Unattended object monitoring device, unattended object monitoring system equipped with the device, and unattended object monitoring method

Abstract

Enables unattended objects that do not require an alarm to be excluded from the alarm target. [Solution means] An unattended object monitoring device (3) is used to detect unattended objects left in a monitoring area based on an image of the monitoring area captured by a camera device, and issue an alarm when an unattended object is detected. The device includes: a processor for performing processing for detecting unattended objects, the processor being configured to: detect objects present in the monitoring area from the captured image; track each detected object between the captured images; based on the tracking results for each detected object, detect an object that has not moved for more than a predetermined time as an unattended object; determine whether the detected unattended object exists in a mask area preset in the monitoring area; and in the case where it is determined that the detected unattended object exists in the mask area, exclude the unattended object from the alarm target.

Description

Technical Field

[0001] The present invention relates to an unattended object monitoring device, an unattended object monitoring system equipped with the unattended object monitoring device, and an unattended object monitoring method, which are used to detect an unattended object left in a monitoring area based on an image of the monitoring area captured by a camera device, and issue an alarm when the unattended object is detected. Background Art

[0002] One aspect of the monitoring of public spaces (such as stations, airports, commercial facilities, shopping areas, and underground shopping areas, etc.) used by the public is to closely monitor suspicious unattended objects in these spaces for the purpose of crime prevention. Usually, in order to closely monitor unattended objects in public spaces, an observer monitors an image of a monitoring area captured by a monitoring camera (camera device). For such a monitoring system, a technique for automatically detecting unattended objects based on an image of the monitoring area is required, thereby reducing the burden on the observer and improving the detection accuracy of unattended objects.

[0003] There are various known techniques (such as the technique disclosed in Patent Document 1, etc.) for automatically detecting unattended objects based on an image of a monitoring area captured by a camera device. Patent Document 1 teaches an image monitoring system configured to: calculate the difference between an initially captured background image of a monitoring area and each image of the same area captured at a predetermined time interval, thereby cutting out a human image; compare the human images sequentially acquired at the predetermined time interval with each other, thereby detecting an object image that can be separated from the human image; when such an object image is detected, measure the unattended time during which the object is unattended; and when the unattended time is greater than a predetermined time period, issue an alarm.

[0004] Prior Art Documents

[0005] Patent Documents

[0006] Patent Document 1: Japanese Patent Laid-Open No. 1-245395 Summary of the Invention

[0007] Problems to be Solved by the Invention

[0008] However, the system of Patent Document 1 issues an alarm for any detected unattended object, regardless of the type of the detected unattended object. Therefore, the following problem exists in this system: an alarm is issued for an unattended object for which an alarm is not required (such as a store sign (storefront sign) on the ground, a closed - store cleaning sign, a construction safety cone, an information board, etc.). When an alarm is issued for an unattended object for which an alarm is not required, the observer needs to observe an increased number of unattended objects, which increases the burden on the observer. This problem deteriorates in the case where there are a large number of unattended objects for which an alarm is not required in the surveillance area, or in the case where the surveillance area is large. From this perspective, it would be beneficial to exclude unattended objects for which an alarm is not required from the alarm - target objects for which an alarm is issued, as this reduces the burden on the observer.

[0009] The present invention has been made in view of such problems of the prior art, and the main object of the present invention is to provide an unattended object surveillance device, an unattended object surveillance system equipped with the unattended object surveillance device, and an unattended object surveillance method, which can exclude detected unattended objects for which an alarm is not required from the alarm - target objects for which an alarm is issued.

[0010] Solution to the problem

[0011] One aspect of the present invention provides an unattended object surveillance device for detecting unattended objects left in a surveillance area based on an image of the surveillance area captured by a camera device and issuing an alarm when an unattended object is detected. The unattended object surveillance device includes: a processing device for performing processing operations required for detecting unattended objects, wherein the processing device is configured to: detect objects present in the surveillance area from the captured image captured by the camera device and track each detected object between captured images; detect an object that has not moved for more than a predetermined time as an unattended object based on the tracking result for each detected object; determine whether the detected unattended object exists in a mask area preset in the surveillance area; and exclude the detected unattended object from the alarm - target objects for which an alarm is issued when it is determined that the detected unattended object exists in the mask area.

[0012] Effect of the invention

[0013] According to the present invention, unattended objects for which an alarm is not required can be excluded from the alarm - target objects for which an alarm is issued. As a result, particularly in the case where there are a large number of unattended objects for which an alarm is not required in the surveillance area, or in the case where the surveillance area is large, the burden on the observer can be reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 FIG. is a diagram showing a schematic configuration of an unattended object monitoring system according to a first embodiment of the present invention;

[0015] Figure 2 FIG. is a block diagram showing a schematic configuration of an unattended object monitoring device of the unattended object monitoring system;

[0016] Figure 3 FIG. is a diagram showing a mask area setting screen and an unattended object type setting screen;

[0017] Figure 4 FIG. is a diagram showing a mask area setting screen and a usage time period setting screen;

[0018] Figure 5 FIG. is a diagram showing an example of mask area information;

[0019] Figure 6 FIG. is a diagram showing an example of a determination operation performed by an alarm target determination unit of the unattended object monitoring device;

[0020] Figure 7 FIG. is a diagram showing an example of a determination operation performed by an alarm target determination unit of the unattended object monitoring device;

[0021] Figure 8 FIG. is a diagram showing an example of a determination operation performed by an alarm target determination unit of the unattended object monitoring device;

[0022] Figure 9 FIG. is a diagram showing an example of a notification image;

[0023] Figure 10 FIG. is a flowchart showing an operation process of an operation performed by the unattended object monitoring device;

[0024] Figure 11 FIG. is a diagram showing a schematic configuration of an unattended object monitoring system according to a second embodiment of the present invention;

[0025] Figure 12 FIG. is a diagram showing an example of a representative position of an unattended object; and

[0026] Figure 13 FIG. is a diagram showing an example of changing the geometry of a mask area. DETAILED DESCRIPTION OF THE INVENTION

[0027] A first aspect of the present invention made to achieve the above object is an unattended object monitoring device for detecting an unattended object left in a monitoring area based on an image of the monitoring area captured by a camera device, and issuing an alarm when an unattended object is detected. The unattended object monitoring device includes: a processing device configured to perform processing operations required for detecting an unattended object, wherein the processing device is configured to: detect an object present in the monitoring area from a captured image captured by the camera device, and track each detected object between captured images; based on the tracking results for each detected object, detect an object that has not moved for more than a predetermined time as an unattended object; determine whether the detected unattended object exists in a mask area preset in the monitoring area; and exclude the detected unattended object from objects to which the alarm is issued when it is determined that the detected unattended object exists in the mask area.

