JP2012174058A - Specific event detection device, system, method and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a specific event detection device which can detect a specific event even in a situation that circumstances for observation vary.SOLUTION: This invention includes: object detection means 201 detecting an object from an image that a photographing means has photographed an appointed position; object discrimination means 202 discriminating a classification of the object which the object detection means 201 has detected; event detection rule identification means 203 identifying an event detection rule corresponding to a mobile body of the event detection rules to detect a specific event stored in advance based on the classification of the object which the object discrimination means 202 has discriminated; and detection determination means 204 to determine whether the state of the object satisfies the event to be provided in the event detection rules that the event detection rule identification means 203 has identified.

Description

  The present invention relates to a specific event detection device, a specific event detection system, a specific event detection method, and a specific event detection program for detecting a specific event from an image.

  Surveillance cameras and image processing devices are sometimes used to detect abnormal events in security operations. In this case, for example, the surveillance camera captures a predetermined place, and the image processing apparatus analyzes the captured image to detect an abnormal event. As a related technique, for example, Patent Document 1 describes a system capable of monitoring from a remote location within a monitoring area.

JP 2002-230654 A

  When an abnormal event is detected using a monitoring camera or an image processing apparatus in a security service, a difference from the normal state is detected from a long-term viewpoint. In other words, the size of the building and the location of entrances and windows are fixed, so the normal state is identified by fixedly setting the warning area and detection point, and the identified normal state and current state To detect the difference.

  However, when a technique for detecting a specific event using a monitoring camera or an image processing apparatus is applied to a business area such as a factory, it must be assumed that the situation constantly changes. For example, factory layout changes and the like are frequently performed as daily operations improve. In addition, in and out of logistics and warehouses, loading and unloading always occurs, and people's movement routes and loading places change over time.

  Therefore, an object of the present invention is to provide a specific event detection device, a specific event detection system, a specific event detection method, and a specific event detection program that can detect a specific event even in a situation where the environment of the monitoring target changes. And

  The specific event detection apparatus according to the present invention includes an object detection unit that detects an object from an image obtained by shooting a predetermined place by an imaging unit, an object identification unit that identifies a type of an object detected by the object detection unit, and an object identification unit. Event detection rule specifying means for specifying an event detection rule corresponding to an object among event detection rules for detecting a specific event stored in advance based on the type of the identified object, and event detection rule specifying means Detecting and determining means for determining whether or not the state of the object satisfies the event indicated by the event detection rule specified by.

  A specific event detection system according to the present invention includes an imaging unit and an information processing device. The information processing device detects an object from an image obtained by imaging the predetermined location by the imaging unit, and the object detection unit detects the object. Object identification means for identifying the type of the detected object, and event detection corresponding to the object in the event detection rule for detecting a specific event stored in advance based on the object type identified by the object identification means An event detection rule specifying unit that specifies a rule, and a detection determination unit that determines whether or not the state of the object satisfies the event indicated by the event detection rule specified by the event detection rule specifying unit .

  In the specific event detection method according to the present invention, an imaging unit detects an object from an image obtained by imaging a predetermined location, identifies the type of the detected object, and stores a specific stored in advance based on the identified type of the object. An event detection rule corresponding to an object among event detection rules for detecting an event is specified, and it is determined whether or not the state of the object satisfies an event indicated by the specified event detection rule .

  The specific event detection program according to the present invention includes a computer that performs object detection processing for detecting an object from an image obtained by photographing a predetermined place, object identification processing for identifying the type of the detected object, and type of the identified object. Event detection rule specifying process for specifying an event detection rule corresponding to an object among event detection rules for detecting a specific event stored in advance, and an event indicated by the specified event detection rule And a detection determination process for determining whether or not the state of the object is satisfied.

  According to the present invention, it is possible to detect a specific event even in a situation where the environment to be monitored changes.

It is a block diagram which shows an example of a structure of the specific event detection system by this invention. It is a functional block diagram which shows an example of a function structure of a specific event detection system. It is explanatory drawing which shows an example of a detection rule. It is a flowchart which shows the process example which a specific event detection system performs.

  Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing an example of the configuration of a specific event detection system according to the present invention. As shown in FIG. 1, the specific event detection system includes a monitoring camera 100, an image analysis device 200, a detection rule database storage device 250, and a display device 300. The monitoring camera 100 and the image analysis device 200 are connected to each other via a communication network such as a LAN or the Internet. The image analysis device 200, the detection rule database storage device 250, and the display device 300 can communicate with each other via a wired or wireless line. Note that the image analysis device 200 and the detection rule database storage device 250 may be realized by the same device.

  The monitoring camera 100 has a function of photographing a predetermined place (for example, an inspection work place in a warehouse) and outputting an image (for example, a moving image or a still image) to the image analysis apparatus 200.

  Specifically, the image analysis apparatus 200 is realized by an information processing apparatus such as a personal computer that operates according to a program.

  Specifically, the detection rule database storage device 250 is realized by a storage device such as a magnetic disk device or an optical disk device. The detection rule database storage device 250 stores a detection rule database (details will be described later).

  Specifically, the display device 300 is realized by a display device or the like. The display device 300 has a function of displaying an alert or the like output from the image analysis device 200.

  Next, the functional configuration of the specific event detection system will be described. FIG. 2 is a functional block diagram illustrating an example of a functional configuration of the specific event detection system. As shown in FIG. 2, the image analysis apparatus 200 includes an object detection unit 201, an object identification unit 202, an event detection rule specifying unit 203, and a detection determination unit 204.

  Specifically, the object detection unit 201 is realized by a CPU of an information processing apparatus that operates according to a program. The object detection unit 201 has a function of detecting a moving body from an image output from the monitoring camera 100. In the present embodiment, the moving body may or may not exist in a certain place with time, and specifically, is a worker or an article.

  Specifically, the object identification unit 202 is realized by a CPU of an information processing apparatus that operates according to a program. The object identification unit 202 has a function of identifying the type of moving object detected by the object detection unit 201.

  The object identification unit 202 identifies the object based on information such as the shape, size, color, and attached characters and marks of the detected moving object. Specifically, the object identification unit 202 compares information such as the detected shape, size, color, and attached characters and marks with feature information stored in advance for each type of moving object. Then, the type is identified by whether or not the similarity between the two is equal to or greater than a predetermined threshold.

  If it demonstrates using the example shown in FIG. 3, for example, if the similarity of the mobile body detected from the image and the template image of a worker or an article is a predetermined value or more, the object identification means 202 judge. For example, the object identification unit 202 determines whether the color information of the moving body detected from the image matches the color of the work clothes or the article, and identifies the type of the moving body based on the determination result. Further, for example, the object identifying unit 202 determines that the object is a suspicious object (that is, an unmanaged object) when the feature of the moving object detected from the image does not match the feature information of any moving object stored in advance. .

  Further, for example, the object identification unit 202 may identify the type of the detected moving object based on information obtained using another technique such as RFID (Radio Frequency Identification) or an authentication technique.

  Specifically, the event detection rule specifying unit 203 is realized by a CPU of an information processing apparatus that operates according to a program. The event detection rule specifying unit 203 has a function of specifying a detection rule corresponding to the moving object from the detection rule database storage device 250 based on the type of the moving object identified by the object identifying unit 202.

  Specifically, the detection determination unit 204 is realized by a CPU of an information processing apparatus that operates according to a program. The detection determination unit 204 has a function of determining whether or not the detected moving object satisfies the detection rule specified by the event detection rule specifying unit 203.

  Next, the detection rules stored in the detection rule database storage device 250 will be described. FIG. 3 is an explanatory diagram illustrating an example of a detection rule. In the example shown in FIG. 3, the operation condition, the position, and the detection event are determined for each type of the detected moving body. In the present embodiment, the operation condition, the position, and the detection event are defined as detection rules for each type of moving object. However, the present invention is not limited to this, and a condition such as a time zone may be added.

  The detection rule shown in FIG. 3 assumes an operation in a warehouse. In a warehouse, when an article is received / shipped, the article is temporarily placed on the wall and inspected. In this case, it is defined in business operations as a space where objects can be placed near the wall. Therefore, if a person stops at the wall, the work is hindered. Therefore, if the person has stopped for a predetermined period, it is necessary to issue a warning and quickly move to another place.

