Estimation device, estimation method, and program

Abstract

The present invention provides an estimation device comprising: a person detection unit that detects a person to be tracked from within an image generated by a plurality of cameras installed at prescribed positions; a camera identification unit that identifies a camera for which the timing at which the person to be tracked is detected in the image satisfies a prescribed condition that is based on a target time; and an estimation unit that, on the basis of the installation position of the identified camera, estimates an area in which the person to be tracked is present at the target time.

Description

Estimation device, estimation method, and recording medium

[0001] The present invention relates to an estimation device, an estimation method, and a program.

[0002] A technique related to the present invention is disclosed in Patent Document 1. The technique disclosed in Patent Document 1 discloses a technique for tracking a person as a tracking target based on images generated by a plurality of cameras.

[0003] International Publication No. 2020 / 115890

[0004] By using a tracking technique such as that disclosed in Patent Literature 1, it is possible to identify the location of a person to be tracked at a target time. However, with this tracking technique, if the person to be tracked is not captured in an image captured at the target time, it is not possible to identify the location of the person to be tracked at the target time.

[0005] In view of the above-mentioned problems, an example of an object of the present invention is to provide an estimation device, an estimation method, and a program for estimating an area in which a person to be tracked is present at a target time.

[0006] According to one aspect of the present invention, there is provided an estimation device having: a person detection means for detecting a person to be tracked from images generated by a plurality of cameras installed at predetermined positions; a camera identification means for identifying the camera at which the timing at which the person to be tracked is detected in the image satisfies a predetermined condition based on a target time; and an estimation means for estimating an area in which the person to be tracked is present at the target time based on the installation positions of the identified cameras.

[0007] According to one aspect of the present invention, there is provided an estimation method in which one or more computers detect a person to be tracked from images generated by multiple cameras installed at predetermined locations, identify the cameras at which the timing at which the person to be tracked was detected in the images satisfies predetermined conditions based on a target time, and estimate an area in which the person to be tracked is located at the target time based on the installation locations of the identified cameras.

[0008] According to one aspect of the present invention, a program is provided that causes a computer to function as: person detection means that detects a person to be tracked from images generated by multiple cameras installed at predetermined locations; camera identification means that identifies the camera at which the timing at which the person to be tracked is detected in the image satisfies predetermined conditions based on a target time; and estimation means that estimates the area in which the person to be tracked is located at the target time based on the installation locations of the identified cameras.

[0009] According to one aspect of the present invention, an estimation device, an estimation method, and a program are realized that estimate an area in which a person to be tracked is present at a target time.

[0010] The above-mentioned objects and other objects, features and advantages will become more apparent from the following description of the preferred embodiments and the accompanying drawings.

[0011] FIG. 1 is a diagram showing an example of a functional block diagram of an estimation device. FIG. 2 is a diagram showing an example of a hardware configuration of an estimation device. FIG. 3 is a diagram for explaining an example of a process in which an estimation device estimates a predetermined area. FIG. 4 is a diagram for explaining another example of a process in which an estimation device estimates a predetermined area. FIG. 5 is a flowchart showing an example of a process flow of an estimation device. FIG. 6 is a diagram for explaining an example of a process in which an estimation device estimates a visit target of a tracking target person. FIG. 7 is a flowchart showing an example of a process flow of an estimation device. FIG. 8 is a diagram showing an example of information output by the estimation device.

[0012] Hereinafter, embodiments of the present invention will be described with reference to the drawings. In all the drawings, like components are designated by like reference numerals, and the description thereof will be omitted as appropriate.

[0013] 1 is a functional block diagram showing an overview of an estimation device 10 according to a first embodiment. The estimation device 10 includes a person detection unit 11, a camera identification unit 12, and an estimation unit 13.

[0014] The person detection unit 11 detects a person to be tracked from images generated by multiple cameras installed at predetermined locations. The camera identification unit 12 identifies a camera from which the timing at which the person to be tracked is detected in the image satisfies a predetermined condition based on a target time. The estimation unit 13 estimates an area in which the person to be tracked is present at the target time based on the installation positions of the identified cameras.

[0015] In this way, the estimation device 10 of this embodiment identifies cameras whose timing of detecting the tracking target person in the image satisfies predetermined conditions based on the target time, and estimates the area where the tracking target person is located at the target time based on the installation positions of the identified cameras. The estimation device 10 of this embodiment can estimate the area where the tracking target person is located at the target time with high accuracy.

[0016] Second Embodiment "Overview" The estimation device 10 of the second embodiment is a specific embodiment of the estimation device 10 of the first embodiment. That is, the estimation device 10 identifies cameras whose timing of detecting a tracking target person in an image satisfies a predetermined condition based on a target time, and estimates an area in which the tracking target person is present at the target time based on the installation positions of the identified cameras. This will be described in detail below.