[0028] In this configuration, an unattended object determined to exist in a preset mask area in the monitoring area can be excluded from objects to which the alarm is issued. This reduces the number of objects to which the alarm is issued, thereby enabling the burden on the observer to be reduced.

[0029] A second aspect of the present invention is the unattended object monitoring device according to the first aspect, wherein the processing device generates a setting screen that enables a user to set the mask area in the monitoring area.

[0030] In this configuration, the user can easily set the mask area through a simple operation.

[0031] A third aspect of the present invention is the unattended object monitoring device according to the first aspect, wherein the processing device determines the object type of the detected unattended object.

[0032] In this configuration, the user can know the type of the detected unattended object.

[0033] A fourth aspect of the present invention is the unattended object monitoring device according to the third aspect, wherein the mask area is set in association with the object type, and wherein the processing device is configured to: when it is determined that the detected unattended object exists in the mask area, further determine whether the object type of the unattended object is the object type associated with the mask area; and exclude the detected unattended object from objects to which the alarm is issued only when it is determined that the object type of the unattended object is the object type associated with the mask area.

[0034] In this configuration, an unattended object can be excluded from the objects to be alerted only when the object type of the unattended object determined to be present in the mask area is the object type associated with the mask area. This configuration can exclude objects of types for which alerts do not need to be issued from the objects to be alerted, thereby further reducing the burden on the observer.

[0035] A fifth aspect of the present invention is the unattended object monitoring device according to the third or fourth aspect, wherein the processing device generates an image in which a frame image surrounding the mask area is superimposed on the captured image, and wherein the processing device determines the form of the frame image according to the object type associated with the mask area.

[0036] In this configuration, the frame image can be a clear visual indication of the mask area, which enables the user to easily grasp the geometry of the mask area in the image. In addition, since the form of the frame image is determined according to the object type of the unattended object, the frame image can also be a clear visual indication of the object type of the unattended object associated with the mask area.

[0037] A sixth aspect of the present invention is the unattended object monitoring device according to the fourth or fifth aspect, wherein the processing device is configured to: detect a representative position of the unattended object whose object type has been determined to be the object type associated with the mask area; store data of the representative position for a predetermined period of time to create statistical data of the representative position; and change the geometry of the mask area based on the statistical data of the representative position.

[0038] In this configuration, the geometry of the mask area can be set more accurately according to the actual position of the unattended object associated with the mask area.

[0039] A seventh aspect of the present invention is the unattended object monitoring device according to any one of the first to sixth aspects, wherein the processing device determines that the mask area is associated with a usage period, which is a period during which the mask area is used, and wherein the processing device determines whether the detected unattended object is present in the mask area only for the mask area having a currently valid usage period.

[0040] In this configuration, the mask area can be used in operation only when an unattended object associated with the mask area is actually used. This reduces the number of judgment operations after the detection of the unattended object, thereby reducing the burden on the device performing the judgment operations.

[0041] The eighth aspect of the present invention is an unattended object monitoring device according to any one of the first to seventh aspects, wherein the processing device detects a representative position of the detected unattended object and determines whether the detected unattended object exists in the mask area based on the detected representative position.

[0042] This configuration enables a more accurate and easier determination of whether a detected unattended object exists in the mask area based on the detected representative position of the detected unattended object.

[0043] The ninth aspect of the present invention is an unattended object monitoring device according to any one of the first to eighth aspects, wherein when the processing device detects a plurality of unattended objects in the mask area, subsequent determination operations are performed on each of the plurality of unattended objects.

[0044] This configuration can ensure that even when two or more unattended objects are detected in a single mask area, the device can appropriately perform subsequent determination operations on each unattended object.

[0045] The tenth aspect of the present invention is an unattended object monitoring system, comprising: an unattended object monitoring device according to any one of the first to ninth aspects; an imaging device for capturing an image of a monitoring area; and a notification device for notifying a user of an alarm issued from the unattended object monitoring device.

[0046] The eleventh aspect of the present invention is an unattended object monitoring method for detecting unattended objects left in a monitoring area based on an image of the monitoring area captured by an imaging device and issuing an alarm when an unattended object is detected. The unattended object monitoring method includes: acquiring a captured image of the monitoring area captured by the imaging device; detecting objects present in the monitoring area from the captured image and tracking each detected object between captured images; based on the tracking results for each detected object, detecting an object that has not moved for more than a predetermined time as an unattended object; determining whether the detected unattended object exists in a mask area preset in the monitoring area; and excluding the detected unattended object from the alarm target objects for which the alarm is issued when it is determined that the detected unattended object exists in the mask area.

[0047] Embodiments of the present invention will be described below with reference to the accompanying drawings.

[0048] (First Embodiment)

[0049] Reference will be made to Figures 1 to 10To describe the unattended object monitoring system 1 according to the first embodiment of the present invention. The unattended object monitoring system 1 is configured to detect an unattended object left in a monitored area based on an image of the monitored area captured by the imaging device 2, and issue an alarm when such an unattended object is detected. The unattended object monitoring system 1 can be used to closely monitor unattended suspicious objects in public spaces used by the public, such as stations, airports, commercial facilities, shopping areas, and underground shopping areas, etc.

[0050] The present invention will be described below with reference to embodiments, in which the unattended object monitoring system 1 according to the present invention is used to closely monitor unattended suspicious objects in public spaces used by the public, such as stations, airports, commercial facilities, shopping areas, and underground shopping areas, etc. In commercial facilities, shopping areas, and underground shopping areas, there are many unattended objects for which no alarm needs to be issued, such as store signs (shop facades) and closed-store cleaning signs on the ground, etc. Therefore, excluding unattended objects for which no alarm needs to be issued from the alarm target objects (hereinafter simply referred to as "(one or more) alarm targets") for which an alarm is issued can reduce the burden on the observer.

[0051] Figure 1 is a diagram showing the schematic structure of the unattended object monitoring system 1 according to the first embodiment of the present invention. As Figure 1 shown, the unattended object monitoring system 1 includes an imaging device 2, an unattended object monitoring device 3, and a display device 4. The display device serves as a notification device for notifying an observer of an alarm. The unattended object monitoring device 3 is connected to the imaging device 2 and the display device 4 via a wired or wireless communication link such as a LAN (local area network) (e.g., Wi-Fi (registered trademark)). The communication link between the unattended object monitoring device 3, the imaging device 2, and the display device 4 is not limited to a specific form, and examples of available communication links can include communication links via a network such as the Internet, or a combination of communication cables and general interfaces (e.g., USB interfaces) forming connections between devices.

