JP7407562B2 - Risky behavior prevention devices, risky behavior prevention systems, risky behavior prevention methods, and risky behavior prevention programs - Google Patents

Description

The present invention relates to a risky behavior prevention device, a risky behavior prevention system, a risky behavior prevention method, and a risky behavior prevention program. More specifically, the present invention relates to a dangerous behavior prevention device, a dangerous behavior prevention system, a dangerous behavior prevention method, and a dangerous behavior prevention program suitable for preventing the occurrence of dangerous human behaviors such as falling, stumbling, and staggering.

There are cases where elderly people break bones when they fall and become bedridden. Therefore, it is necessary to prevent falls in homes and facilities where elderly people live. Furthermore, since behaviors such as stumbling and staggering may lead to falls in the future, it is also important to reduce the risk of such behaviors occurring.

For example, Patent Document 1 discloses a technology that uses a self-propelled traveling device to automatically detect obstacles that may impede walking, such as steps and slopes, in a residence, and to notify residents, etc. There is.

Patent No. 6393199

However, in the technique described in Patent Document 1, since obstacles are detected by the self-propelled traveling device, when a new obstacle occurs, it may not be possible to respond instantly. Furthermore, it only detects obstacles to objects using a group of sensors, and does not take into account human behavior, for example. Therefore, it is not possible to detect unnaturalness in a person's behavior and prevent falls.

Therefore, in the past, there was room for improvement in terms of more effectively preventing dangerous behaviors such as falling, stumbling, and staggering.

The present invention has been made in view of the above-mentioned current situation, and an object of the present invention is to provide a dangerous behavior prevention device, a dangerous behavior prevention system, a dangerous behavior prevention method, and a dangerous behavior prevention program that are excellent in preventing dangerous behavior. It is something.

In order to solve the above-mentioned problems and achieve the purpose, the present invention provides a dangerous behavior prevention device for preventing the occurrence of dangerous human behavior, and the present invention provides area information that is information about a predetermined area. an area information acquisition unit that acquires image information; a storage unit that stores detection information used to detect a dangerous situation in the predetermined area; and a storage unit that stores detection information used to detect a dangerous situation in the predetermined area based on the detection information and the area information. a detection unit that detects a situation; a notification unit that notifies when a dangerous situation is detected in the predetermined area by the detection unit; and an estimation that estimates the cause of the dangerous situation detected by the detection unit. and an updating unit that updates information on the cause to the detection information when the information on the cause of the dangerous situation estimated by the estimation unit is not included in the detection information. Features.

Moreover, the present invention is characterized in that, in the above-mentioned invention, the detection information includes information regarding dangerous behavior of a person.

Further, the present invention is characterized in that, in the above-mentioned invention, the detection information includes information regarding the environment of the predetermined area that corresponds to the cause of the dangerous behavior of the person.

Further, the present invention is characterized in that, in the above-mentioned invention, the detection information includes at least one of information regarding the movement of a person in a non-problematic state and information regarding an object in a non-problematic state.

Further, the present invention is characterized in that, in the above-mentioned invention, the detection information includes information regarding a situation when a person takes a dangerous action in the predetermined area.

Further, the present invention is characterized in that, in the above-mentioned invention, the cause of the dangerous situation is a common situation when a person takes a dangerous action a plurality of times in the predetermined area.

Further, the present invention is characterized in that, in the above-mentioned invention, the dangerous situation is a situation in which a dangerous object is placed in an area where a person's movement trajectory can take in the predetermined area.

Moreover, the present invention is characterized in that, in the above-mentioned invention, the detection unit detects a dangerous situation regarding an object placed on the floor in the predetermined area.

Moreover, the present invention is characterized in that, in the above-mentioned invention, the detection unit detects a dangerous situation regarding the object by changing the weight depending on the location where the object is placed.

Moreover, the present invention is characterized in that, in the above-mentioned invention, the detection unit detects an abnormal movement trajectory of a person in the predetermined area.

Moreover, the present invention is characterized in that, in the above-mentioned invention, the area information acquisition unit further acquires distance information from the imaging device to the measurement point in the predetermined area as the area information.

Moreover, the present invention is characterized in that, in the above-mentioned invention, the area information acquisition unit further acquires skeletal information of a person in the predetermined area as the area information.

Moreover, the present invention is characterized in that, in the above-mentioned invention, the estimation unit estimates the cause of the dangerous situation for each time period.

Moreover, the present invention is characterized in that, in the above-mentioned invention, the estimation unit estimates the cause of the dangerous situation for each person.

Further, in the above invention, the present invention provides that the area information acquisition unit acquires area information of a plurality of predetermined areas, and the detection unit detects a dangerous situation in the predetermined area for each of the predetermined areas, and The unit is characterized in that the cause of the dangerous situation detected by the detection unit is estimated for each of the predetermined areas.