[0017] "Hardware Configuration" An example of the hardware configuration of the estimation device 10 will be described. Each functional unit of the estimation device 10 is realized by any combination of hardware and software. Those skilled in the art will understand that there are various variations in the realization method and device. The software includes programs that are pre-stored in the device before shipping, and programs downloaded from recording media such as CDs (Compact Discs) or servers on the Internet.

[0018] FIG. 2 is a block diagram illustrating an example of the hardware configuration of the estimating device 10. As shown in FIG. 2, the estimating device 10 has a processor 1A, a memory 2A, an input / output interface 3A, a peripheral circuit 4A, and a bus 5A. The peripheral circuit 4A includes various modules. The estimating device 10 does not necessarily have to have the peripheral circuit 4A. Note that the estimating device 10 may be composed of multiple devices that are physically and / or logically separated. In this case, each of the multiple devices can have the above hardware configuration.

[0019] The bus 5A is a data transmission path for the processor 1A, memory 2A, peripheral circuit 4A, and input / output interface 3A to mutually transmit and receive data. The processor 1A is, for example, a processing unit such as a CPU or a graphics processing unit (GPU). The memory 2A is, for example, a random access memory (RAM) or a read-only memory (ROM). The input / output interface 3A includes interfaces for acquiring information from input devices, external devices, external servers, external sensors, cameras, etc., and interfaces for outputting information to output devices, external devices, external servers, etc. The input / output interface 3A also includes an interface for connecting to a communication network such as the Internet. Examples of input devices include a keyboard, mouse, microphone, physical buttons, touch panel, etc. Examples of output devices include a display, speaker, printer, mailer, etc. The processor 1A can issue commands to each module and perform calculations based on the results of those calculations.

[0020] "Functional Configuration" Next, the functional configuration of the estimation device 10 of this embodiment will be described in detail. Fig. 1 shows an example of a functional block diagram of the estimation device 10 of this embodiment. As shown in the figure, the estimation device 10 of this embodiment has a person detection unit 11, a camera identification unit 12, and an estimation unit 13.

[0021] The person detection unit 11 detects a person to be tracked from images generated by a plurality of cameras installed at predetermined positions.

[0022] "Image" is a concept that includes moving images.

[0023] A "camera" captures images. The camera may be a surveillance camera. The camera is installed at a predetermined location and captures the surrounding area of ​​that location. The installation location of the camera is not particularly limited. The camera may be installed on the street or inside a facility. Examples of facilities include, but are not limited to, department stores, museums, and art galleries. The camera may also be installed outdoors or indoors. Information indicating the installation location of each of the multiple cameras is registered in advance in the estimation device 10. The installation location of the camera may be indicated by latitude and longitude, or by address. Furthermore, in the case of a camera installed inside a facility, the installation location of the camera may be indicated by location information specific to the facility, such as a corridor number or room number.

[0024] Images generated by each of the multiple cameras are input to the estimation device 10 by any means. For example, the estimation device 10 and the cameras may be communicably connected. The cameras may then transmit the images they generate to the estimation device 10. Alternatively, the images generated by the cameras may be accumulated in any storage means. The images stored in the storage means may then be input to the estimation device 10 by manual operation by a user. The input of images to the estimation device 10 may be performed by real-time processing or batch processing. The person detection unit 11 can acquire the images input to the estimation device 10 in this manner. Note that the person detection unit 11 may acquire images by other means.

[0025] "Acquisition" includes at least one of the following: a device going to retrieve data or information stored in another device or storage medium (active acquisition), and a device inputting data or information output from another device (passive acquisition). Examples of active acquisition include making a request or inquiry to another device and receiving a reply, and accessing and reading information from another device or storage medium. An example of passive acquisition is receiving information that is distributed (or transmitted, pushed, etc.). Furthermore, "acquisition" may also mean selecting and acquiring data or information from received data or information, or selecting and receiving distributed data or information.

[0026] The person detection unit 11 detects the person to be tracked from the image acquired in this manner. The person detection unit 11 detects the person to be tracked from the image based on information indicating the feature amounts of the appearance of the person to be tracked.

[0027] The "information indicating the appearance feature quantities of the tracked person" is input to the estimation device 10 by the user. For example, the user may input the appearance feature quantities of the tracked person to the estimation device 10. Alternatively, the user may input an image of the tracked person to the estimation device 10. The person detection unit 11 may then analyze the image and extract the appearance feature quantities of the tracked person.