[0052] The imaging device 2 is a commonly used imaging device such as a CCD camera, and is installed on any part of a wall, ceiling, pole, or other part in various places such as commercial facilities, shopping areas, and underground shopping areas, so that the imaging device 2 captures a predetermined surveillance target area (hereinafter simply referred to as "surveillance area") as still images and / or moving images in time series. The form, function, arrangement, number, and other characteristics of the imaging device 2 are not limited to specific forms, functions, arrangements, numbers, and other characteristics; that is, the imaging device 2 can have any characteristics and / or various modifications can be made to the imaging device 2, as long as the device can capture images of the surveillance area in time series. For example, the imaging device 2 can be a panoramic camera installed on a wall or any other surface and capable of capturing a 180-degree view with one device, or an omnidirectional camera installed on a ceiling or any other surface and capable of capturing a 360-degree view with one device. The time-series images of the surveillance area captured by the imaging device 2 are provided to the unattended object monitoring device 3.

[0053] The display device 4 is a commonly used display device such as a monitor (display) installed in a surveillance room or any other place where an observer closely monitors unattended objects, and is configured to display a notification image provided from the unattended object monitoring device 3 (see Figure 9 ). Details of the notification image will be described later. The form, function, arrangement, number, and other characteristics of the display device 4 are not limited to specific forms, functions, arrangements, numbers, and other characteristics; that is, the display device 4 can have any characteristics and / or various modifications can be made to the display device 4, as long as the device can display the notification image. For example, the display device 4 can be the display screen of a mobile terminal carried by an observer used when the observer leaves the surveillance room due to patrol or other reasons.

[0054] The unattended object monitoring device 3 is a commonly used computer device with a known structure, and is configured to detect unattended objects left in the surveillance area based on the images captured by the imaging device 2, generate a notification image for the detected unattended object (if any) as an alarm, and provide the generated notification image to the display device 4. Until an unattended object is detected, the unattended object monitoring device 3 continuously receives the captured images at the image acquisition device 13 and provides these captured images as they are to the display device 4.

[0055] Figure 2It is a block diagram showing the schematic structure of the unattended object monitoring device 3. The unattended object monitoring device 3 has a known hardware structure and includes: a processing controller 101 for comprehensively executing the processing operations in the unattended object monitoring system 1 based on a predetermined control program (e.g., an unattended object monitoring program); a RAM 102, i.e., a volatile memory that can be used as the working area of the processing controller 101; a ROM 103, i.e., a non-volatile memory for storing the control program executable by the processing controller 101 and other data; a storage unit 104 for storing the programs and data required for the operation of the unattended object monitoring device 3; an indicator device 105 provided with a monitor for indicating information to the user; an input receiver 106 including input devices such as a keyboard or a mouse for the input operations of the user; and a communication device 107 provided with a communication module configured to communicate via a network.

[0056] In Figure 2 In the shown hardware structure, each function of the unattended object monitoring device 3 can be realized by the processing controller 101 executing the corresponding control program. The unattended object monitoring device 3 is not limited to a computer device and can be any other data processing device (such as a server, etc.) that can provide the same functions as a computer device. In other embodiments, at least some parts of the functions of the unattended object monitoring device can be realized by (one or more than one) operations performed by other known hardware instead of being realized by the unattended object monitoring device 3.

[0057] Return reference Figure 1 , the unattended object monitoring device 3 includes a mask area information controller 11, a mask area information storage unit 12, an image acquisition device 13, an image processing device 14, and a notification image generator 15. The image processing device 14 includes an object tracker 21, an unattended object detector 22, an object type determination unit 23, and an alarm object determination unit 24. Each of the above functional units is controlled by a control device (not shown).

[0058] The mask area information controller 11 generates a mask area information setting screen for inputting information related to each mask area to be preset in the monitoring area (hereinafter referred to as "mask area information"), and provides the required information to the indicator device 105 (see Figure 2 ) to display the mask area information setting screen. The user of the unattended object monitoring device 3 can operate the input device (such as a keyboard or a mouse, etc.) to input various information on the mask area information setting screen and then set the mask area information.

[0059] The mask area information for each mask area includes the name of the mask area, the geometry of the mask area, and the type of unattended object associated with the mask area (hereinafter referred to as the "specified unattended object"). The geometry of the mask area refers to the area in the surveillance area corresponding to the mask area. In this embodiment, although the geometry of the mask area is judged to have a rectangular shape, the geometry of the mask area is not limited to a rectangle and can be any of various other shapes.

[0060] The mask area information for each mask area further includes the form of the mask area frame image indicating the geometry of the mask area, and the usage time period of the mask area.

[0061] In this embodiment, the mask area frame image is a rectangular dotted line image surrounding the mask area and is colored with different colors according to the type of the specified unattended object. The mask area frame image is not limited to a dotted line image and can be an image in any of other forms. For example, the mask area frame image can be a solid line image. The color of the mask area frame image indicating the corresponding type of the specified unattended object is not an essential feature and can be omitted if not needed.

[0062] The usage time period of the mask area is information indicating the time period (e.g., time band and day of the week) during which the mask area is used; that is, when the usage time period is valid, the mask area is used. As described above, the mask area is set in association with the specified unattended object. Therefore, the mask area can be set in association with the specified unattended object and further with its usage time period. In this case, only when the specified unattended object is actually used (placed), the mask area corresponding to the specified unattended object can be used. This configuration can reduce the number of judgment operations (described later) performed by the alarm object judgment unit 24, thereby enabling the burden on the device performing the judgment operations to be alleviated.

[0063] For example, when the specified unattended object is a store sign, the store sign is only used during the business hours and business days of the corresponding store, so its mask area only needs to be used during the time band and day of the week when the store sign is actually used. In other cases, when the specified unattended object is a construction safety cone, the construction safety cone is only used when the corresponding construction work is being performed, so its mask area only needs to be used during the time band and day of the week in the building time period when the construction safety cone is actually used.

[0064] In addition, in addition to including the above information records, the mask area information can further include various other types of information. For example, the mask area information can further include image data of the specified unattended object of the mask area.

[0065] Figure 3 FIG. Figure 3 is a diagram showing an example of a mask area information setting screen 31 generated by a mask area information controller 11. The mask area information setting screen 31 includes a mask area setting screen 32 (the left screen portion) and a designated unattended object setting screen 33 (the right screen portion). Below the designated unattended object setting screen 33, there is a screen switching button 34 for switching the designated unattended object setting screen 33 to a usage time period setting screen 36 described later.

[0066] The mask area setting screen 32 displays an image of a monitoring area captured by the imaging device 2. In Figure 3 the example, the mask area setting screen 32 displays a captured image of a predetermined monitoring area in a shopping area.

