entering the room through a desired entrance, and an image captured by the image capturing unit. A determination unit that determines an authentication target person by collating an image including at least a face with an image including at least a face registered by the registration unit; and a setting unit that sets information indicating a non-authentication target person. Corresponding to the information indicating the non-authenticated person set by the setting means, storage means for storing an image including at least the face imaged by the imaging means, and the authentication target person judged by the judging means Management means for managing the respective staying status for the non-authenticated person set by the setting means, and the management status by the management means Extract the chamber's, and has a display means for displaying a list of images including the face corresponding to the occupants.

  An entrance management system according to the present invention includes a registration unit that registers an image including at least a face of a person to be authenticated, an image capturing unit that captures an image including at least a face of a person entering from a desired entrance, and the image capturing unit captures the image. A determination unit that determines an authentication target person by collating an image including at least a face with an image including at least a face registered by the registration unit; and a setting unit that sets information indicating a non-authentication target person. Corresponding to the information indicating the non-authenticated person set by the setting means, storage means for storing an image including at least the face imaged by the imaging means, and the authentication target person judged by the judging means Management means for managing each entry time and staying time for the non-authenticated person set by the setting means, and a management status by the management means Extracting excess's time spent by those having display means for displaying a list of images including the face corresponding to the excess users of this residence time.

  As described above, according to the present invention, it is possible to provide an entrance management system that can perform stay time management of a person to be authenticated and a non-authenticated person with a single system centering on face authentication.

  Hereinafter, an embodiment of an entrance management system according to the present invention will be described with reference to the drawings.

  This entrance management system manages the entrance and exit of humans using facial images at the entrances and exits of buildings and rooms, etc., so that employees (internal) who are subject to authentication and occupants (outside) who are non-authenticated Time management is performed.

  That is, as shown in FIG. 1, the entrance / exit management system is provided on the entrance side of the entrance / exit 1, and includes a gate device (entrance device) 2 having a door opening / closing function and a face image matching function, and an exit side of the entrance / exit 1. A gate device (exit device) 3 provided with a door opening / closing function and a face image matching function, a management server 5 and a monitoring device 6 connected via a communication line 4 are provided.

  The gate device 2 (3) includes an operation display unit 10, a camera 11, a human detection unit 12, a door opening / closing unit 13, an image memory 14, a main memory 15, a face image matching unit 16, a control unit 17, and a transmission interface 18. ing.

  The operation display unit 10 performs various settings and guides and displays the processing status, and has a numeric keypad that is used when a vendor inputs a permission code.

  The camera 11 is an imaging unit that captures an image including at least a face of the approaching human when the human detection unit 12 detects that the human has approached.

  The human detection unit 12 detects that a human has approached the gate device 2.

  The door opening / closing part 13 opens and closes a door provided at the doorway 1. The image memory 14 stores a captured image from the camera 11. The main memory 15 stores a control program and is used for storing work data.

  The face image matching unit 16 includes a face image dictionary 16a in which a template as a feature vector (face pattern) as a face feature amount with respect to the face image of the employee as the authentication target is registered in advance along with the employee number, and the camera 11. And an authentication unit 16b that performs authentication by comparing and comparing a feature vector (face pattern) with respect to the face image obtained from the captured image from the registered content of the face image dictionary 16a. The control unit 17 controls the entire gate device 2.

  The face image dictionary 16a in the gate device 3 is registered with the data of the people who have entered the room by the gate device 2.

  The control unit 17 performs processing for generating a feature vector (face pattern: corresponding to a template) based on a face image from an image including at least a face captured by the camera 11.

  For example, a photographed image including at least a face is taken into the image memory 14, and an area of the image including at least a face is searched from the entire captured image. When the face image area is detected, a pupil area and a nostril area that can be regarded as a substantially circular area are detected in the detected face image area, and a center position of these detected areas is extracted as a feature point of the face image.

  When feature points are extracted in this way, collation areas are extracted in order based on the positions of the feature points. Thus, when the collation area is cut out, the size of each collation area is normalized by geometric correction. Further, the density distribution is normalized by density correction for each collation area. When these processes are completed, a feature vector (face pattern) as a feature amount of the face image is calculated (generated) based on the extracted feature points.

  The control unit 17 performs processing for generating a feature vector (face pattern) of a face image from an image captured by the camera 11 when the human detection unit 12 detects that a person approaches the gate device 2. is there. Further, the control unit 17 permits entry / exit from the entrance / exit 1 by the authentication processing by the face image matching unit 16 with respect to the feature vector of the generated face image, and the management server 5 stores the face image, information indicating the employee, and the entry time. To send to.