[0028] The "appearance feature amount of the person to be tracked" includes, but is not limited to, facial feature amount, body feature amount, clothing feature amount, belongings feature amount, shoe feature amount, and the like.

[0029] The person detection unit 11 detects a person to be tracked from images taken by multiple cameras based on the feature amounts of the appearance of the person to be tracked. The person detection unit 11 can then record the detection results in a detection history. The detection history indicates the timing at which the person to be tracked was detected for each camera. The timing at which the person to be tracked was detected is the shooting date and time of the frame image in which the person to be tracked was detected. The person detection unit 11 can identify the shooting date and time of the frame image in which the person to be tracked was detected based on the timestamp attached to the image.

[0030] The camera identification unit 12 identifies a camera whose timing of detecting a tracking target person in an image satisfies a predetermined condition. The predetermined condition is defined based on the target time. The camera identification unit 12 can identify a camera that satisfies the predetermined condition based on, for example, the detection history described above.

[0031] The "target time" is the timing at which an area where the tracked person is located is estimated. That is, the estimation device 10 estimates the area where the tracked person is located at the target time. The target time is the current time, or a current, future, or past time specified by the user. For example, the current time may be automatically set as the target time. Alternatively, the target time may be set by user input. The user can set any current, future, or past time as the target time.

[0032] The "predetermined condition" may include at least one of the following conditions 1 to 4.

[0033] (Condition 1) The person to be tracked was detected among the multiple cameras at the timing closest to the target time. (Condition 2) The person to be tracked was detected among the multiple cameras before the target time at the timing closest to the target time. (Condition 3) The person to be tracked was detected among the multiple cameras after the target time at the timing closest to the target time. (Condition 4) The person to be tracked was detected within a reference time from the target time.

[0034] When Condition 1 is a predetermined condition, the camera identification unit 12 identifies, from among the multiple cameras, a camera that detected the person to be tracked at a timing closest to the target time. The detection timing may be either before or after the target time.

[0035] If condition 2 is a predetermined condition, the camera identification unit 12 identifies, from among the multiple cameras, a camera that detected the person to be tracked before the target time and at the timing closest to the target time.

[0036] If condition 3 is a predetermined condition, the camera identification unit 12 identifies, from among the multiple cameras, a camera that detected the person to be tracked at a timing that is after and closest to the target time.

[0037] When Condition 4 is the predetermined condition, the camera identification unit 12 identifies, from among the multiple cameras, a camera that detected the person to be tracked within a reference time from the target time. The detection timing may be before or after the target time. When Condition 4 is the predetermined condition, the number of cameras identified by the camera identification unit 12 varies.

[0038] It should be noted that a condition obtained by connecting two or more of the above conditions 1 to 4 with an AND condition or an OR condition may be used as the predetermined condition.

[0039] For example, the predetermined conditions may be Condition 1 and Condition 4. In this case, the camera identification unit 12 identifies, from among the multiple cameras, a camera that detected the tracking target person within a reference time from the target time and at a timing closest to the target time. The detection timing may be either before or after the target time.

[0040] Alternatively, the predetermined conditions may be Condition 2 and Condition 4. In this case, the camera identification unit 12 identifies, from among the multiple cameras, a camera in which the tracking target person was detected within a reference time from the target time, and which detected the tracking target person before the target time and at a timing closest to the target time.

[0041] Alternatively, the predetermined conditions may be both Condition 3 and Condition 4. In this case, the camera identification unit 12 identifies, from among the multiple cameras, a camera in which the tracking target person was detected within a reference time from the target time, and in which the tracking target person was detected after the target time and at a timing closest to the target time.

[0042] Alternatively, condition 2 or condition 3 may be used as the predetermined condition. In this case, the camera identification unit 12 identifies, from among the multiple cameras, both a camera in which the tracking target person was detected before the target time and closest to the target time, and a camera in which the tracking target person was detected after the target time and closest to the target time. Condition 2 or condition 3 may be combined with condition 4 using an AND condition. A predetermined condition in which condition 2 or condition 3, or condition 2 or condition 3 and condition 4 are combined using an AND condition, may be used, for example, when the target time is a time in the past.

[0043] The estimation unit 13 estimates an area where the person to be tracked is present at a target time, based on the installation positions of the cameras identified by the camera identification unit 12. The estimation unit 13 can execute at least one of the following area estimation processes 1 to 3.

[0044] Area Estimation Process 1 Area estimation process 1 is suitable for use when one camera is identified by the camera identification unit 12 .

[0045] First, the estimation unit 13 calculates the travel time, which is the time difference between the target time and the most recent detection timing at which the person to be tracked was detected in the image generated by the camera identified by the camera identification unit 12.