[0067] The designated unattended object setting screen 33 includes check boxes for designating, for each mask area, a designated unattended object associated with the mask area, and the color of a mask area frame image corresponding to the designated unattended object. In Figure 3 the example, the listed designated unattended objects include store signs, billboards, closed store cleaning signs, and construction safety cones. The listed colors of the mask area frame images include red for store signs, blue for billboards, yellow for closed store cleaning signs, and green for construction safety cones.

[0068] When a user of the unattended object monitoring device 3 inputs various information records in the mask area information setting screen 31 to set mask area information, first, the user needs to click a check box in the designated unattended object setting screen 33 to select a designated unattended object. In Figure 3 the example, a store sign is selected. Then, the user uses a mouse or any other input device to specify a desired area in the mask area setting screen 32, thereby setting the geometry of the mask area.

[0069] In the case where the designated unattended object is a store sign, the position of the store sign can change every day. In addition, even after the store sign is placed, the position of the store sign may change due to a collision with a passerby or other reasons. From this perspective, the geometry of the mask area is preferably wider than the area where the store sign is expected to be placed, so that even if the position of the store sign changes, the mask area can still be used.

[0070] In Figure 3In the example, the mask area A1 is set on the lower right side of the mask area setting screen 32. The mask area A1 is indicated by a mask area frame image 35 that surrounds the mask area. A store sign is designated as the unattended object specified for the mask area A1, and red is associated with the store sign. Accordingly, the mask area frame image 35 is colored red. Thus, in Figure 3 the example, the mask area A1 is set in association with the store sign.

[0071] In Figure 3 the example, after the unattended object specified is set, the geometry of the mask area is set. However, the order of setting the unattended object specified and the geometry of the mask area is not limited to this order. The user can set the geometry of the mask area before setting the unattended object specified for the mask area. In this case, when the geometry of the mask area is set, the mask area frame image 35 can be indicated in black, and then when the unattended object specified is set, the color of the mask area frame image 35 can be changed to the color associated with the unattended object specified.

[0072] Figure 4 A usage period setting screen 36 for setting the usage period of the mask area is shown. As described above, the user can operate the screen switching button 34 to switch the unattended object specified setting screen 33 to the usage period setting screen 36.

[0073] The usage period setting screen 36 includes an input field for specifying the usage period for using the mask area. In the present embodiment, the time band and the day of the week are input to set the usage period. However, the unit of the usage period is not limited to this example, and the usage period can be set in units of days, weeks, months, or years. In addition, the user can specify the start date and time and the end date and time, thereby setting the usage period as a continuous period between the two. In the case where the usage period is periodically effective (i.e., a periodic usage period), it is preferable to set the usage period such that the usage period is effective multiple times (each time at the same time band or day of the week) within a predetermined period.

[0074] In Figure 4 the example, the time band when the mask area is used (that is, the mask area function is effective) is set to "8:00 to 22:00", and the time band when the mask area is not used (that is, the mask area function is ineffective) is set to "22:00 to 8:00". This time band is selected because the store using the store sign is open from 8:00 to 22:00 and does not use the store sign at other times. In Figure 4In the example, the days of the week for which the mask area is used (the days of the week when the mask area function is effective) are set to "Monday to Friday", and the days of the week for which the mask area is not used (the days of the week when the mask area function is ineffective) are "Saturday and Sunday". The days of the week are selected because the business days of the store using the store sign are from Monday to Friday, and the store sign is not used on Saturday and Sunday.

[0075] The mask area information set by the mask area information controller 11 is stored in the mask area information storage unit 12. In the present embodiment, the geometry of the mask area is defined by the upper left coordinate and the lower right coordinate of the rectangular frame of the mask area frame image indicated on the mask area setting screen 32.

[0076] Figure 5 is a diagram showing an example of the mask area information stored in the mask area information storage unit 12. As Figure 5 shown, the mask area information for each mask area includes "mask area name" (the name of the mask area), "designated unattended object" (the designated unattended object in the mask area), "coordinates of the frame image" (the coordinates of the mask area frame image), "color of the frame image" (the color of the mask area frame image), and "usage time period" (the usage time period of the mask area). In Figure 5 the example, the information of mask area A1 in the "mask area name" is associated with the following information records, which include the designated unattended object "store sign", the coordinates of the frame image "(x800, y300), (x900, y200)", the frame coordinate color "red", and the usage time period "time zone: 8:00 to 22:00, days of the week: Monday to Friday".

[0077] The image acquisition device 13 is connected to the imaging device 2 and is configured to acquire an image of the surveillance area captured by the imaging device 2. The image acquisition device 13 is also connected to the image processing device 14 and the notification image generator 15, and is configured to input the captured image acquired from the imaging device 2 to the image processing device 14 and the notification image generator 15.

[0078] The image processing device 14 acquires a captured image from the image acquisition device 13. When the captured image is a still image captured at a predetermined time interval, the image processing device 14 acquires these images as still image data. When the captured image is a moving image, the image processing device 14 acquires these images as a set of still image data (image frame data) extracted from the moving image at a predetermined time interval. The captured image (still image data) acquired by the image processing device 14 from the image acquisition device 13 undergoes processing operations by the object tracker 21, the unattended object detector 22, the object type determination unit 23, and the alarm target determination unit 24.

[0079] The object tracker 21 detects one or more objects (hereinafter referred to as "appearing objects") that are present in the monitoring area from the captured image, and tracks each detected appearing object between consecutive captured images after the detection of the appearing object. Examples of appearing objects include paper bags, store signs, closed shop cleaning signs, and bags. The object tracker 21 detects the appearing object by comparing a pre-captured background image with the acquired captured image, or by comparing between adjacent pairs of captured images in chronological order. Specifically, the object tracker 21 calculates the degree of difference (difference or correlation value) of pixel values (e.g., luminance values) between the captured images to be compared, and then extracts the pixels for which the calculated difference exceeds a predetermined threshold. Then, the object tracker 21 determines the appearing object as a region formed by grouping the extracted adjacent pixels together into one region. When one or more appearing objects are detected from the captured image, the object tracker 21 assigns a unique ID to each detected appearing object, and tracks the appearing object between the captured images based on the corresponding ID.

[0080] The unattended object detector 22 detects an unattended object based on the tracking result for each appearing object provided by the object tracker 21. Specifically, the unattended object detector 22 determines the movement of each appearing object detected by the object tracker 21 over time in the captured image, and detects an appearing object that has not shown any movement within a predetermined time (that is, an appearing object that has been stationary within a predetermined time) as an unattended object. When an unattended object is detected, the unattended object detector 22 assigns a unique ID to each detected unattended object, and tracks the unattended object between consecutive captured images based on the corresponding ID. In addition, in the case where an unattended object is detected, the unattended object detector 22 inputs the detection result to the notification image generator 15.