  Furthermore, when an article is temporarily placed near the wall, inspection work must be performed in the vicinity. In this case, in order to collate the article with the article list, a person works while stopping for a predetermined period. Therefore, people must be allowed to stop around objects placed near the wall. Therefore, when a moving body is detected by the wall, the type of the detected moving body is determined as being an article or a person. If the type is an article, it may be placed near the wall, so an alert is not sent, and if it is a person, an alert is sent. In addition, if an article is placed near the wall, the person is allowed to stop nearby and does not send an alert. In the example shown in FIG. 3, such a detection rule is shown.

  Next, the operation of the specific event detection system will be described. FIG. 4 is a flowchart showing an example of processing executed by the specific event detection system. In this embodiment, it is assumed that it is operating in a warehouse.

  In a state where the specific event detection system is activated, the monitoring camera 100 captures an image of the warehouse and outputs an image to the image analysis apparatus 200 (step S1). Note that the monitoring camera 100 continuously captures the inside of the warehouse and outputs an image to the image analysis apparatus 200.

  Next, the object detection unit 201 detects a moving body from the image output by the monitoring camera 100 (step S2).

  Next, the object identification unit 202 identifies the type of the moving body detected by the object detection unit 201 (steps S3 and S4). For example, the object identification unit 202 identifies information (step S4) such as the shape and size of the detected moving body (step S3), color, attached character, mark, etc., and the identified information and the type of the moving body in advance. The feature information stored for each is compared, and the type of the moving object is identified based on whether the similarity between the two is equal to or greater than a predetermined threshold. Note that the object identification unit 202 may identify the type of the detected moving object based on information obtained by using another technique such as RFID or an authentication technique.

  Next, the event detection rule specifying unit 203 specifies a detection rule corresponding to the moving body detected from the detection rule database storage device 250 based on the type of the moving body identified by the object identifying unit 202. Then, the event detection rule specifying unit 203 returns the process to step S1 if it cannot be specified, and proceeds to step S6 if it can be specified (step S5). Note that the order of steps S5 and S6 may be reversed or may be performed in parallel.

  When the detection rule is specified, the detection determination unit 204 specifies the position of the moving object from the image, and tracks the trajectory of the moving object in the image continuously output by the monitoring camera 100 (that is, the change in the position of the moving object). Seeking). Here, for example, the detection determination unit 204 may obtain the distance between the moving body and a wall, a pillar, or another detected moving body as the position information indicating the position of the moving body. Specifically, the detection determination unit 204 obtains a distance between an object in the image and a person in the image plane, and determines whether the distance is within a threshold. For example, when the monitoring camera 100 is fixed, the detection determination unit 204 may determine that there is an object at the inspection location if there is an object in a predetermined area in the image at a predetermined ratio or more. .

  And the detection determination means 204 transfers a process to step S7, when the position of a moving body can be specified, and complete | finishes a process when it cannot specify (step S6).

  When the position of the moving body can be specified, the detection determination unit 204 determines whether or not the moving body satisfies the detection rule (step S7). For example, the detection determination unit 204 determines based on the operation condition, position, and detection event determined for each moving object shown in FIG.

  Specifically, the detection determination unit 204 determines whether or not another moving body (object) detected in the image exists as an operation condition when the type of the detected moving body is a worker. To do. Further, the detection determination unit 204 determines whether or not the position specified in step S6 matches the position predetermined for each moving object shown in FIG. Further, the detection determination unit 204 determines whether or not the track of the moving body being tracked matches the detection event shown in FIG. Specifically, the detection determination unit 204 determines whether or not the detected moving object exists at a position determined by the detection rule for a predetermined period.

  According to the example shown in FIG. 3, when the detected moving body is a worker and there is another object detected as a moving body, the detection determining unit 204 assumes that the worker is within 1 m around the object. It is determined that it may be stopped. Further, when the detected moving body is a worker and there is no other object detected as a moving body, the detection determining unit 204 determines whether or not the worker has stopped for 10 seconds or more in all areas. . When the worker has stopped for 10 seconds or more, the detection determination unit 204 determines that the state of the detected moving object (worker) satisfies the event indicated by the detection rule.