[0046] The "most recent detection timing" is the timing closest to the target time among the timings at which the tracking target person was detected in the image generated by the camera identified by the camera identification unit 12. Note that, if a camera is identified as satisfying the above condition 2, the most recent detection timing is the timing before and closest to the target time among the timings at which the tracking target person was detected in the image generated by that camera. Also, if a camera is identified as satisfying the above condition 3, the most recent detection timing is the timing after and closest to the target time among the timings at which the tracking target person was detected in the image generated by that camera.

[0047] The estimation unit 13 also determines an estimated movement speed of the person to be tracked. Here, the process of calculating the estimated movement speed will be described. The estimation unit 13 determines the estimated movement speed based on the characteristics of the person to be tracked acquired by user input or analysis of images generated by the camera. The characteristics of the person to be tracked include at least one of the age, gender, whether or not the person is carrying luggage, the size of the luggage, whether or not the person is injured, the means of transportation (walking, bicycle, motorcycle, car, etc.), and the movement speed in the image of the person to be tracked.

[0048] The estimation unit 13 can calculate the estimated movement speed of the person to be tracked based on, for example, this person characteristic information. A speed calculation model is generated in advance, with this person characteristic information as input and an estimated movement speed calculated based on the input person characteristic information as output. The speed calculation model may be a function, a learning model generated by machine learning, or other models. The estimation unit 13 inputs the person characteristic information of the person to be tracked into such a speed calculation model, and obtains the estimated movement speed output from the speed calculation model.

[0049] Alternatively, the estimation unit 13 may calculate the movement speed of the person to be tracked in the image based on the image, and use the calculation result as the estimated movement speed of the person to be tracked. Calculation of the movement speed of the person detected in the image can be realized using any technology.

[0050] After calculating the travel time and estimated travel speed, the estimation unit 13 calculates the estimated travel distance of the person to be tracked between the most recent detection timing and the target time based on the travel time and estimated travel speed. The estimated travel distance can be simply calculated as the product of the travel time and the estimated travel speed, but other methods may also be used.

[0051] Then, as shown in Figure 3, the estimation unit 13 estimates area B within an estimated movement distance D from the installation position of camera C identified by the camera identification unit 12 as area A where the person to be tracked is located at the target time.

[0052] Area Estimation Process 2 Area estimation process 2 is suitable for use when two or more cameras are identified by the camera identification unit 12. For example, when the predetermined condition is condition 2 or condition 3, two cameras may be identified. Also, when the predetermined condition is condition 4, two or more cameras may be identified.

[0053] The estimation unit 13 calculates a travel time for each identified camera using a process similar to that described in area estimation process 1. The estimation unit 13 also calculates an estimated travel speed as a common value applied to all cameras using a process similar to that described in area estimation process 1. The estimation unit 13 then calculates an estimated travel distance for each identified camera using a process similar to that described in area estimation process 1.

[0054] Then, as shown in Fig. 4, the estimation unit 13 identifies an area within an estimated movement distance from the installation position of the camera for each camera identified by the camera identification unit 12. In Fig. 4, the camera identification unit 12 identifies two cameras C １ and C ２ In this example, the camera C １ Estimated travel distance D from the installation position １ Area B within １ And Camera C ２ Estimated travel distance D from the installation position ２ Area B within ２ The estimation unit 13 estimates the area B １ and Area B ２ When the camera identification unit 12 identifies M cameras (M is an integer of 2 or more), the estimation unit 13 estimates the overlapping area as an area A where the person to be tracked exists at the target time. １ ~B ｍ The area where all of these overlap can be estimated as area A where the person to be tracked exists at the target time.

[0055] Area Estimation Process 3 In area estimation process 3, the estimation unit 13 determines whether the following three conditions, which will be described with reference to FIG. 5, are satisfied.

[0056] The target time is the current time. As shown in FIG. 5, the camera C identified by the camera identification unit 12 １ is installed on one road. １ After the most recent detection timing at which the tracking target person was detected in the image generated by the other camera C ２ The person to be tracked is not detected in the image generated by the other camera C.２ is Camera C １ The camera is installed in the direction of movement of a person to be tracked on a single road (the direction indicated by the arrow in the figure) identified based on the image generated by the camera.

[0057] If all of the above three conditions are satisfied, the estimation unit 13 determines that the camera C identified by the camera identification unit 12 is the １ Imaging area E １ and the other cameras C mentioned above ２ Imaging area E ２ The area between is estimated as the area where the person to be tracked exists at the target time.