[0081] The object type determination unit 23 determines the type (object type) of the unattended object detected by the unattended object detector 22. The object type determination unit 23 can determine the object type of the unattended object by using known image recognition techniques such as a classifier using deep learning techniques. For example, the object type determination unit 23 preferably uses a classifier configured to identify the object type of the unattended object based on the shape features and texture features of the unattended object detected from the captured image. In this case, the object type determination unit 23 uses a database containing the shape features and texture features of the unattended object corresponding to each type of the unattended object, which is prepared in advance. The method used by the object type determination unit 23 to identify the object type of the unattended object is not limited to the above method, and can be any method among known appropriate methods. The object type determination unit 23 provides the determined object type of the unattended object to the notification image generator 15.

[0082] The alarm target determination unit 24 determines whether the unattended object detected by the unattended object detector 22 is an alarm target, that is, an object for which an alarm needs to be issued. In the case where the unattended object detector 22 detects a plurality of unattended objects, the alarm target determination unit 24 performs an operation (alarm target determination operation) for determining each unattended object (that is, for each object).

[0083] The alarm target determination unit 24 performs the alarm target determination operation by referring to the mask area information stored in the mask area information storage unit 12. As described above, the mask area information includes, for each mask area, the geometric shape of the mask area (coordinates of the mask area frame image) and the specified unattended object(s) of the mask area. In the alarm target determination operation, the alarm target determination unit 24 only uses the mask areas having a currently valid usage period. In the case where there are a plurality of mask areas in the monitoring area, the alarm target determination unit 24 performs the alarm target determination operation for each mask area.

[0084] Specifically, the alarm target determination unit 24 first determines whether the unattended object detected by the unattended object detector 22 exists in the mask area by using known appropriate techniques such as known image analysis techniques. In the case where it is determined that the detected unattended object does not exist in the mask area, the alarm target determination unit 24 determines that the unattended object is an alarm target.

[0085] In the case where it is determined that the unattended object detected exists in the mask area, the alarm target determination unit 24 further determines whether the unattended object is the designated unattended object in the mask area. In the case where it is determined that the object type of the unattended object is the designated unattended object, the alarm target determination unit 24 determines that the unattended object is an unattended object for which no alarm needs to be issued (hereinafter referred to as a "non-alarm target"). In the case where it is determined that the object type of the unattended object is not the designated unattended object in the mask area, the alarm target determination unit 24 determines that the unattended object is an alarm target.

[0086] In other words, the alarm target determination unit 24 determines only the (one or more) unattended objects that exist in the mask area and are also the designated unattended objects as non-alarm targets, and otherwise determines the unattended object as an alarm target.

[0087] Figure 6 is a diagram showing an example of the alarm target determination operation performed by the alarm target determination unit 24. In Figure 6 In the example shown in (a), a mask area A1 is preset on the lower right side of the monitoring area, and the designated unattended object in the mask area A1 is a store sign. Since the store sign is associated with red, the mask area frame image 35 is colored red.

[0088] In addition, as Figure 6 shown in (b), a paper bag 41 and a store sign 42 are detected in the monitoring area. In this case, the alarm target determination unit 24 determines whether the paper bag 41 and the store sign 42 exist in the mask area A1. As described above, the alarm target determination operation is performed for each unattended object. In Figure 6 the example of (b), the alarm target determination unit 24 determines that the paper bag 41 does not exist in the mask area A1 and that the store sign 42 exists in the mask area A1. Therefore, the alarm target determination unit 24 determines the paper bag 41 as an alarm target.

[0089] Next, the alarm target determination unit 24 further determines whether the store sign 42 in the mask area A1 is the designated unattended object in the mask area A1. Since the designated unattended object in the mask area A1 is a store sign, the alarm target determination unit 24 determines that the store sign 42 is the designated unattended object. Therefore, the alarm target determination unit 24 determines that the store sign 42 is a non-alarm target (see Figure 6 (c)).

[0090] In this way, the alarm target determination unit 24 determines only the (one or more) unattended objects that exist in the mask area and are also the designated unattended objects as non-alarm targets, and otherwise determines the unattended object as an alarm target.

[0091] In some cases, the unattended object detector 22 detects two or more unattended objects in a mask area. Figure 7 is a diagram showing an example of the alarm object determination operation in these cases. As Figure 7 shown in (a) of, a mask area A2 is preset at the lower center of the monitoring area. The designated unattended object in the mask area A2 is a closed store cleaning sign. Since the closed store cleaning sign is associated with yellow, the mask area frame image 35 is colored yellow.

[0092] In addition, as Figure 7 shown in (b) of, a closed store cleaning sign 43 and a bag 44 are detected in the monitoring area. In this case, the alarm object determination unit 24 determines whether the closed store cleaning sign 43 and the bag 44 each exist in the mask area A2. As described above, the alarm object determination operation is performed for each unattended object. In Figure 7 the example of (b) of, the alarm object determination unit 24 determines that the closed store cleaning sign 43 and the bag 44 each exist in the mask area A2.

[0093] Next, the alarm object determination unit 24 further determines whether the closed store cleaning sign 43 and the bag 44 in the mask area A2 are each the designated unattended object in the mask area A2. Since the designated unattended object in the mask area a2 is the closed store cleaning sign, the alarm object determination unit 24 determines that the closed store cleaning sign 43 is the designated unattended object. Therefore, the alarm object determination unit 24 determines that the closed store cleaning sign 43 is a non-alarm object. The alarm object determination unit 24 determines that the bag 44 is not the designated unattended object. Therefore, the alarm object determination unit 24 determines that the bag 44 is an alarm object (see Figure 7 (c) of).

[0094] In this way, even when there are two or more unattended objects in a mask area, the alarm object determination unit 24 performs the alarm object determination operation for each unattended object and determines whether the unattended object is an alarm object or a non-alarm object.

[0095] In some cases, the unattended object detector 22 detects two or more unattended objects that appear in an overlapping manner in a mask area. Since it is rare for two or more unattended objects to appear simultaneously, the alarm object determination unit 24 makes determinations regarding whether the unattended object is the designated unattended object and whether the unattended object is an alarm object or a non-alarm object in the order of appearance of the two unattended objects. In this case, as Figure 7 in the example of, the determination operation is performed for each unattended object. First, as Figure 8As shown in (a), a mask area A1 is preset at the lower right side of the monitoring area, and the designated unattended object in the mask area A1 is a store sign. Since the store sign is associated with red, the mask area frame image 35 is colored red.

[0096] In addition, as Figure 8 shown in (b), a store sign 42 and a bag 44 are detected in the monitoring area. In this case, the alarm target determination unit 24 determines whether the store sign 42 and the bag 44 are each present in the mask area A1. As described above, the alarm target determination operation is performed on each unattended object. In Figure 8 the example of (b), the alarm target determination unit 24 determines that the store sign 42 and the bag 44 are each present in the mask area A1.