Further, in the above invention, the present invention is characterized in that the detection unit determines whether the cause of the dangerous situation in a certain predetermined area estimated by the estimation unit occurs in another predetermined area. .

The present invention also provides a dangerous behavior prevention system for preventing the occurrence of dangerous human behavior, including an area information acquisition unit that acquires at least image information of a predetermined area as area information that is information of a predetermined area; a storage unit that stores detection information used to detect a dangerous situation in a predetermined area; a detection unit that detects a dangerous situation in the predetermined area based on the detection information and the area information; a notification unit that notifies when a dangerous situation is detected in the predetermined area; an estimation unit that estimates the cause of the dangerous situation detected by the detection unit; and the danger estimated by the estimation unit. The present invention is characterized by comprising an updating unit that updates information on the cause to the detection information when the information on the cause of the situation is not included in the detection information.

The present invention also provides a dangerous behavior prevention method for preventing the occurrence of dangerous human behavior, which comprises: acquiring at least image information of a predetermined area as area information that is information of a predetermined area; a detection step of detecting a dangerous situation in the predetermined area based on the area information and detection information stored in a storage unit to detect a dangerous situation in the predetermined area; a notification step for notifying when the dangerous situation is detected; an estimating step for estimating the cause of the dangerous situation detected by the detecting step; and information on the cause estimated by the estimating step is the detection information. If the detection information is not included in the detection information, the detection information may include an updating step of updating information on the cause to the detection information.

Further, the present invention is a dangerous behavior prevention program, and the present invention provides an acquisition means for acquiring at least image information of a predetermined area as area information, which is information of a predetermined area, from a computer in order to prevent the occurrence of dangerous behavior of a person. , a detection means for detecting a dangerous situation in the predetermined area based on the area information and detection information stored in a storage unit for detecting a dangerous situation in the predetermined area; Notification means for notifying when a dangerous situation is detected; estimation means for estimating the cause of the dangerous situation detected by the detection means; and information on the cause of the dangerous situation estimated by the estimation means. is characterized in that it functions as an updating means for updating information on the cause to the detection information when the detection information does not include the detection information.

According to the present invention, it is possible to realize a dangerous behavior prevention device, a dangerous behavior prevention system, a risky behavior prevention method, and a dangerous behavior prevention program that have excellent performance in preventing dangerous behavior.

FIG. 2 is a schematic diagram for explaining the outline of a method for preventing dangerous human behavior in Embodiment 1, and shows a scene in which a person trips over laundry scattered on the floor and falls. FIG. 2 is another schematic diagram for explaining the outline of the method for preventing dangerous human behavior in Embodiment 1, and shows a scene where laundry is scattered on the floor. 1 is a diagram illustrating the overall configuration of a dangerous behavior prevention system according to Embodiment 1. FIG. FIG. 2 is a schematic diagram showing image information (two-dimensional image) acquired by an imaging device and human skeletal information acquired from the image information. 5 is a schematic diagram showing distance information (depth image) acquired by the imaging device in the same scene as FIG. 4. FIG. 1 is a block diagram illustrating the configuration of a dangerous behavior prevention device according to Embodiment 1. FIG. 3 is a flowchart illustrating an example of a procedure for dangerous behavior prevention processing performed by the dangerous behavior prevention system according to the first embodiment.

EMBODIMENT OF THE INVENTION Hereinafter, preferred embodiments of a risky behavior prevention device, a risky behavior prevention system, a risky behavior prevention method, and a risky behavior prevention program according to the present invention will be described with reference to the drawings. The dangerous behavior prevention system according to the present invention can be applied to any desired area indoors or outdoors, but in the following, the system will be targeted at a nursing home where elderly people live, and will mainly be used to prevent dangerous behavior among elderly people. The intended purpose will be explained.

<Overview of methods for preventing dangerous human behavior>
First, an overview of the method for preventing dangerous human behavior in Embodiment 1 will be described using FIGS. 1 and 2. In this embodiment, first, as shown in FIG. image) and distance information (depth image).

Then, based on the acquired area information and detection information stored in advance in the storage unit, a dangerous situation in the target area is detected, and the detected dangerous situation is notified to the administrator. This detection information is reference information used to detect the presence or absence of a dangerous situation in the target area. For example, the detection information includes motion information when a person falls, obtained from past cases, and if motion information corresponding to the motion information when falling is obtained from area information, the motion information It is detected that the person has fallen and is in a dangerous condition.

Furthermore, in this embodiment, the cause of the dangerous situation is estimated from the detected dangerous situation. For example, as shown in FIG. 1, when it is detected that laundry is scattered on the floor around a person who has fallen, the laundry scattered on the floor is estimated to be the cause of the fall.

Then, the information on the estimated cause is compared with the detection information, and if the information on the estimated cause is not included in the detection information, the information on the cause is updated to the detection information. For example, in the above example, information on the laundry scattered on the floor (for example, class information indicating the category (type) and position information) is added to the detection information. As a result, as shown in Figure 2, it becomes possible to detect laundry scattered on the floor as a dangerous situation in the target area, and it is possible to detect the occurrence of a dangerous situation in the target area before a person actually falls. It becomes possible to notify the administrator. Therefore, dangerous behavior can be effectively prevented.