  In addition, when the authentication using the face image cannot be performed, the control unit 17 guides the input of the permission code by the contractor through the operation display unit 10, and the entrance / exit 1 is confirmed by confirming the permission code input by the numeric keypad according to the guidance. Is permitted to enter the room, and the face image, the permission code, and the room entry time are transmitted to the management server 5.

  The management server 5 includes an operation display unit (not shown), a control unit, and an imaging unit using a camera. The management server 5 captures an image including at least the face of a user permitted to enter and exit from the entrance 1 in advance. A face image feature vector (face pattern: corresponding to a template) as a face feature quantity for an image including at least a face is registered in the face image dictionary 16a in the face image matching unit 16 of the gate devices 2 and 3. . When registering a face pattern in the face image dictionary 16a, an image taken by a camera is used, but it may be captured by a photograph or the like.

  The management server 5 manages the staying status of employees, traders, and the like entering and leaving the gate devices 2 and 3, and includes a control unit 21, a memory unit 22, and an interface unit 23. The memory unit 22 is provided with a stay permission management master 31, a resident master 32, and an entry refusal list 33.

  As shown in FIG. 2, the stay permission management master 31 is registered with the stay permission time for each type of employee, trader A, trader B, etc. One hour after entry, Contractor B is limited to three hours after entry.

  As shown in FIG. 3, the resident master 32 is registered with a stay restriction time, a room entry time, an ID (employee number), the presence or absence of an employee, and a face image file. .09.518: 33 is registered, 2004.09.05 is registered as the entry time, 0 is registered as the ID (employee number), 0 (contractor) is registered as the presence / absence of the employee, and 040905 is registered as the face image file. . JPG is registered.

  As shown in FIG. 4, the entry refusal list 33 is registered with a time and a face image file. For example, 2004.8.01 is registered as a time, and 0401103. IMG (image file) is registered.

  The monitoring device 6 includes a personal computer including a keyboard 41, a display unit 42, and a control unit 43. The monitoring device 6 monitors an excess of stay time as a monitoring of this system. The excess list is extracted as shown in FIG. 5 and the contents of the generated list are displayed on the display unit 42 as shown in FIG.

  Next, with reference to the flowcharts shown in FIG. 7 and FIG. 8, the entrance processing from the entrance / exit 1 to the employee who performs face authentication and the trader who does not perform face authentication in the above configuration will be described.

  That is, when the human detection unit 12 detects that a human is approaching the gate device 2, this detection signal is supplied to the control unit 17 (ST1). Then, the control unit 17 captures an image including at least the face of the person approaching the camera 11 and registers it in the image memory 14 (ST2). Thereby, the control part 17 performs the process which produces | generates the feature vector (face pattern) of a face image from the image imaged with the camera 11 (ST3).

  Further, the control unit 17 collates the feature vector of the face image from the captured image with the feature vector of the face image registered for each employee in the employee face image dictionary 16a (ST4). As a result of this collation, the control unit 17 determines that at least one collation degree (similarity) with the feature vectors of all face images registered in the employee face image dictionary 16a has reached a predetermined value (ST5). ), Permission to enter from the entrance 1 is determined, the door opening / closing part 13 is driven, and the door is opened while the employee passes (ST6).

  At this time, the employee enters the room from the entrance 1.

  Further, the control unit 17 transmits the face image, the feature vector of the face image, information indicating the employee, and the room entry time to the management server 5 (ST7).

  Thereby, the management server 5 reads the stay permission time of the employee from the stay permission management master 31 based on the information indicating the employee supplied from the gate device 2 (ST8), and the stay permission restriction time as the stay permission time and the gate device 2 Is registered in the occupant master 32 by assigning the occupant number (unique number) to the face image supplied from the user, the information indicating the employee, and the entry time (ST9). The management server 5 transmits the feature vector of the face image to which the occupant number is assigned as occupant information to the gate device 3 (ST10). Thereby, the gate apparatus 3 registers the feature vector of the face image to which the occupant number supplied from the management server 5 is assigned in the face image dictionary 16a (ST11).