[0097] Next, the alarm target determination unit 24 further determines whether the store sign 42 and the bag 44 in the mask area A1 are each the designated unattended object in the mask area A1. Since the designated unattended object in the mask area A1 is a store sign, the alarm target determination unit 24 determines that the store sign 42 is the designated unattended object. Therefore, the alarm target determination unit 24 determines that the store sign 42 is a non-alarm target. The alarm target determination unit 24 determines that the bag 44 is not the designated unattended object. The alarm target determination unit 24 determines that the bag 44 is an alarm target (see Figure 8 (c)).

[0098] As described above, since the unattended object detector 22 detects unattended objects discriminately and assigns unique IDs to each detected unattended object, even when two or more unattended objects are detected in one mask area and these unattended objects appear in an overlapping manner, each detected unattended object can be identified. Therefore, even when two or more unattended objects are detected and these unattended objects appear in an overlapping manner, the alarm target determination unit 24 can make a determination regarding each unattended object as to whether the unattended object is the designated unattended object and whether the unattended object is an alarm target or a non-alarm target.

[0099] The alarm target determination unit 24 provides the determination result to the notification image generator 15.

[0100] The notification image generator 15 generates a notification image as an alert for the alert object based on the determination result from the alert object determination unit 24. Non-alert objects are excluded from the alert object objects for which the alert is issued. Specifically, the notification image generator 15 generates a notification image for issuing an alert as an alert box image surrounding the alert object, an indication of the object type, and a mask area box image indicating the mask area, which are superimposed on the captured image provided by the image acquisition device 13. The mask area box image superimposed on the notification image is not essential. The notification image generator 15 may be configured such that the user can select to show or hide the mask area box image superimposed on the notification image.

[0101] The alert box image 51 is preferably formed to have a distinct color such as red or yellow. In some cases, the alert box image 51 can be made to flash, and / or characters or pictures can be displayed adjacent to the alert box image to enhance the alert. In the present embodiment, characters indicating the object type are used to notify the user of the type of the unattended object. The indication form of the type of the unattended object is not limited to these characters and can be other forms of indication such as pictures or symbols indicating the object type, etc.

[0102] Figure 9 (a) to (c) of FIG. are diagrams showing examples of the notification image generated by the notification image generator 15.

[0103] Figure 9 (a) of FIG. shows an example corresponding to the Figure 6 alert object determination operation shown in FIG. In the example of (a) of FIG., a paper bag 41 and a store sign 42 are detected as unattended objects, and the paper bag 41 is determined as an alert object, while the store sign 42 is determined as a non-alert object. Therefore, the notification image generator 15 generates a notification image P1 as a notification box image 51 surrounding the paper bag 41 as the alert object, which is superimposed on the captured image. In the example of (a) of FIG., since the paper bag 41 is the unattended object to be notified, the character 52 "paper bag" is further superimposed on the captured image as information indicating the object type. Additionally, in the example of (a) of FIG., a mask area box image 35 showing the mask area A1 is further superimposed on the captured image. Since the store sign is associated with red, the mask area box image 35 is colored red. Figure 9 (a) of FIG. shows an example corresponding to the Figure 9 alert object determination operation shown in FIG. In the example of (a) of FIG., a paper bag 41 and a store sign 42 are detected as unattended objects, and the paper bag 41 is determined as an alert object, while the store sign 42 is determined as a non-alert object. Therefore, the notification image generator 15 generates a notification image P1 as a notification box image 51 surrounding the paper bag 41 as the alert object, which is superimposed on the captured image. In the example of (a) of FIG., since the paper bag 41 is the unattended object to be notified, the character 52 "paper bag" is further superimposed on the captured image as information indicating the object type. Additionally, in the example of (a) of FIG., a mask area box image 35 showing the mask area A1 is further superimposed on the captured image. Since the store sign is associated with red, the mask area box image 35 is colored red. Figure 9 (a) of FIG. shows an example corresponding to the

[0104] Figure 9 (b) of FIG. shows an example corresponding to the Figure 7 alert object determination operation shown in FIG. In the example of (b) of FIG., Figure 9In the example of (b), the closing cleaning sign 43 and the bag 44 are detected as unattended objects, and the closing cleaning sign 43 is judged as a non-alarm object, while the bag 44 is judged as an alarm object. Therefore, the notification image generator 15 generates a notification image P2 as a notification box image 51 surrounding the bag 44 that is superimposed on the captured image. In Figure 9 In the example of (b), since the bag 44 is an unattended object to be notified, the character 52 "(bag)" is further superimposed on the captured image as information indicating the object type. Additionally, in Figure 9 In the example of (b), a mask area frame image 35 showing the mask area A2 is further superimposed on the captured image. Since the closing cleaning sign is associated with yellow, the mask area frame image 35 is colored yellow.

[0105] Figure 9 (c) is a diagram showing an example corresponding to the Figure 6 alarm object determination operation shown. In Figure 9 In the example of (c), the store sign 42 and the bag 44 are detected as unattended objects, and the store sign 42 is judged as a non-alarm object, while the bag 44 is judged as an alarm object. Therefore, the notification image generator 15 generates a notification image P3 as a notification box image 51 surrounding the bag 44 that is superimposed on the captured image. In Figure 9 In the example of (c), since the bag 44 is an unattended object to be notified, the character 52 "(bag)" is further superimposed on the captured image as information indicating the object type. Additionally, in Figure 9 In the example of (c), a mask area frame image 35 showing the mask area A1 is further superimposed on the captured image. Since the store sign is associated with red, the mask area frame image 35 is colored red.

[0106] The notification image generator 15 provides the generated notification image to the display device 4. In addition, the notification image generator 15 acquires the captured image from the image acquisition device 13 and provides these captured images as they are to the display device 4 until the unattended object detector 22 detects an unattended object.

[0107] When the notification image is displayed on the display device, the notification image generator 15 can be configured to output an alarm voice by using the voice output function of the display device or cause an alarm light to flash by using an alarm device associated with the display device to provide an additional alarm. This feature can provide a stronger alarm to warn the observer.

[0108] Next, the following description of the flowchart with reference to Figure 10 is about the Figure 1The operation process of the unattended object monitoring device 3 of the unattended object monitoring system 1 according to the first embodiment as shown.

[0109] First, the image acquisition device 13 acquires the images captured by the imaging device 2 and provides these captured images to the image processing device 14 and the notification image generator 15 (step ST101). Subsequently, the object tracker 21 detects the appearing objects present in the monitoring area from the captured images (step ST102). When an appearing object is detected (Yes in step ST102), the process proceeds to step ST103. When no appearing object is detected (No in step ST102), the process returns to step ST101.