As mentioned above, in this specification, a dangerous situation refers to a situation in which a person is in danger and a situation in which the environment (including objects) poses a danger to a person (or a situation in which there is a possibility of causing a danger). It is inclusive. The former situation is preferably a situation in which a person is engaging in risky behavior, and the latter situation is preferably a situation in which the environment (including objects) causes or is likely to cause the person to engage in risky behavior. .

Further, the object may be any object that can exist in the target area, and includes not only easily movable objects such as laundry and baggage, but also equipment such as a built-in bed and chest of drawers.

<Overall structure of dangerous behavior prevention system>
Next, the overall configuration of the dangerous behavior prevention system of this embodiment will be explained using FIG. 3. As shown in FIG. 3, the dangerous behavior prevention system 1 of this embodiment is constructed for each private room in a nursing home where elderly people live, and the interior of each private room is the target area (predetermined area). becomes. The dangerous behavior prevention system 1 is connected to a plurality of imaging devices 2 that respectively image a plurality of target areas, and is communicably connected to each of the imaging devices 2, and based on area information acquired from each imaging device 2, each target area is The system is equipped with a dangerous behavior prevention device 3 as a management server that detects dangerous situations and notifies an administrator of the detected dangerous situations.

The dangerous behavior prevention device 3 also includes a monitor (display device) 4 with a built-in speaker that displays images acquired from the imaging device 2, and an input device 5 (for example, a mouse or keyboard) that allows the administrator to perform various input operations. are connected. Note that the monitor 4 and the input device 5 may be configured from a display device with an input function such as a touch panel display.

The imaging device 2 is installed on the ceiling of each private room in the nursing home, and acquires an image, here a moving image (video), of a target area including the living space in the private room of an elderly resident. The imaging device 2 also includes an RGB camera (video camera, not shown) capable of acquiring RGB color image information (two-dimensional image), as shown in FIG. It is equipped with a 3D camera (depth sensor, not shown) capable of acquiring a depth image in which each pixel has distance information (depth information), and acquires image information and distance information of the target area, as well as Image information and distance information are output to the dangerous behavior prevention device 3. Note that in this embodiment, grayscale image information (including infrared image information obtained by an IR camera) may be used as the image information instead of RGB color image information. Further, distance information (depth information) is information indicating the distance from the imaging device 2 (particularly the 3D camera) to the measurement point (the point being imaged).

In addition, the imaging device 2 further acquires human skeletal information in each target area, as shown in FIG. More specifically, the imaging device 2 uses existing person detection technology to detect body parts such as the head, shoulders, elbows, hands, hips, knees, and feet of a person from image information, and also detects the body parts from the image information. By tracking, human skeletal information is obtained in each target area. Since human image information is abstracted by converting it into skeletal information, the following effects can be obtained.
(1) It becomes impossible to determine who the target is with the human eye, leading to privacy protection.
(2) Leads to a reduction in the amount of information processing.

Regarding (1) above, for example, for an HD image, it is necessary to process 1920×1080×3 pieces of data, whereas for skeleton points, it is only necessary to process the amount of data for a dozen or so points at most. In addition, the amount of communication with the server side can be reduced. As a result, data communication time that tends to become a bottleneck when processing is performed by the dangerous behavior prevention device 3 can be reduced. In addition, when it is necessary to recognize each person in a video, such as in behavior recognition, processing time varies greatly depending on the number of people in the video, which negates the disadvantage of having to change the behavior recognition algorithm depending on the scene. is possible.

Here, the skeletal information of a person also includes information related to a movement trajectory (hereinafter also simply referred to as a trajectory) of the person. A trajectory is generated by connecting specific points at which specific parts (for example, the head) are detected.

<Configuration of dangerous behavior prevention device>
Next, the configuration of the dangerous behavior prevention device 3 will be further explained using FIG. 6. As shown in FIG. 6, the dangerous behavior prevention device 3 is composed of an information processing device having functions equivalent to a general server, and includes a control section (control means) 10, a storage section (storage means) 20, and a notification section (notification section). means) 30.

The control unit 10 has the functions of an area information acquisition unit (acquisition means) 11, a detection unit (detection unit) 12, an estimation unit (estimation unit) 13, and an update unit (update unit) 14. The control unit 10 includes, for example, a software program for implementing various processes, a CPU (Central Processing Unit) that executes the software program, and various hardware controlled by the CPU. Software programs and data necessary for the operation of the control unit 10 are stored in the storage unit 20.

Note that each part of the control unit 10 shown in FIG. 6 is realized by causing the CPU of the control unit 10 to execute a dangerous behavior prevention program. The dangerous behavior prevention program may be installed in the dangerous behavior prevention device 3 in advance, or may be operated as an application program that can run on a general-purpose OS, recorded on a computer-readable recording medium, or via a network. may be provided to the person.