[0058] A "single road" is a road with no branches. A single road may be a road, or a corridor or passageway within a facility. Map information or a floor map of a facility, in which single roads are identifiable, is registered in advance in the estimation device 10. The estimation unit 13 can determine whether each road is a single road or not based on the information.

[0059] The "direction of movement of the person to be tracked on a single road" can be identified using any technique based on the direction of movement of the person to be tracked in the image.

[0060] "Another camera C installed ahead in the direction of movement ２ " can be identified based on information indicating the installation location of the camera that has been registered in advance, as well as the above-mentioned map information, a floor map of the facility, etc.

[0061] Next, an example of the processing flow of the estimation device 10 will be described with reference to the flowchart of FIG.

[0062] First, the estimation device 10 executes a process of detecting a tracking target person from images generated by multiple cameras installed at predetermined locations (S10). Next, the estimation device 10 identifies cameras whose timing of detecting the tracking target person in the images satisfies predetermined conditions based on the target time (S11). Next, the estimation device 10 estimates the area in which the tracking target person is present at the target time based on the installation locations of the cameras identified in S11 (S12). Note that if no cameras are identified in S11, the estimation device 10 can end the process without executing S12.

[0063] Although not shown, the estimation device 10 can output the estimation result. The estimation device 10 can output the estimation result via an output device such as a display or a projection device. For example, as shown in FIGS. 3 and 4 , the estimation device 10 may output, as the estimation result, information indicating an area A on a map (or a floor map of a facility) where the tracking target person is estimated to be present at the target time. The estimation device 10 may also output information indicating the installation position of the camera identified by the camera identification unit 12, the most recent detection timing, etc., as auxiliary information to the estimation result. Furthermore, the estimation device 10 may output the estimated movement speed and estimated movement distance as auxiliary information to the estimation result.

[0064] "Effects" The estimation device 10 of this embodiment identifies a camera that detected the tracked person at a timing close to the target time, and estimates the area where the tracked person is located at the target time based on the installation location of that camera and the estimated movement speed of the tracked person. With this estimation device 10, even if the tracked person is not captured in an image taken at the target time, it is possible to estimate the area where the tracked person is located at the target time. In this embodiment, even if the capture areas of multiple cameras do not overlap and there are areas not captured by any of the cameras, it is possible to estimate the area where the tracked person is located at the target time.

[0065] The estimation device 10 of this embodiment includes a unit for estimating the visit destinations of the tracked person in facilities located in an area where the tracked person is estimated to be present. This will be described in detail below.

[0066] The estimation unit 13 estimates an area in which the tracked person is located at a target time using the method described in the first and second embodiments. Then, the estimation unit 13 estimates a visit target of the tracked person among facilities located within the estimated area based on at least one of the tracked person's clothing, belongings, age, gender, accompanying person, means of transportation, travel route, and target time. Note that the facility estimated to be a visit target may be one or multiple.

[0067] The clothing, belongings, age, gender, accompanying persons, mode of transportation, and route of travel of the tracked person are determined through user input or analysis of camera-generated images.

[0068] "Clothing" refers to the type of clothing. For example, it can be athletic wear, fashionable wear, casual wear, a suit, etc. There are various ways to identify the type of clothing through image analysis. For example, the classification can be performed based on the brand, design, shape, etc. of the clothing. For example, the classification can be achieved using a classifier generated by machine learning, or other means can be used.

[0069] "Belongings" indicates the type of belongings. For example, sports equipment, business bags, shopping bags, etc. There are various means for identifying the type of belongings using image analysis. For example, the above classification can be performed based on the characteristics of their appearance. For example, the classification can be achieved using a classifier generated by machine learning, or other means can be used.

[0070] "Accompanying person" indicates whether or not the person is accompanied by another person, and the age and sex of the person.

[0071] "Means of transportation" include walking, bicycle, motorcycle, car, etc.

[0072] The "movement route" is determined based on the detection results from multiple cameras.

[0073] Next, an example of a process for estimating a visit destination of a tracked person in a facility located within an estimated area will be described. The estimation unit 13 can execute at least one of the following facility estimation processes 1 to 4.

[0074] Facility Estimation Process 1 First, the estimation unit 13 identifies facilities that exist within the estimated area based on map information, a floor map of the facility, and the like that are registered in advance in the estimation device 10. Examples of facilities whose locations are indicated by map information, a floor map of the facility, and the like include parks, supermarkets, department stores, hospitals, and the like. Examples of facilities provided within the facility include a kids' corner, a nursing room, a diaper changing station, and exercise facilities. Note that the examples here are merely examples and are not limited to these.