[0110] In step ST103, after detecting the appearing objects, the object tracker 21 starts tracking each detected appearing object between the captured images. In step ST104, the unattended object detector 22 detects the appearing objects that have not shown any movement within a predetermined time as unattended objects based on the tracking results of each appearing object from the object tracker 21. When an unattended object is detected (Yes in step ST104), the process proceeds to step ST105. When no unattended object is detected (No in step ST104), the process returns to step ST101. When returning to step ST101, in the next step ST102, the object tracker 21 detects new appearing objects (that is, appearing objects other than the already detected (one or more) appearing objects).

[0111] In step ST105, the object type determination unit 23 determines the object type of the unattended objects detected by the unattended object detector 22. As described above, the object type determination unit 23 performs this object type determination operation on each unattended object. The same applies to other subsequent determination operations performed in the image processing device.

[0112] Next, in step ST106, the alarm target determination unit 24 determines whether the unattended objects detected by the unattended object detector 22 exist in a preset mask area in the monitoring area (see Figure 6 (b) of Figure 7 (b) of Figure 8 (b) of

[0113] When the unattended object determination unit 24 determines that the unattended object detected by the unattended object detector 22 exists in the mask area (Yes in step ST106), the process proceeds to step ST107. When the unattended object determination unit 24 determines that the detected unattended object does not exist in the mask area (No in step ST106), the process proceeds to step ST109.

[0114] In step ST107, the unattended object determination unit 24 further determines whether the object type of the unattended object is the specified unattended object in the mask area (see Figure 6 (c) of Figure 7 (c) of Figure 8 (c) of). When the unattended object determination unit 24 determines that the object type of the unattended object is the specified unattended object in the mask area (Yes in step ST107), the process proceeds to step ST108. When the unattended object determination unit 24 determines that the object type of the unattended object is not the specified unattended object in the mask area (No in step ST107), the process proceeds to step ST109.

[0115] In step ST108, the unattended object determination unit 24 determines that the unattended object is a non-alarm object, and then ends the alarm object determination operation.

[0116] In step ST109, the unattended object determination unit 24 determines that the unattended object is an alarm object, and then ends the alarm object determination operation.

[0117] As described above, according to the first embodiment, an unattended object determined to exist in a preset mask area in the monitoring area can be excluded from the alarm target objects for which an alarm is issued. This reduces the number of alarm target objects for which an alarm is issued, thereby enabling the burden on the observer to be reduced. In addition, according to the first embodiment, an unattended object can be excluded from the alarm target objects for which an alarm is issued only when the object type of the unattended object determined to exist in the mask area is the object type associated with the mask area. This configuration can exclude objects of types for which an alarm does not need to be issued from the alarm target objects for which an alarm is issued, thereby enabling the burden on the observer to be further reduced.

[0118] (Second Embodiment)

[0119] Next, the unattended object monitoring system 1 according to the second embodiment of the present invention will be described with reference to Figures 11 to 13 The second embodiment is the same as Figure 1The difference in the first embodiment shown is that the image processing device 14 of the unattended object monitoring device 3 further includes a representative position detector 25. Except for using the representative position detector, the structure of the second embodiment is the same as that of the first embodiment. Therefore, the same features as those in the first embodiment are denoted by the same reference numerals, and their descriptions will not be repeated.

[0120] When the unattended object detector 22 detects an unattended object, the representative position detector 25 detects the representative position of the detected unattended object by using known image analysis techniques. Examples of the representative position of an unattended object include the centroid of the image of the unattended object or the position of the lower end center.

[0121] Figure 12 is a diagram showing examples of the representative positions of unattended objects. Figure 12 (a) of shows an example of using the centroid of an object as the representative position. Figure 12 (b) of shows an example of using the lower end center of an object as the representative position. In Figure 12 In the example of (a), the store sign 42 is detected as an unattended object, and the representative position detector 25 detects the position 61, which is the centroid of the store sign 42, as its representative position. In Figure 12 In the example of (b), the closed store cleaning sign 43 is detected as an unattended object, and the representative position detector 25 detects the position 61, which is the lower end center of the closed store cleaning sign 43, as its representative position.

[0122] In the second embodiment, the alarm target determination unit 24 determines whether an unattended object exists in the mask area based on the representative position of the unattended object detected by the representative position detector 25. Specifically, when the representative position of the unattended object exists in the mask area, the alarm target determination unit 24 determines that the unattended object exists in the mask area. In Figure 12 In the example of (a), although the right side of the store sign 42 protrudes from a part of the mask area A1, the position 61 of the centroid of the store sign 42 exists in the mask area A1. Therefore, the alarm target determination unit 24 determines that the store sign 42 exists in the mask area A1. In Figure 12 In the example of (b), although the left side of the closed store cleaning sign 43 protrudes from a part of the mask area A2, the position 62 of the lower end center of the closed store cleaning sign exists in the mask area A2. Therefore, the alarm target determination unit 24 determines that the closed store cleaning sign 43 exists in the mask area A2.

[0123] In this way, the unattended object determination unit 24 determines whether the unattended object exists in the mask area based on the representative position of the unattended object (such as the position of the center of gravity or the center of the lower end, etc.), thereby enabling a more accurate and easier determination of whether the unattended object is in the mask area.

[0124] The position that can be the representative position of the unattended object is not limited to the center of gravity or the center of the lower end of the object, and can be various other positions, such as the center of the upper end, the upper left end, the upper right end, the lower left end, and the lower right end of the object, etc. However, the lower end part or any other part of the unattended object is sometimes hidden behind passers-by or other unattended objects, which prevents the representative position detector 25 from detecting the position of the center of gravity or the center of the lower end of the object. In this case, the position of the center of the upper end, the upper left end, the upper right end, or any other point can be used as the representative position of the unattended object.

[0125] In the second embodiment, the image processing device 14 can change the geometry of the mask area based on the statistical data of the representative positions of the specified unattended objects in the mask area. For example, signs such as the store sign 42 and the closed store cleaning sign 43 can be placed at different positions every day. Even after placing such signs, the position of the signs may change due to collisions with passers-by or other reasons. Therefore, it is beneficial to change the geometry of the mask area according to the actual positions of the specified unattended objects.

[0126] Specifically, the image processing device 14 first accumulates the position data of the representative positions of the specified unattended objects detected by the unattended object detector 22 for a predetermined period of time, and creates statistical data of the representative positions of the specified unattended objects. Then, the image processing device 14 uses known statistical analysis techniques to analyze the created statistical data, and changes the geometry of the mask area according to the actual positions of the specified unattended objects.