  In step 4, if the face image feature vector having a matching degree equal to or higher than a predetermined value is not registered in the employee face image dictionary 16 a, the control unit 17 inputs an authorization code using the operation display unit 10. Guide (ST12). In response to this guidance, the supplier inputs a permission code using the numeric keypad of the operation display unit 10. Thereby, when the control part 17 confirms that the inputted permission code is correct (ST13), it judges permission of the entrance from the said entrance / exit 1 (ST14), drives the door opening / closing part 13, and a trader passes. While doing this, the door is opened (ST15).

  At this time, the trader enters the room from the entrance 1.

  Further, the control unit 17 transmits the face image, the feature vector of the face image, the permission code indicating the trader A, and the room entry time to the management server 5 (ST16).

  As a result, the management server 5 reads one hour after entering the room as the stay permission time of the dealer A from the stay permission management master 31 using the permission code indicating the dealer A supplied from the gate device 2 (ST17). The stay permission restriction time is calculated from the time, and the resident person number (unique number) is assigned to the stay permission restriction time, the face image supplied from the gate device 2 and the entry time, and registered in the resident master 32. (ST18).

  Further, the management server 5 transmits the feature vector (supplied from the gate device 2) of the face image to which the occupant number is assigned as the occupant information to the gate device 3 (ST19). Thereby, the gate apparatus 3 registers the feature vector of the face image to which the occupant number supplied from the management server 5 is assigned in the face image dictionary 16a (ST20).

  Further, in the case of the trader B, similarly to the case of the trader A, the processing is performed by the above steps 12 to 20. However, the contents of the permission code are different, and the stay permission time is 3 hours after entering the room.

  In step 11, when the correct permission code is not confirmed, the control unit 17 determines that entry into the room through the doorway 1 is not permitted (ST 21), does not drive the door opening / closing unit 13, and opens the door. Keep it tightened.

  In addition, the control unit 17 transmits the face image and the room entry time to the management server 5 (ST22).

  Thereby, the management server 5 registers the face image supplied from the gate device 2 and the room entry time in the room refusal list 33 (ST23).

  Thereby, an alarm display is performed by the display unit 42 of the monitoring device 6 (ST24).

  Next, the exit process from the entrance 1 will be described with reference to the flowchart shown in FIG.

  That is, when the human detection unit 12 detects that a human is approaching the gate device 3, this detection signal is supplied to the control unit 17 (ST31). Then, the control unit 17 captures an image including at least a face by the camera 11 and registers it in the image memory 14 (ST32). Thereby, the control part 17 performs the process which produces | generates the feature vector (face pattern) of a face image from the image imaged with the camera 11 (ST33).

  Further, the control unit 17 collates the feature vector of the face image from the captured image with the feature vector of the face image of the occupant registered in the face image dictionary 16a (ST34). As a result of this collation, when the collation degree (similarity) with the feature vector of the face image registered in the face image dictionary 16a has reached a predetermined value (ST35), the control unit 17 determines whether or not it The permission to enter the room is determined, the door opening / closing part 13 is driven, and the door is opened while the person leaving the room passes (ST36).

  At this time, the person leaving the room leaves the entrance 1.

Further, the control unit 17 deletes the feature vector of the face image registered in the face image dictionary 16a corresponding to the person leaving the room (ST37), and transmits the resident person number and information indicating the room leaving to the management server 5. (ST38). Thereby, the management server 5 deletes the occupant information of the occupant number supplied from the gate device 3 from the occupant master 32 (ST39).
Next, processing when the staying time of the occupant exceeds the set time will be described with reference to the flowchart shown in FIG.

  For example, now, the control unit 21 of the management server 5 checks whether the stay limit time of each occupant information in the occupant master 32 exceeds the current time (ST41), as shown in FIG. The excess occupant information list 41 is generated in the internal memory 21a (ST42), and the excess occupant information corresponding to this list is transmitted to the monitoring device 6 (ST43). In addition, the control unit 21 checks the number of people outside the room and the number of people who have exceeded time based on the contents of the resident master 32 and transmits them to the monitoring device 6 (ST44).

  Thereby, as shown in FIG. 6, the monitoring device 6 displays the excess time corresponding to the face image of the resident who has exceeded the stay limit time on the display unit 42, as well as the number of outside occupants and the excess time. The number of persons is displayed (ST45).

  As a result, the number of people who have exceeded their stay time and the display of the face image of the person concerned can be displayed on the screen of the monitoring device, and strict management can be performed.

  When the occupant list is specified by the monitoring device, the face images of all occupants read from the occupant information in the occupant master 32 of the management server can be displayed on the display unit.