[0075] Then, characteristic information for each of a plurality of facilities is registered in advance in the estimation device 10. The characteristic information for the facilities indicates the purpose of each facility, the characteristics of the people who use each facility, and the time of use of each facility. The purposes of each facility are exercise, shopping, play, etc. The characteristics of the people who use each facility are indicated by clothing, belongings, age, gender, means of transportation, etc.

[0076] The estimation unit 13 can estimate the purpose of the tracked person from, for example, the clothing and belongings of the tracked person, and estimate facilities that match the purpose as the tracked person's visit destination. For example, information that associates the types of clothing and belongings with the purpose may be registered in advance in the estimation device 10. Then, the estimation unit 13 may estimate the purpose of the tracked person based on the information.

[0077] Facility Estimation Process 2 The estimation unit 13 can estimate, for example, that a facility is a destination of the tracked person if the similarity between the characteristics of the person using the facility and the characteristics of the tracked person is equal to or greater than a reference value. The similarity of characteristics can be calculated using any technique. For example, the similarity may be calculated based on the number of items whose values ​​match. In this case, the greater the number of items whose values ​​match, the higher the similarity.

[0078] The characteristics of the person using the facility and the characteristics of the person being tracked are as described in Facility Estimation Process 1. The items are items included in the characteristics of a person, such as clothing, belongings, age, gender, and means of transportation. Note that the characteristics of the person using the facility may be linked to each item and set to multiple values. For example, a facility that is used by both men and women may be linked to gender and set to both male and female. In such a case, "the characteristics of the person using the facility and the characteristics of the person being tracked match" means that the characteristics of the person being tracked are included in the characteristics of the person using the facility.

[0079] Facility Estimation Process 3 The estimation unit 13 can estimate the facilities visited by the person to be tracked based on the movement route.

[0080] A specific example of facility estimation process 3 will be described with reference to Fig. 7. Fig. 7 shows a route R of a person to be tracked, an area A where the person to be tracked is estimated to be present at a target time, and a plurality of facilities F present in area A. １ ~F ３ The person to be tracked visits facility F. １ and Facility F ３ On the other hand, if the destination of the person to be tracked is facility F, then the person to be tracked will take a detour to reach those facilities. ２ If so, the person being tracked will reach the facility via the shortest route.

[0081] It is unlikely that a person will take a detour, and it is considered that the person will usually head to the destination facility via the shortest route. ２ The estimation unit 13 estimates the facilities to be visited by the person to be tracked. The estimation unit 13 identifies the facilities that the person to be tracked will reach via the shortest route based on the movement route of the person to be tracked and the positional relationship between the person to be tracked and each of the facilities. The estimation unit 13 then estimates the identified facilities as the places to be visited by the person to be tracked.

[0082] There are various means for determining whether the route to each facility is the shortest or a longer route. For example, the shortest route can be determined by a route search that sets any position on the movement route R of the tracked person as the starting point and each facility as the destination. Then, if the route is the same as the shortest route calculated by the route search or if the deviation from the shortest route calculated by the route search is within a reference value, it may be determined to be the shortest route. If these conditions are not met, it may be determined to be a longer route. The starting point may be changed to another position on the movement route R and the above process may be performed multiple times. Then, a facility that is determined to be the shortest route in either case, or that has been determined to be the shortest route a predetermined number of times or more, may be estimated to be a facility visited by the tracked person.

[0083] The deviation from the shortest route calculated by route search is indicated by the difference in distance between the first route and the second route, or the difference in the time required for travel. The larger the difference, the greater the deviation from the shortest route. The first route is the "shortest route calculated by route search." The second route is "a route that has the same starting point and destination as the first route, moves to the end point indicated by the movement route R of the person to be tracked, and then moves from there to the destination point via the shortest route calculated by route search."

[0084] Facility Estimation Process 4 The estimation unit 13 can exclude facilities whose target time is not within the usage hours from the destinations of the tracked person. For example, the estimation unit 13 may estimate the remaining facilities that have not been excluded as the destinations of the tracked person. Alternatively, the estimation unit 13 may estimate the destinations of the tracked person from the remaining facilities that have not been excluded using any of the above-described facility estimation processes 1 to 3.

[0085] Next, an example of the processing flow of the estimation device 10 will be described with reference to the flowchart of FIG.

[0086] First, the estimation device 10 executes a process of detecting a tracking target person from images generated by multiple cameras installed at predetermined locations (S20). Next, the estimation device 10 identifies cameras whose timing of detecting the tracking target person in the images satisfies predetermined conditions based on the target time (S21). Next, the estimation device 10 estimates the area where the tracking target person is located at the target time based on the installation locations of the cameras identified in S21 (S22). Thereafter, the estimation device 10 estimates the destinations of the tracking target person within the facilities located within the area estimated in S22 (S23). Note that if no cameras are identified in S21, the estimation device 10 can end the process without executing S22 and S23.