The storage unit 20 is composed of a storage device such as a hard disk device or a non-volatile memory, and stores detection information 21 used to detect dangerous situations in the target area and historical information regarding the dangerous situations detected by the detection unit 12. A certain dangerous situation history 22 is stored.

The notification unit 30 notifies the administrator when the detection unit 12 detects a dangerous situation in the target area. Specifically, the monitor 4, for example, can be used as the notification section 30. In this case, when a dangerous situation is detected in the target area, information indicating the specific details of the dangerous situation is displayed on the monitor 4 along with the video of the target area, and the speaker (sound output means) of the monitor 4 outputs the danger situation. A sound is output to notify that the situation has been detected. Further, the notification unit 30 may notify the administrator of the cause of the dangerous situation estimated by the estimation unit 13.

The area information acquisition unit 11 performs a process of acquiring image information and distance information of each target area from each imaging device 2, and also performs a process of acquiring human skeletal information from each imaging device 2. Then, the acquired image information, distance information, and human skeleton information, that is, area information, are output to the detection unit 12.

The following effects can be obtained by acquiring distance information.
(1) Be able to determine whether an object is hanging on a wall or falling on the floor.
(2) Can generally distinguish whether a person is sleeping, sitting, or standing.
(3) It is possible to obtain a three-dimensional trajectory, which makes it possible to determine at a glance what kind of action the person took.

Regarding (2) above, if the angle of depression and the appearance are known, it is possible to make a rough estimate based on the aspect ratio of the upper and lower body and the whole body using only image information, but if the angle of depression is not determined, Estimation is difficult when there are a mixture of fat people, slender and tall people, or when the appearance of images is unknown. In other words, two-dimensional information requires various parameter adjustments according to the camera's attitude, arrangement, and situation, but three-dimensional information that includes distance information can be discriminated only from information about the camera's attitude and arrangement.

Regarding (3) above, for example, if you look back at the video later, it is not possible to tell whether you crouched or moved at each point just by looking at the trajectory on the video. It is necessary to confirm the situation. On the other hand, if the trajectory is based on three-dimensional information that includes distance information, the type of movement made can be seen at a glance.

The detection unit 12 performs a process of detecting a dangerous situation in each target area based on the detection information 21 stored in the storage unit 20 and the area information acquired from the area information acquisition unit 11. Specifically, the detection information 21 includes information indicating a dangerous situation (hereinafter also referred to as danger standard information), and the detection unit 12 detects that information corresponding to the danger standard information has been obtained from the area information. detect that the person or thing related to the information is in a dangerous situation. Further, the detection information 21 includes information indicating a non-dangerous situation (hereinafter also referred to as safety standard information), and the detection unit 12 detects when information that does not correspond to safety standard information is obtained from the area information. Detecting that a person or thing related to the information is in a dangerous situation. Information regarding the dangerous situation detected by the detection unit 12 (for example, information indicating the content of the detected dangerous situation, information indicating a situation other than the dangerous situation at the time the dangerous situation was detected, and information indicating the date and time when the dangerous situation was detected) information, etc.) are updated to the dangerous situation history 22 in the storage unit 20.

The detection unit 12 also performs processing to detect dangerous situations regarding objects placed on the floor of each target area. Thereby, objects placed on the floor that are likely to lead to dangerous human behavior can be detected more effectively, and the detection information 21 can be updated as necessary. Therefore, dangerous behavior can be more effectively prevented.

Further, the detection unit 12 changes the weight depending on the location where the object is placed and detects a dangerous situation regarding the object. Thereby, objects placed in places that are likely to lead to dangerous behavior can be detected more effectively, and the detection information 21 can be updated as necessary. Therefore, dangerous behavior can be more effectively prevented.

More specifically, for example, if an object is placed on the floor, and if it is placed higher than the floor (for example, a desk), it is more dangerous if it is placed on the floor. Highly sexual. Therefore, the former case is given higher risk than the latter case. That is, so that the detection unit 12 can more reliably detect objects located in more dangerous locations, the detection unit 12 increases the risk level of objects placed in more dangerous locations.

As a method for detecting a dangerous situation by the detection unit 12, for example, there is a method of using the usual flow line, distance information, and background difference. Specifically, for example, first, it is detected from the background difference that some object has been placed. Next, it detects that there is a possibility of stepping on or tripping over that object from the normal flow line. Note that the flow line that the user normally follows can be generated from the trajectory information. Then, the height, width, etc. of the object can be detected from the distance information, and it can be determined whether there is a possibility of actually stepping on or tripping over the object. If there is a possibility, it will be detected as a dangerous situation.

The estimating unit 13 performs a process of estimating the cause of the dangerous situation detected by the detecting unit 12 based on the dangerous situation history 22.

