JP2008252706A - Person supervisory system and person monitoring method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a person supervisory system capable of obtaining the facial image of a suspicious person in a short period of time, and preparing a listing of habitual offenders, in the order of registration. <P>SOLUTION: The figure monitoring system monitors a person who has committed an act of unauthorized removing of goods out of various kinds of shops, and when a doubtful person is detected, the facial image of a suspicious person is identified from the detected pictorial image for registration; and by having the registered facial image retrieved, as required, the list of registration by the same person is prepared, in the order of registration. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a person monitoring system and a person monitoring method for monitoring, for example, a person who has illegally carried out a product from various stores, a patient leaving without permission from a hospital room in various hospitals, and the like.

  Conventionally, for example, as a person monitoring system for monitoring a person who has illegally taken goods from various stores, a so-called shoplifting prevention system, a monitoring element such as a shoplifting prevention wireless tag is attached to the product in advance, A device that detects the passage of a monitoring element attached to a product taken out by a customer from the store using a monitoring sensor composed of a gate having a sensor function installed at an entrance / exit is known (for example, Patent Document 1).

In addition, in this type of shoplifting prevention system, when shoplifting is detected, a shoplifted person is visually identified by playing back a moving image that was shot at the same time. In addition, the determination of whether or not an addict is based on human memory, and the frequency of repeat offenses for the same person cannot be determined.
JP 2001-126158 A

  Since the person is visually identified by human judgment, it is necessary to carefully check the recorded moving image, which requires a lot of time and labor. In addition, because human beings determine whether they are addicts, they cannot be recognized because of ambiguous memory, they cannot be determined to be the same person because they are watching separately, and the subjectivity and the shape of the subject Therefore, it is often impossible to make an accurate judgment.

  Accordingly, an object of the present invention is to provide a person monitoring system and a person monitoring method that can automatically acquire a face image of a suspicious person in a short time and can list an addict in order of the number of registrations. .

  The person monitoring system of the present invention uses a monitoring sensor installed in the vicinity of a doorway to be monitored, a detection means for detecting the passage of a monitoring element that moves with the person passing through the doorway, and the detection means. When the passage of the monitoring element is detected, an imaging unit that captures an image including at least the vicinity of the monitoring sensor at that time, and a face detection unit that detects a human face image from the image captured by the imaging unit And a face image registration means for registering the face image detected by the face detection means in the face image storage means, and when the face image is registered by the face image registration means, the current face image and the face registered in the past The registration control means that controls to match the images and register the face images of the same person in the past if they exist in the past, and registration list creation conditions as required Registration list creation condition input means, search means for searching for the corresponding face image from the face image storage means based on the registration list creation conditions input by the registration list creation condition input means, and the face searched by the search means Registration list creating means for creating a registration list by counting images for each person and sorting the face images in the order of the count values;

  In addition, the person monitoring method of the present invention includes a detection step of detecting passage of a monitoring element that moves together with a person passing through the entrance / exit using a monitoring sensor installed in the vicinity of the entrance / exit to be monitored, and this detection When the passage of the monitoring element is detected by the step, an imaging step of capturing an image including at least the vicinity of the monitoring sensor at that time by an imaging unit, and a human face image from the image captured by the imaging step A face detection step for detecting, a face image registration step for registering the face image detected by the face detection step in the face image storage means, and when registering a face image by the face image storage step, the current face image and the past A registration control process that controls to match face images registered in, and if face images of the same person exist in the past, and creates a registration list if necessary A registration list creation condition input step for inputting a search result, a search step for searching for the corresponding face image from the face image storage means based on the registration list creation conditions input in the registration list creation condition input step, and the search step And a registration list creating step of creating a registration list by counting the searched face images for each person and sorting the face images in the order of the count values.

  According to the present invention, when a suspicious person is detected, the face image of the person is detected and registered from the captured image at that time, and the registered face image is searched as necessary. A person monitoring system and person monitoring method that can automatically obtain a face image of a suspicious person in a short time by creating a registration list in order of the number of registrations, and can list addicts in order of the number of registrations Can provide.

Hereinafter, embodiments of the present invention will be described with reference to the drawings.
FIG. 1 schematically shows a configuration of a person monitoring system according to the present embodiment. This person monitoring system includes a transmission gate 11 and a reception gate 12 as monitoring sensors, an abnormality detection unit 13 as detection means, a television camera (hereinafter simply referred to as a camera) 14 as imaging means, and face detection. A face detection unit 15 as a means, a notification unit 16 as a notification means, a face comparison unit and a face image registration means and a face comparison unit 17 as a registration control means, a face image database 18 as a face image storage means, a registration list creation condition It has a condition input unit 19 as an input unit, a search unit, and a list creation unit 20 as a registration list creation unit.