[0087] Although not shown, the estimation device 10 can output the estimation result. The estimation device 10 can output the estimation result via an output device such as a display or a projection device. For example, as shown in FIGS. 3 and 4 , the estimation device 10 may output, as the estimation result, information indicating an area A on a map (or a floor map of a facility) where the tracked person is estimated to be present at the target time. The estimation device 10 may also highlight facilities estimated to be visited by the tracked person on the map (or the floor map of the facility). The estimation device 10 may also output information indicating the installation locations of the cameras identified by the camera identification unit 12, the most recent detection timing, and the like, as auxiliary information to the estimation result. Furthermore, the estimation device 10 may output the estimated movement speed and estimated movement distance as auxiliary information to the estimation result.

[0088] Other configurations of the estimation device 10 are similar to those of the first and second embodiments.

[0089] The estimation device 10 of this embodiment achieves the same effects as the estimation devices 10 of the first and second embodiments. Furthermore, the estimation device 10 of this embodiment can estimate the destinations (facilities) of a tracked person based on at least one of the tracked person's clothing, belongings, age, gender, accompanying person, vehicle, travel route, and target time.

[0090] <Modification> As shown in FIG. 9, the estimation device 10 estimates that a person to be tracked is present at a target time by capturing a predetermined camera C in an area A. ２ and C ３ Imaging area E ２ and E ３ It is possible to output information indicating the predetermined camera C. ２ and C ３ is a camera that has not been identified by the camera identification unit 12. That is, the predetermined camera C ２ and C ３ is a camera that does not satisfy the predetermined conditions detailed in the second embodiment. ２ and C ３ Imaging area E ２ and E ３ 9 can be excluded from the candidate areas where the person to be tracked is present at the target time. Based on the information as shown in FIG. 9, the user can ascertain the area where the person to be tracked is present at the target time.

[0091] Although the embodiments of the present invention have been described above with reference to the drawings, these are merely examples of the present invention, and various other configurations may be adopted. The configurations of the above-described embodiments may be combined with each other, or some of the configurations may be replaced with other configurations. Furthermore, various modifications may be made to the configurations of the above-described embodiments without departing from the spirit of the invention. Furthermore, the configurations and processes disclosed in the above-described embodiments and modified examples may be combined with each other.

[0092] In addition, in the flowcharts used in the above description, multiple steps (processes) are described in order. However, the order of the steps performed in each embodiment is not limited to the order described. In each embodiment, the order of the steps shown in the drawings can be changed to the extent that the content is not affected. Furthermore, the above-described embodiments can be combined to the extent that the content is not contradictory.

[0093] Some or all of the above embodiments may be described as, but are not limited to, the following supplementary notes. 1. An estimation device comprising: person detection means for detecting a tracking target person from images generated by multiple cameras installed at predetermined positions; camera identification means for identifying the cameras at which the timing at which the tracking target person was detected in the images satisfies a predetermined condition based on a target time; and estimation means for estimating an area in which the tracking target person is present at the target time based on the installation positions of the identified cameras. 2. The estimation device according to 1, wherein the predetermined condition includes at least one of: the tracking target person was detected among the multiple cameras at a timing closest to the target time; the tracking target person was detected among the multiple cameras before the target time at a timing closest to the target time; the tracking target person was detected among the multiple cameras after the target time at a timing closest to the target time; and the tracking target person was detected within a reference time from the target time. 3. The estimation device according to 1 or 2, wherein the target time is the current time, or a current, future, or past time specified by a user. 4. 4. The estimation device according to any one of 1 to 3, wherein the estimation means calculates an estimated movement distance of the tracked person between the most recent detection time and the target time based on the time difference between the most recent detection time at which the tracked person was detected in an image generated by the specified camera and the target time, and an estimated movement speed of the tracked person, and estimates an area within the estimated movement distance from the installation position of the specified camera as the area where the tracked person is present at the target time. 5. The estimation device according to 4, wherein the estimation means estimates, as the area where the tracked person is present at the target time, an overlapping area of ​​an area within the estimated movement distance from the installation position of the camera at which the tracked person was detected before the target time and closest to the target time, and an area within the estimated movement distance from the installation position of the camera at which the tracked person was detected after the target time and closest to the target time.6. The estimation device according to 4 or 5, wherein the estimation means determines the estimated movement speed based on characteristics of the tracked person acquired by user input or by analysis of images generated by the camera. 7. The estimation device according to 6, wherein the characteristics of the tracked person include at least one of the tracked person's age, sex, whether or not they have luggage, the size of the luggage, whether or not they are injured, their means of transportation, and their movement speed in the images. 8. The estimation device according to any of 1 to 7, wherein the estimation means estimates an area between the identified camera and another camera as an area where the tracked person is present at the target time, when the target time is the current time, the identified camera is installed on a single road, and after the tracked person is detected in an image generated by the identified camera, the tracked person is not detected in an image generated by another camera installed ahead of the tracked person identified based on the image generated by the identified camera. 9. 10. An estimation method in which one or more computers detect the tracked person from images generated by multiple cameras installed at predetermined positions, identify the cameras whose timing at which the tracked person was detected in the images satisfies a predetermined condition based on the target time, and estimate the area in which the tracked person is located at the target time based on the installation positions of the identified cameras. 11. A program that causes a computer to function as: person detection means that detects the tracked person from images generated by multiple cameras installed at predetermined positions; camera identification means that identifies the cameras whose timing at which the tracked person was detected in the images satisfies a predetermined condition based on the target time; and estimation means that estimates the area in which the tracked person is located at the target time based on the installation positions of the identified cameras.