  In addition, for example, when the detected moving body is a management target object, the detection determination unit 204 is 30 seconds or more other than the vicinity of the wall / column or the management target object determined by the detection rule. Or it is determined whether it exists for 15 seconds or more. If both conditions are satisfied, the detection determination unit 204 determines that the state of the detected moving object (managed object) satisfies the event indicated by the detection rule.

  As described above, the detection determination unit 204 determines that the detection rule is satisfied when the detected moving object satisfies all the operating conditions, positions, and detection events determined in advance for each type of the moving object. .

  If it is determined that the detection rule is satisfied, the detection determination unit 204 controls to display an alert on the display device 300 (step S8). When it is determined that the detection rule is not satisfied, the detection determination unit 204 ends the process. In this embodiment, an alert is displayed on the display device 300 to warn the worker. However, the present invention is not limited to this. For example, a warning sound may be emitted.

  Next, the effect of this embodiment will be described. First, as an existing technique, there is a method of defining and using a detection rule based on position information in order to detect a specific event from an image by an image analysis apparatus. For example, a detection area on a line is defined at a specific position in a camera image, and a moving object that passes through the line is detected. In addition, for example, a polygonal detection area is defined at a specific position, and the entry into the detection area is detected.

  However, in a place where the environment changes, such as a business area such as a factory, a detection rule based on position information cannot be defined. For example, in the inspection work place, in the stage before the inspection object is placed, it is dangerous for a person to stop at the place to place the inspection object. He must be allowed to stop for inspection work. If we try to define this with a detection rule based on location information, we define exactly where the object is placed, define it as a place where only the object can be placed, and around it It is defined as a place where you can stop. However, in reality, it is difficult to accurately define a place where an object is placed and to place an object within the definition range. Also, when an object is brought and placed, it cannot be defined by a detection rule based on position information even when a person should not approach the area.

  In contrast, in the present embodiment, detection rules are defined in advance for each type of mobile object, and when a mobile object is detected, the detected mobile object type is identified, and the detection rule defined for the identified type is detected. Alerts only when the conditions are met. In the present embodiment, a detection rule in which an event is defined based on a positional relationship with a surrounding object (for example, a wall, a pillar, or another moving body) can be used.

  Therefore, in this embodiment, a specific action can be detected even in a situation where the environment of the monitoring target such as a factory changes. This is because a detection rule determined for each type of mobile object is used instead of a detection rule determined based on position information used in a general method.

  Moreover, in this embodiment, a flexible detection rule can be set easily. When using a detection rule based on position information, it is necessary to accurately define the place where an object is placed or where a person may stop, but in this embodiment, an event that is detected for each type of mobile object This is because the conditions (for example, the positional relationship with other moving bodies) can be determined.

  In addition, not only an image but various sensors, RFID, a biometrics authentication technique, etc. may be used for determination of the classification of a moving body.

  From the above, in this embodiment, the characteristic configuration of the specific event detection apparatus as shown in the following (1) to (7) is shown.

  (1) The specific event detection device is an object detection unit (for example, an object) that detects an object (for example, an operator or an article) from an image in which a shooting unit (for example, realized by the monitoring camera 100) captures a predetermined place. Based on the type of the object identified by the object identifying means (for example, realized by the object identifying means 202) and the object identifying means for identifying the type of the object detected by the object detecting means. Event detection rule specifying means (for example, the detection rule shown in FIG. 3) corresponding to the object among the event detection rules for detecting a specific event stored in advance (for example, the detection rule shown in FIG. 3) The state of the object satisfies the event indicated by the event detection rule specified by the event detection rule specifying means). Determining detection determining means whether Luke (e.g., as implemented by the detection determining means 204), characterized in that it comprises a.

  (2) In the specific event detection device, when the detection determination unit determines that the state of the object satisfies the event indicated by the event detection rule, a warning unit that issues a warning (for example, realized by the image analysis device 200 and the display device 300) May be included.

  (3) In the specific event detection device, the warning unit may be configured to control the display device (for example, the display device 300) to display the warning.