Hereinafter, each component will be described in detail.
The transmission gate 11 and the reception gate 12 are installed, for example, in the vicinity of a store entrance, and a customer M who visits or leaves the store passes between them. A weak radio wave is radiated from the transmission gate 11 toward the reception gate 12, and the radio wave is received by the reception gate 12.

  In addition, a monitoring element 10 such as a wireless tag is attached to each product in the store, and the product that the customer M takes out passes between the transmission gate 11 and the reception gate 12 together with the gate 11. , 12 are detected by changing the state of weak radio waves transmitted and received between them.

  Here, as the wireless tag, the transmission gate 11 and the reception gate 12 as the monitoring element 10, for example, a technique described in JP 2001-126158 A can be used.

  The abnormality detection unit 13 controls the transmission gate 11 and the reception gate 12 respectively, and the state change of the weak radio wave when the monitoring element 10 passes between the transmission gate 11 and the reception gate 12 is detected. By detecting it, an abnormality (shoplifting) is detected, and an abnormality detection signal is sent to the face detection unit 15 and the notification unit 16.

  The camera 14 is installed so as to capture at least an image between the transmission gate 11 and the reception gate 12, specifically an image including at least the face of the person M passing between the gates 11 and 12. For example, an image sensor or the like is mainly used.

  The face detection unit 15 detects the face area (face image) of the person M from the image captured by the camera 14. This is described in, for example, the literature (Fukui, Yamaguchi. “Face Feature Point Extraction by Combination of Shape Extraction and Pattern Matching”, Science Theory, (D), vol. J80-DH, No. 8, pp. 2170-. 2177, 1997) can be used to detect a face area with high accuracy. The detected face area information is sent to the face matching unit 17.

  When the notification unit 16 receives an abnormality detection signal from the abnormality detection unit 13 (when the passage of the monitoring element 10 through the gate is detected), the notification unit 16 outputs a warning to that effect to the clerk (or store clerk) by an alarm or the like, At that time, the face image of the person M detected by the face detection unit 15 is visually displayed on the screen.

  The face collation unit 17 registers (stores) the face image detected by the face detection unit 15 in the face image database 18, and at this time, collates the current face image with the face image registered in the past, and the same. If human face images exist in the past, they are registered together as the same group.

  The condition input unit 19 is used to input registration list creation conditions as necessary, and is mainly composed of, for example, a keyboard. As the registration list creation condition, for example, the number of cases registered by the same person to be listed, the period of listing, or the like can be considered.

  The list creation unit 20 searches for the corresponding face image from the face image database 18 based on the registration list creation conditions input by the condition input unit 19, counts the searched face images for each person, and the count value A registration list is created by sorting the face images in order, and the created registration list is printed out or displayed on the screen.

Next, face image registration processing in such a configuration will be described with reference to the flowchart shown in FIG.
The abnormality detection unit 13 controls the transmission gate 11 and the reception gate 12, and the abnormality detection unit 13 detects that the monitoring element 10 has passed between the transmission gate 11 and the reception gate 12. When the abnormality detection signal is output (step S1), the notification unit 16 outputs a warning to that effect to the staff (or store clerk) by an alarm or the like (step S2).

  Next, the face detection unit 15 detects a face area (face image) of a person (customer who took out the product without permission) M from the image captured by the camera 14 at this time, and the detected face image is notified to the notification unit 16 and The image is sent to the face collation unit 17 (step S3). The notification unit 16 displays the face image of the person M detected by the face detection unit 15 at that time on the screen to the attendant (or clerk) (step S4).

  The face collation unit 17 registers (stores) the face image detected by the face detection unit 15 in the face image database 18, and collates the current face image with the face image registered in the past (step). S5) It is checked whether or not face images of the same person exist in the past (step S6). If they exist, they are collectively registered as the same group (step S7).

  If the face image of the same person does not exist in the past as a result of the check in step S6, the current face image is registered as a new image (step S8). When the processes of steps S7 and S8 are completed, the next abnormality detection signal standby state is set (step S9).

Next, registration list creation processing for creating a registration list of face images will be described with reference to the flowchart shown in FIG.
First, a registration list creation condition is input by the condition input unit 19 (step S11). For example, it is assumed that the registration list creation condition is set so that five (five) or more face images registered as the same person are listed in descending order of the number of cases.