A specific example of the cause of the dangerous situation estimated by the estimating unit 13 includes items such as laundry scattered on the floor as described above. In addition, if it is detected that there are objects such as clothes or abandoned items around a person who has engaged in dangerous behavior in the normal flow line, the object on the flow line is presumed to be the cause of the dangerous behavior. Good too. Furthermore, if a situation where a carpet edge or a level difference in the floor surface exists around a person who has taken a dangerous action is detected, the step at a low position may be estimated as the cause of the dangerous action. Further, if dangerous behavior is detected in a state where the trajectory (location and speed) is different from usual, the trajectory may be estimated as the cause of the dangerous behavior. Further, if it is detected that the lighting environment at the time of the dangerous behavior is different from usual, this lighting environment may be estimated as the cause of the dangerous behavior. Furthermore, when it is detected that a specific person who always requires an attendant has taken a dangerous action without an attendant, the lack of an attendant may be presumed to be the cause of this person's dangerous action. Further, if it is detected that the floor was cleaned immediately before the dangerous behavior, the cleaning of the floor may be inferred as the cause of the dangerous behavior. It should be noted that all of the causes exemplified here can be said to be the causes of dangerous situations caused by humans, among other causes of dangerous situations.

Furthermore, the estimation unit 13 estimates the cause of the dangerous situation detected by the detection unit 12 for each time period. There is a high risk of dangerous behavior such as falling during the western sun and night hours, and the cause of dangerous situations can change depending on the time of day, so by estimating the cause of dangerous situations for each time period, behavior can be more effectively prevented.

Examples of time slots include time slots divided by predetermined time periods such as daytime and nighttime, morning and afternoon, and three hours.

Furthermore, the estimation unit 13 estimates the cause of the dangerous situation detected by the detection unit 12 for each person. Since the cause of a dangerous situation may differ from person to person, dangerous behavior can be more effectively prevented by estimating the cause of a dangerous situation for each person. In this case, information (for example, ID) for identifying each target individual may be set and used.

As a method for estimating the cause of a dangerous situation by the estimation unit 13, for example, the situation when a dangerous behavior occurs and the situation when an action similar to the dangerous action (but not a dangerous action) is taken are used. One method is to take the difference between the two. Specifically, for example, if it was daytime when the user traveled on the same trajectory last time, but it was night when he stumbled on the same trajectory this time, the difference in night is assumed to be the cause of the trip. Also, the last time I passed through the same flow line, the floor cleaning action was not recognized within a predetermined time (for example, within an hour), but this time when I slipped on that flow line, it was within a predetermined time (for example, within an hour). If a floor cleaning action is detected, the difference in floor cleaning is assumed to be the cause of the slip.

The cause of the dangerous situation estimated by the estimation unit 13 may be a common situation when a person takes a dangerous action multiple times in the target area. This allows a more accurate cause to be estimated as the cause of the dangerous situation.

When the information on the cause of the dangerous situation estimated by the estimation unit 13 is not included in the detection information 21, the updating unit 14 performs a process of updating the information on the cause into the detection information 21. As a result, it becomes possible to detect the cause as a dangerous situation in the target area, and the administrator can detect that a dangerous situation (the cause of the above dangerous situation) is occurring in the target area before a person actually takes a dangerous action. It becomes possible to notify the

On the other hand, if the detection information 21 includes information on the cause of the dangerous situation estimated by the estimating unit 13, the updating process by the updating unit 14 is not performed.

In this specification, "updating" means adding cause information to the detection information 21 and replacing the cause information with other information in the detection information 21 (i.e., replacing other information with other information). and adding information about the cause).

Furthermore, the detection unit 12 detects a dangerous situation in the target area for each target area, and the estimation unit 13 estimates the cause of the dangerous situation detected by the detection unit 12 for each target area. Thereby, dangerous situations can be detected and their causes can be estimated for each target area. As a result, the detection information 21 can be updated for each target area. That is, it becomes possible to adjust the detection information 21 to suit each target area and each person. Therefore, it is possible to detect dangerous situations specific to each target area and each person based on the adjusted detection information 21.

Further, the detection unit 12 determines whether the cause of the dangerous situation in a certain target area estimated by the estimation unit 13 occurs in another target area. As a result, if the cause of a dangerous situation detected in one target area also exists in another target area, the cause can be detected, making it possible to more effectively prevent dangerous behavior in each of multiple target areas. Can be done.

The detection unit 12 then detects an abnormal trajectory of a person in the target area. Thereby, dangerous behavior can be more effectively prevented. This is because if a person's trajectory is abnormal, there is a high possibility that it will lead to dangerous behavior.

Hereinafter, suitable information included in the detection information 21 will be explained.

The detection information 21 includes information regarding dangerous behavior of a person (hereinafter also referred to as dangerous behavior information) as risk reference information. Since dangerous human behavior is common to some extent regardless of individuals, by including the dangerous behavior information in the detection information 21, it is possible to identify dangerous human behavior as a dangerous situation based on the general detection information 21. can be detected more effectively, and the detection information 21 can be updated as necessary. Therefore, dangerous behavior can be more effectively prevented.