[0094] This application claims priority based on Japanese Patent Application No. 2023-041791, filed March 16, 2023, the disclosure of which is incorporated herein by reference in its entirety.

[0095] 10 Estimation device 11 Person detection unit 12 Camera identification unit 13 Estimation unit 1A Processor 2A Memory 3A Input / output I / F 4A Peripheral circuit 5A Bus

Claims

1. a person detection means for detecting a person to be tracked from images generated by a plurality of cameras installed at predetermined positions; a camera identification means for identifying the camera in which the timing at which the person to be tracked is detected in the image satisfies a predetermined condition based on a target time; an estimation means for estimating an area in which the person to be tracked is present at the target time based on the identified installation position of the camera; An estimation device having:

2. The predetermined condition is: The person to be tracked was detected at the timing closest to the target time among the plurality of cameras. The person to be tracked is detected before the target time among the plurality of cameras at a timing closest to the target time. The tracking target person is detected after the target time among the plurality of cameras at a timing closest to the target time, and The person to be tracked is detected within a reference time from the target time. The estimation device according to claim 1 , comprising at least one of:

3. The target time is 3. The estimation device according to claim 1, wherein the time is the current time, or a current, future or past time designated by the user.

4. The estimation means calculating an estimated movement distance of the tracked person between the most recent detection time and the target time based on a time difference between the most recent detection time at which the tracked person was detected in an image generated by the identified camera and the target time, and an estimated movement speed of the tracked person; The estimation device according to claim 1 or 2, wherein an area within the estimated movement distance from the specified installation position of the camera is estimated as an area in which the person to be tracked is present at the target time.

5. The estimation means 5. The estimation device according to claim 4, wherein an overlapping area of ​​an area within the estimated movement distance from the installation position of the camera where the tracked person was detected before the target time and closest to the target time, and an area within the estimated movement distance from the installation position of the camera where the tracked person was detected after the target time and closest to the target time, is estimated as the area where the tracked person is present at the target time.

6. The estimation means The estimation device of claim 4 , wherein the estimated moving speed is determined based on a user input or a characteristic of the person being tracked obtained by analyzing an image generated by the camera.

7. The estimation means the target time is the current time, The identified camera is installed on a single road, and After the tracking target person is detected in the image generated by the identified camera, if the tracking target person is not detected in the image generated by another camera installed ahead of the tracking target person identified based on the image generated by the identified camera in the moving direction of the tracking target person, The estimation device according to claim 1 or 2, wherein an area between the identified camera and the other camera is estimated as an area in which the person to be tracked is present at the target time.

8. The estimation means 3. The estimation device according to claim 1, wherein the estimation device estimates the destinations of the tracked person among facilities present in an area where the tracked person is estimated to be present at the target time, based on at least one of the tracked person's clothing, belongings, age, sex, accompanying person, means of transportation, travel route, and the target time.

9. One or more computers Detects the person to be tracked from images generated by multiple cameras installed at predetermined locations, Identifying the camera in which the timing at which the person to be tracked is detected in the image satisfies a predetermined condition based on a target time; An estimation method for estimating an area in which the person to be tracked is present at the target time based on the identified installation position of the camera.

10. Computer, a person detection means for detecting a person to be tracked from images generated by a plurality of cameras installed at predetermined positions; a camera identification means for identifying the camera in which the timing at which the person to be tracked is detected in the image satisfies a predetermined condition based on a target time; an estimation means for estimating an area in which the person to be tracked is present at the target time based on the identified installation position of the camera; A program that functions as a