  When the registration list creation condition is input, the list creation unit 20 searches for the corresponding face image from the face image database 18 based on the input registration list creation condition (step S12), and uses the searched face image. Counting for each group of the same person, creating a registration list by sorting the face images in descending order of the count value (step S13), printing out the created registration list or displaying it on the screen (step S13) S14).

As described above, the registration list can be listed in descending order of the number of cases registered by the same person. Persons with a large number of registrations can be used as cautionary addicts and as shared information for all store workers, and can be used for future shoplifting prevention measures.
Further, since the face image of a suspicious person can be automatically acquired in a short time, the person's identification work by the supervisor is not required as in the prior art, and the person can be detected early.
In addition, since person detection / collation is automatically performed, mistakes caused by carelessness of the supervisor, intentional acquiescence, and the like can be prevented.
Furthermore, since the addict is automatically discriminated, an accurate discrimination result that does not depend on the ambiguity of the person's memory can be obtained.

  In the above-described embodiment, a case has been described in which a monitoring sensor is installed in the vicinity of an entrance of a store to detect passage (shoplifting) of a monitoring element attached to a product taken by a customer passing through the entrance. However, the present invention is not limited to this. For example, by installing a monitoring sensor in the vicinity of the entrance / exit of a hospital room in various hospitals, the monitoring element attached to the patient passing through the entrance / exit (patient The present invention can be applied in the same manner to detect the unintentional going-out), and the same effect as the above-described embodiment can be expected.

1 is a block diagram schematically showing a configuration of a person monitoring system according to an embodiment of the present invention. The flowchart explaining the registration process of a face image. The flowchart explaining the registration list creation process of a face image.

Explanation of symbols

  M ... person (customer), 10 ... monitoring element (wireless tag), 11 ... transmission gate (monitoring sensor), 12 ... reception gate (monitoring sensor), 13 ... abnormality detection unit (detection means), 14 ... Camera (imaging means), 15 ... Face detection section (face detection means), 16 ... Notification section (notification means), 17 ... Face verification section (face verification means, face image registration means, registration control means), 18 ... Face Image database (face image storage means), 19... Condition input section (registration list creation condition input means), 20... List creation section (search means, registration list creation means).

Claims (6)

Detection means for detecting the passage of the monitoring element that moves together with the person passing through the entrance / exit using a monitoring sensor installed near the entrance / exit to be monitored;
When the detection unit detects passage of the monitoring element, an imaging unit that captures an image including at least the vicinity of the monitoring sensor at that time;
Face detection means for detecting a face image of a person from an image captured by the imaging means;
Face image registration means for registering the face image detected by the face detection means in the face image storage means;
When registering a face image by this face image registration means, the current face image is compared with the face image registered in the past, and if face images of the same person exist in the past, they are registered together Registration control means to control;
Registration list creation condition input means for entering registration list creation conditions as required,
Search means for searching for a corresponding face image from the face image storage means based on the registration list creation condition input by the registration list creation condition input means;
A registration list creating unit that counts the face images retrieved by the retrieval unit for each person and creates a registration list by sorting the face images in the order of the count values;
A person monitoring system comprising:   When the detection means detects the passage of the monitoring element, the alarm means outputs a warning to that effect, and further includes a notification means for displaying and outputting the face image detected by the face detection means. The person monitoring system according to claim 1, wherein:   2. The person monitoring system according to claim 1, further comprising output means for outputting the registration list created by the registration list creation means as a visible image.   The person monitoring system according to claim 1, wherein the monitoring target is an entrance / exit of various stores, and the monitoring element is attached to a product brought out by the customer from the store.   2. The person monitoring system according to claim 1, wherein the monitoring target is an entrance / exit of a hospital room in various hospitals, and the monitoring element is attached to a patient going out of the hospital room. A detection process for detecting the passage of a monitoring element that moves together with a person passing through the doorway using a monitoring sensor installed near the doorway to be monitored;
When the passage of the monitoring element is detected by this detection step, an imaging step of capturing an image including at least the vicinity of the monitoring sensor at that time with an imaging unit;
A face detection step of detecting a human face image from the image captured by the imaging step;
A face image registration step of registering the face image detected by the face detection step in the face image storage means;
When registering a face image by this face image storing step, the current face image is compared with a face image registered in the past, and if face images of the same person exist in the past, they are registered together A registration control process to control;
Registration list creation condition input process for entering registration list creation conditions as necessary,
A search step of searching for a corresponding face image from the face image storage means based on the registration list creation condition input by the registration list creation condition input step;
A registration list creation step of counting the face images searched by this search step for each person and creating a registration list by sorting the face images in the order of the count values;
A person monitoring method comprising:

Images