Types of information included in the dangerous behavior information include trajectory aspects (location and speed), distance fluctuations between predetermined parts of the person, and the like.

In this specification, "dangerous behavior" refers not only to human behavior that can be truly dangerous, such as falling or going wild (hereinafter also referred to as inherently dangerous behavior), but also to behavior that may lead to inherently dangerous behavior. It also includes human behavior (hereinafter also referred to as secondary dangerous behavior). Examples of secondary dangerous behaviors include stumbling, staggering, and wandering.

The inherently dangerous behavior may be known from past cases. Alternatively, the action may be an action that can be definitely or almost determined to be an inherently dangerous action, such as when the same action as a past case occurs.

The secondary risky behavior may be something that has not occurred in the past. It may also be an action that is considered dangerous, although it cannot be determined with certainty, because the change in the speed of a person's movements is abnormal, although it has not been seen in the past.

Furthermore, the detection information 21 includes, as danger standard information, information regarding the environment of the target area that corresponds to the cause of a person's dangerous behavior (hereinafter also referred to as dangerous environment information). As a result, it is possible to more effectively detect, as a dangerous situation, a situation in the environment of the target area that poses a danger to people (or a situation that has the potential to pose a danger) based on the hazardous environment information, and The detection information 21 can be updated accordingly. Therefore, dangerous behavior can be more effectively prevented.

It is preferable that the dangerous environment information is information about an environment that invites dangerous behavior by people living in the target area, and specifically, for example, information about the brightness (including the lighting environment) and temperature of the target area, and information about the time. Examples of information include information such as band, season, and weather, information regarding objects at a specific location within the target area (for example, class information, position information, etc.), and information about the floor plan of the target area.

Information regarding objects at specific locations within the target area includes, for example, if objects (e.g., small objects such as laundry or ornaments) are placed in a passageway (or a hallway), it could cause a person to fall. Therefore, information on objects (small objects) placed in the aisles can be mentioned.

In this way, the detection information 21 includes information about objects that are dangerous if placed in an area where a person can trace a trajectory in the target area. That is, the dangerous situation in the target area detected by the detection unit 12 may be a situation in which a dangerous object is placed in an area where a person can trace a trajectory in the target area. Thereby, dangerous behavior can be more effectively prevented. This is because if a dangerous object is placed in an area that can be traced by a person's trajectory, there is a high probability that it will induce dangerous behavior.

Note that the term "dangerous object" as used herein means an object that causes or is likely to cause dangerous behavior in humans.

In addition, the detection information 21 includes, as safety standard information, information regarding the movement of people in a non-problematic state (hereinafter also referred to as safe behavior information) and information regarding objects in a non-problematic state (hereinafter also referred to as safe object information). ). As a result, based on the detection information 21, dangerous situations can be detected such as problematic human movements (for example, abnormal trajectory of a person) or situations that pose a danger to people (even situations that may pose a danger). (good) can be detected more effectively, and the detection information 21 can be updated as necessary. Therefore, dangerous behavior can be more effectively prevented.

The safe behavior information includes, for example, the trajectory (location and speed) of a person in a safe state, the distance fluctuation of a predetermined part of a person in a safe state, and the like.

Examples of safe object information include the type and location information of objects in a safe condition.

Furthermore, the detection information 21 includes information regarding a causal relationship between an object and a person's dangerous behavior (hereinafter also referred to as causal relationship information) as risk reference information. As a result, by detecting objects that have a causal relationship with a person's risky behavior based on causal relationship information, it is possible to more effectively detect the dangerous behavior of a person who has a causal relationship with that object, and detect it as necessary. information 21 can be updated. Therefore, dangerous behavior can be more effectively prevented.

Note that the causal relationship information is information in which information about an object that corresponds to a cause and information about a person's dangerous behavior that occurs as a result are linked. Specifically, for example, there is information that a person trips because of an object (for example, a small object such as laundry or an ornament), and in this case, information related to the object (small object) is the cause. The resulting information is information related to dangerous behavior such as a person tripping and falling.

Furthermore, the detection information 21 includes, as danger reference information, information regarding the situation when a person takes a dangerous action in the target area (hereinafter also referred to as situation information at the time of occurrence of danger). Thereby, it is possible to more effectively detect the dangerous situation related to the situation information when a danger occurs, and the detection information 21 can be updated as necessary. Therefore, dangerous behavior can be more effectively prevented.