In the above example, the case where the management server manages the entrance / exit at one entrance / exit has been described, but the entrance / exit at a plurality of entrances / exits may be managed. However, in this case, information indicating the gate device is given during data transmission between the management server and the gate device, and a plurality of occupant masters are required.
In the above example, ecological information based on a face image is used. However, ecological information such as fingerprints and retinas may be used.

  As described above, the occupancy time management of employees (authenticated persons) and outpatients (non-authenticated persons) can be performed by one system with face authentication at the center.

  Also, location management can be ensured with a face image taken from the front (captured with a face orientation and size that can be clearly seen by humans). The occupant list and the occupant time excess list can display face images side by side in real time by the monitoring device.

  Further, the non-authenticated person can perform time management in the same manner as the authentication subject person by using the “stay permission time management master” for each trader by the auxiliary input means.

  In addition, since the person who has exceeded the occupancy time is displayed as a face image, occupancy management (monitoring) in which the person is accurately identified is possible.

  In addition, since the monitoring device can display the face image of the occupant in real time, the visibility is high.

The figure which shows schematic structure of the entrance management system for demonstrating one Embodiment of this invention. The figure for demonstrating a stay permission management master. The figure for demonstrating a resident master. The figure for demonstrating an entry refusal list. The figure for demonstrating an excess list. The figure for demonstrating the example of a display of an excess list. The flowchart for demonstrating a room entry process. The flowchart for demonstrating a room entry process. The flowchart for demonstrating a leaving process. The flowchart for demonstrating the process at the time of staying person's stay time exceeding setting time.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 1 ... Entrance / exit 2.3 ... Gate apparatus 4 ... Communication line 5 ... Management server 6 ... Monitoring apparatus 10 ... Operation display part 11 ... Camera 12 ... Human detection part 13 ... Door opening / closing part 14 ... Image memory 15 ... Main memory 16 ... Face Image collation unit 16a ... face image dictionary 16b ... authentication unit 17 ... control unit 21 ... control unit 22 ... memory unit 23 ... interface unit 31 ... stay permission management master 32 ... resident master 33 ... entry refusal list 41 ... keyboard 42 ... Display unit 43 ... Control unit.

Claims (4)

Registration means for registering an image including at least the face of the person to be authenticated;
Imaging means for capturing an image including at least a face of a person entering from a desired entrance;
A determination unit that determines whether the person is an authentication target person or an authentication non-target person by comparing an image including at least the face imaged by the imaging unit with an image including at least the face registered by the registration unit;
A management means for managing each staying time for the authentication target person and the authentication non-target person determined by the determination means;
An entrance management system characterized by comprising: Registration means for registering an image including at least the face of the person to be authenticated;
Imaging means for capturing an image including at least a face of a person entering from a desired entrance;
A determination unit that determines an authentication target person by collating an image including at least a face imaged by the imaging unit with an image including at least a face registered by the registration unit;
A setting means for setting information indicating a non-authenticated person;
Management means for managing respective stay times for the authentication target person determined by the determination means and the authentication non-target person set by the setting means;
An entrance management system characterized by comprising: Registration means for registering an image including at least the face of the person to be authenticated;
Imaging means for capturing an image including at least a face of a person entering from a desired entrance;
A determination unit that determines an authentication target person by collating an image including at least a face imaged by the imaging unit with an image including at least a face registered by the registration unit;
A setting means for setting information indicating a non-authenticated person;
Storage means for storing an image including at least a face imaged by the imaging means, corresponding to the information indicating the non-authenticated person set by the setting means;
Management means for managing each staying state for the authentication target person determined by the determination means and the authentication non-target person set by the setting means;
The display means for extracting the occupants according to the management state by the management means and displaying a list of images including faces corresponding to the occupants;
An entrance management system characterized by comprising: Registration means for registering an image including at least the face of the person to be authenticated;
Imaging means for capturing an image including at least a face of a person entering from a desired entrance;
A determination unit that determines an authentication target person by collating an image including at least a face imaged by the imaging unit with an image including at least a face registered by the registration unit;
A setting means for setting information indicating a non-authenticated person;
Storage means for storing an image including at least a face imaged by the imaging means, corresponding to the information indicating the non-authenticated person set by the setting means;
Management means for managing each entry time and stay time for the person to be authenticated determined by the determination means and the non-authenticated person set by the setting means;
A display means for extracting a person who has exceeded the stay time according to a management state by the management means and displaying a list of images including faces corresponding to the person who has exceeded the stay time;
An entrance management system characterized by comprising:

Images