  (4) In the specific event detection apparatus, the object identification means includes feature information indicating the characteristics of the object detected by the object detection means (for example, information on the shape and size of the object, color, attached characters and marks, etc. For example, the information is acquired using various sensors, RFID, and other technologies), and the type of the object is identified based on the acquired feature information and the feature information of the object stored in advance. May be.

  (5) The specific event detection device includes storage means (for example, realized by the detection rule database storage device 250) that stores an event detection rule determined for each object type, and the event detection rule specification means Based on the type of the object identified by the identification means, the event detection rule corresponding to the object among the event detection rules stored by the storage means is specified, and the detection determination means is the event detection rule specified by the event detection rule specification means It may be configured to determine whether or not the state of the object satisfies the event indicated by.

  (6) In the specific event detection apparatus, the detection determination unit may be configured to determine whether or not there is an object at a position indicated by the event detection rule.

  (7) In the specific event detection apparatus, the detection determination unit may be configured to determine whether or not the object is in a predetermined position for a period of time indicated by the event detection rule.

  The present invention can be applied to applications for detecting specific events in the logistics / warehouse industry, stores, and the like.

DESCRIPTION OF SYMBOLS 100 Surveillance camera 200 Image analysis apparatus 201 Moving body detection means 202 Moving body identification means 203 Event detection rule specification means 204 Detection determination means 250 Detection rule database storage device 300 Display apparatus

Claims (10)

Object detection means for detecting an object from an image obtained by the imaging means capturing a predetermined place;
Object identifying means for identifying the type of the object detected by the object detecting means;
Event detection rule specifying means for specifying an event detection rule corresponding to the object among event detection rules for detecting a specific event stored in advance based on the type of the object identified by the object identification means When,
A specific event detection apparatus comprising: a detection determination unit that determines whether or not the state of the object satisfies the event indicated by the event detection rule specified by the event detection rule specifying unit. The specific event detection device according to claim 1, further comprising a warning unit that issues a warning when the detection determination unit determines that the state of the object satisfies the event indicated by the event detection rule. The specific event detection device according to claim 2, wherein the warning means controls the display device to display a warning. The object identification unit acquires feature information indicating the feature of the object detected by the object detection unit, and identifies the type of the object based on the acquired feature information and the feature information of the object stored in advance. The specific event detection device according to claim 1. Including storage means for storing event detection rules determined for each type of object,
The event detection rule specifying means specifies an event detection rule corresponding to the object among the event detection rules stored in the storage means based on the type of the object identified by the object identification means,
The detection determination unit determines whether or not the state of the object satisfies an event indicated by the event detection rule specified by the event detection rule specifying unit. The specific event detection device according to 1. The specific event detection apparatus according to any one of claims 1 to 5, wherein the detection determination unit determines whether or not there is an object at a position indicated by the event detection rule. The specific event detection device according to any one of claims 1 to 6, wherein the detection determination unit determines whether or not the object is at a predetermined position for a period of time indicated by the event detection rule. Photographing means;
An information processing device,
The information processing apparatus includes:
Object detecting means for detecting an object from an image obtained by photographing the predetermined place by the photographing means;
Object identifying means for identifying the type of the object detected by the object detecting means;
Event detection rule specifying means for specifying an event detection rule corresponding to the object among event detection rules for detecting a specific event stored in advance based on the type of the object identified by the object identification means When,
A specific event detection system comprising: a detection determination unit that determines whether or not the state of the object satisfies the event indicated by the event detection rule specified by the event detection rule specifying unit. An imaging means detects an object from an image obtained by photographing a predetermined place,
Identifying the type of the detected object;
Based on the identified type of the object, identify an event detection rule corresponding to the object among the event detection rules for detecting a specific event stored in advance,
A method for detecting a specific event, comprising: determining whether the state of the object satisfies an event indicated by the specified event detection rule. On the computer,
An object detection process in which an imaging unit detects an object from an image obtained by imaging a predetermined place;
An object identification process for identifying the type of the detected object;
An event detection rule specifying process for specifying an event detection rule corresponding to the object among event detection rules for detecting a specific event stored in advance based on the type of the identified object;
A specific event detection program for executing a detection determination process for determining whether or not the state of the object satisfies an event indicated by the specified event detection rule.

Images