Furthermore, the detection information 21 includes, as danger reference information, information regarding a common situation when a person takes a dangerous action multiple times in the target area (hereinafter also referred to as common situation information at the time of occurrence of danger). That is, the dangerous situation in the target area detected by the detection unit 12 may be a common situation when a person takes a dangerous action multiple times in the target area. Thereby, the dangerous situation (common situation) related to the common situation information when a danger occurs can be detected more effectively, and the detection information 21 can be updated as necessary. Therefore, dangerous behavior can be particularly effectively prevented. This is because situations that are common when dangerous actions occur multiple times have a high probability of being the cause of those dangerous actions.

Here, the "situation" in the situation information at the time of danger occurrence and the common situation information at the time of danger occurrence includes the situation of the person who took the dangerous action and the environment at that time.

Information on the situation at the time of a danger involving a person and common situation information at the time of the occurrence of a danger include, for example, information that identifies the person at the time of the dangerous behavior (e.g. ID), and the trajectory of the person at the time of the dangerous behavior. (location and speed), changes in the distance of a predetermined part of the person when the dangerous behavior occurs, etc.

Information on the situation when an environmental hazard occurs includes, for example, information on the brightness (including the lighting environment) and temperature of the target area when a person engages in dangerous behavior, and the time period when a person engages in dangerous behavior. , information on the season, weather, etc., information on objects at a specific location within the target area when a person engages in dangerous behavior (e.g., class information, location information, etc.), target when a person engages in dangerous behavior Examples include information on the floor plan of the area.

According to the common situation information when a danger occurs, for example, if a person is likely to fall while walking in a specific place at a certain time, It can detect when a person is in a dangerous situation while walking in the area.

<Procedures for preventing dangerous behavior>
Next, the procedure of dangerous behavior prevention processing performed by the dangerous behavior prevention system 1 will be described using FIG. 7.

First, each imaging device 2 acquires image information, distance information, and human skeleton information of the target area (step S11).

Next, the area information acquisition unit 11 acquires image information and distance information of each target area from each imaging device 2 as area information, and also acquires human skeletal information in each target area from each imaging device 2. (Step S12).

Next, the detection unit 12 detects each target based on the detection information 21 stored in the storage unit 20 and the area information (image information, distance information, and human skeletal information) acquired from the area information acquisition unit 11. A dangerous situation in the area is detected (step S13).

If the detection unit 12 does not detect a dangerous situation in any target area (step S14: No), the process ends.

When the detection unit 12 detects a dangerous situation in any target area (step S14: Yes), the notification unit 30 notifies the administrator to that effect (step S15).

Next, the estimation unit 13 estimates the cause of the dangerous situation detected by the detection unit 12 (step S16). At this time, the cause of the dangerous situation estimated by the estimation unit 13 may be notified to the administrator by the notification unit 30.

If the information on the cause of the dangerous situation estimated by the estimation unit 13 is included in the detection information 21 (step S17: Yes), the process ends.

If the information on the cause of the dangerous situation estimated by the estimating unit 13 is not included in the detection information 21 (Step S17: No), the updating unit 14 updates the information on the cause to the detection information 21 (Step S17: No). S18), the process ends.

As explained above, in this embodiment, when a dangerous situation is detected in the target area (predetermined area), this is notified, the cause of the detected dangerous situation is estimated, and the estimated dangerous situation is corrected. If the information on the cause is not included in the detection information 21, the information on the cause is updated to the detection information 21. Therefore, when a dangerous situation is detected, not only is it notified, but also the danger The cause of the situation is estimated, and if the cause information is not included in the detection information 21 that is referred to for detecting a dangerous situation, the cause information is updated to the detection information 21. Therefore, thereafter, the situation in which the cause occurred can also be detected as a dangerous situation. Therefore, dangerous behavior can be effectively prevented.

In addition, in the above embodiment, a case has been described in which the dangerous behavior prevention device 3 is configured as one device, but each function of the dangerous behavior prevention device 3 may be realized by a distributed processing system that is appropriately distributed to a plurality of devices. .

Furthermore, in the above embodiment, a case has been described in which dangerous behavior of a person is prevented based on a dangerous situation in a target area detected from the detection information 21 and area information. or the cumulative amount of movement). Thereby, it is possible to understand the condition of a Parkinson's disease patient using the acquired motion amount information.

Although the embodiments of the present invention have been described above with reference to the drawings, the present invention is not limited to the above embodiments. Furthermore, the configurations of each embodiment may be combined or modified as appropriate without departing from the gist of the present invention.

As described above, the present invention is a technique useful for preventing dangerous human behavior.