[0127] In the second embodiment, the image processing device 14 changes the geometry of the mask area by moving the position of the mask area. Specifically, the image processing device 14 can move the position of the mask area so that the representative position of the specified unattended object is within a predetermined distance from the center position of the geometry of the mask area. In this embodiment, the geometry of the mask area is defined by the upper left coordinates and the lower right coordinates of the rectangular frame of the mask area frame image. Therefore, the image processing device 14 determines the center position (coordinates) of the geometry of the mask area based on the upper left coordinates and the lower right coordinates of the mask area frame image. The predetermined distance from the center position of the geometry of the mask area is preferably determined such that the entire specified unattended object can be included within the geometry of the mask area.

[0128] Figure 13 This is a diagram showing an example of changing the geometry of a mask area. Figure 13 (a) of this shows an example of changing the geometry of mask area A1, where A1-1 represents mask area A1 before the geometry change, and A1-2 represents mask area A1 after the geometry change. As shown in this diagram, after the geometry change, the position of mask area A1 moves to the right. Figure 13 (b) of this shows an example of changing the geometry of mask area A2, where A2-1 represents mask area A2 before the geometry change, and A2-2 represents mask area A2 after the geometry change. As shown in this diagram, after the geometry change, the position of mask area A2 moves to the left.

[0129] In Figure 13 the example, the geometry of the mask area is changed by moving the position of the mask area. However, the geometry change is not limited to the change in the position of the mask area. For example, the geometry of the mask area can be changed by changing the size of the mask area. In this case, the geometry of the mask area can be changed by increasing or decreasing the size of the mask area based on the statistical data of the representative position of the specified unattended object. In other cases, the geometry of the mask area can be changed by changing the shape of the mask area. In this case, the geometry of the mask area can be changed by changing the shape of the mask area from the initial rectangular shape to any different shape based on the statistical data of the representative position of the specified unattended object.

[0130] Thus, since the geometry of the mask area is changed based on the statistical data of the representative position of the specified unattended object, the image processing device 14 can more accurately set the geometry of the mask area according to the actual position of the unattended object.

[0131] Specific embodiments of the present invention are described herein for illustrative purposes. However, the present invention is not limited to these specific embodiments. In the embodiments of the unattended object monitoring device, the unattended object monitoring system equipped with the unattended object monitoring device, and the unattended object monitoring method as described above, not all elements are necessary. Therefore, various modifications (including deleting some elements) can be appropriately made to the embodiments without departing from the scope of the present invention.

[0132] Industrial Applicability

[0133] An unattended object monitoring device according to the present invention, an unattended object monitoring system equipped with the unattended object monitoring device, and an unattended object monitoring method are useful as follows. The unattended object monitoring device, the unattended object monitoring system equipped with the unattended object monitoring device, and the unattended object monitoring method can exclude detected unattended objects that do not require an alarm from the alarm target objects for which the alarm is issued.

[0134] Explanation of Reference Numerals

[0135] 1 Unattended object monitoring system

[0136] 2 Imaging device

[0137] 3 Unattended object monitoring device

[0138] 4 Display device (notification device)

[0139] 11 Mask area information controller

[0140] 12 Mask area information storage unit

[0141] 13 Image acquisition device

[0142] 14 Image processing device

[0143] 15 Notification image generator

[0144] 21 Object tracker

[0145] 22 Unattended object detector

[0146] 23 Object type determination unit

[0147] 24 Alarm target determination unit

[0148] 25 Representative position detector

Claims

1. An unattended object monitoring device for detecting an unattended object left in a monitoring area based on an image of the monitoring area captured by a camera device and issuing an alarm when an unattended object is detected. The unattended object monitoring device includes: A processing device for performing processing operations required to detect an unattended object. Wherein, the processing device is configured to: Detect objects present in the monitoring area from the captured images captured by the camera device and track each detected object between the captured images; Based on the tracking results of the detected objects, detect an object that has not moved for more than a predetermined time as an unattended object; Judge the object type of the detected unattended object; Judge whether the detected unattended object exists in a mask area preset in the monitoring area, where the mask area is set in association with the object type; In the case where it is judged that the detected unattended object exists in the mask area, further judge whether the object type of the unattended object is the object type associated with the mask area; and Only in the case where it is judged that the object type of the unattended object is the object type associated with the mask area, exclude the detected unattended object from the alarm target objects for which the alarm is issued.

2. The unattended object monitoring device according to claim 1, wherein, The processing device generates a setting screen that enables a user to set the mask area in the monitoring area.

3. The unattended object monitoring device according to claim 1, wherein, The processing device generates an image in which a frame image surrounding the mask area is superimposed on the captured image, and Wherein, the processing device determines the form of the frame image according to the object type associated with the mask area.

4. The unattended object monitoring device according to claim 1, wherein, The processing device is configured to: Detect the representative position of the unattended object whose object type has been judged to be the object type associated with the mask area; Store the data of the representative position and continue for a predetermined time period to create statistical data of the representative position; And Change the geometric shape of the mask area based on the statistical data of the representative position.

5. The unattended object monitoring device according to claim 1, wherein, The processing device determines that the mask area is associated with a usage period, which is the period during which the mask area is used, and Wherein, the processing device judges whether the detected unattended object exists in the mask area only for the mask area with a currently valid usage period.

6. The unattended object monitoring device according to claim 1, wherein, The processing device detects the representative position of the detected unattended object and judges whether the detected unattended object exists in the mask area based on the detected representative position.

7. The unattended object monitoring device according to claim 1, wherein, When multiple unattended objects are detected in the mask area, the processing device performs subsequent judgment operations on each of the multiple unattended objects.

8. An unattended object monitoring system, including: The unattended object monitoring device according to any one of claims 1 to 7; A camera device for capturing an image of a monitoring area; And A notification device for notifying a user of an alarm issued from the unattended object monitoring device.

9. An unattended object monitoring method for detecting an unattended object left in a monitoring area based on an image of the monitoring area captured by a camera device and issuing an alarm when an unattended object is detected, the unattended object monitoring method comprising: Obtaining a captured image of the monitoring area captured by the camera device; Detecting objects present in the monitoring area from the captured image and tracking each detected object between the captured images; Based on the tracking results for each detected object, detecting an object that has not moved for more than a predetermined time as an unattended object; Judging the object type of the detected unattended object; Judging whether the detected unattended object exists in a mask area preset in the monitoring area, wherein the mask area is set in association with the object type; In the case where it is judged that the detected unattended object exists in the mask area, further judging whether the object type of the unattended object is an object type associated with the mask area; And Excluding the detected unattended object from the alarm target objects for which the alarm is issued only when it is judged that the object type of the unattended object is an object type associated with the mask area.

Images