1: Dangerous behavior prevention system 2: Imaging device 3: Dangerous behavior prevention device 10: Control unit 11: Area information acquisition unit 12: Detection unit 13: Estimation unit 14: Update unit 20: Storage unit 21: Detection information 22: Danger Situation history 30: Information department

Claims (19)

A dangerous behavior prevention device for preventing the occurrence of dangerous human behavior,
an area information acquisition unit that acquires at least image information of the predetermined area as area information that is information of the predetermined area;
a storage unit that stores detection information used to detect a dangerous situation in the predetermined area;
a detection unit that detects a dangerous situation in the predetermined area based on the detection information and the area information;
a notification unit that notifies when a dangerous situation in the predetermined area is detected by the detection unit;
an estimation unit that estimates the cause of the dangerous behavior included in the dangerous situation detected by the detection unit;
an updating unit that updates information on the cause to the detection information when the information on the cause estimated by the estimation unit is not included in the detection information;
A dangerous behavior prevention device characterized by comprising: 2. The dangerous behavior prevention device according to claim 1, wherein the detection information includes information regarding dangerous behavior of a person. 3. The dangerous behavior prevention device according to claim 1, wherein the detection information includes information regarding an environment in the predetermined area that corresponds to a cause of a person's dangerous behavior. The dangerous behavior according to any one of claims 1 to 3, wherein the detection information includes at least one of information regarding the movement of a person in a non-problematic state and information regarding an object in a non-problematic state. Preventive device. 5. The dangerous behavior prevention device according to claim 1, wherein the detection information includes information regarding a situation when a person takes a dangerous behavior in the predetermined area. The danger according to any one of claims 1 to 5, wherein the cause estimated by the estimation unit is a common situation when a person takes a dangerous action multiple times in the predetermined area. Behavior prevention device. The dangerous situation includes a situation where a dangerous object is placed in an area where a person's movement trajectory can take in the predetermined area.
The dangerous behavior prevention device according to any one of claims 1 to 6. The dangerous behavior prevention device according to any one of claims 1 to 7, wherein the detection unit detects a dangerous situation regarding an object placed on the floor in the predetermined area. The dangerous behavior prevention device according to any one of claims 1 to 8, wherein the detection unit detects a dangerous situation regarding the object by changing a weight depending on a location where the object is placed. The dangerous behavior prevention device according to any one of claims 1 to 9, wherein the detection unit detects an abnormal movement trajectory of a person in the predetermined area. The dangerous behavior prevention device according to any one of claims 1 to 10, wherein the area information acquisition unit further acquires distance information from an imaging device to a measurement point in the predetermined area as the area information. 12. The dangerous behavior prevention device according to claim 1, wherein the area information acquisition unit further acquires human skeletal information in the predetermined area as the area information. The dangerous behavior prevention device according to any one of claims 1 to 12, wherein the estimating unit estimates the cause for each time period. The dangerous behavior prevention device according to any one of claims 1 to 13, wherein the estimating unit estimates the cause for each person. The area information acquisition unit acquires area information of a plurality of predetermined areas,
The detection unit detects a dangerous situation in the predetermined area for each of the predetermined areas,
The method for preventing dangerous behavior according to any one of claims 1 to 14, wherein the estimation unit estimates the cause of the dangerous behavior included in the dangerous situation detected by the detection unit for each of the predetermined areas. Device. 16. The dangerous behavior prevention device according to claim 15, wherein the detection unit determines whether the cause in a certain predetermined area estimated by the estimation unit occurs in another predetermined area. A dangerous behavior prevention system that prevents the occurrence of dangerous human behavior,
an area information acquisition unit that acquires at least image information of the predetermined area as area information that is information of the predetermined area;
a storage unit that stores detection information used to detect a dangerous situation in the predetermined area;
a detection unit that detects a dangerous situation in the predetermined area based on the detection information and the area information;
a notification unit that notifies when a dangerous situation in the predetermined area is detected by the detection unit;
an estimation unit that estimates the cause of the dangerous behavior included in the dangerous situation detected by the detection unit;
an updating unit that updates information on the cause to the detection information when the information on the cause estimated by the estimation unit is not included in the detection information;
A dangerous behavior prevention system characterized by comprising: A dangerous behavior prevention method for preventing the occurrence of dangerous human behavior,
an acquisition step of acquiring at least image information of the predetermined area as area information that is information of the predetermined area;
a detection step of detecting a dangerous situation in the predetermined area based on detection information stored in a storage unit and the area information to detect a dangerous situation in the predetermined area;
When a dangerous situation is detected in the predetermined area by the detection step, a notification step of notifying the user to that effect;
an estimating step of estimating the cause of the dangerous behavior included in the dangerous situation detected by the detecting step;
an updating step of updating the cause information to the detection information when the cause information estimated in the estimation step is not included in the detection information;
A dangerous behavior prevention method characterized by comprising: computers to prevent the occurrence of dangerous human behavior,
acquisition means for acquiring at least image information of the predetermined area as area information that is information of the predetermined area;
Detection means for detecting a dangerous situation in the predetermined area based on detection information stored in a storage unit to detect a dangerous situation in the predetermined area and the area information;
Notification means for notifying when a dangerous situation in the predetermined area is detected by the detection means;
an estimating means for estimating the cause of the dangerous behavior included in the dangerous situation detected by the detecting means, and when information on the cause estimated by the estimating means is not included in the detection information; , updating means for updating information on the cause to the detection information;
A risky behavior prevention program that is characterized by functioning as a

Images