JP7374301B2 - Information processing device and information processing method - Google Patents

Description

The present disclosure relates to an information processing device and an information processing method that generate and output data that takes personal privacy into consideration.

In recent years, a system has been developed that visualizes people's flow lines and crowding conditions within facilities by analyzing the attributes and behaviors of people included in images taken with cameras installed within facilities (hereinafter referred to as "camera images"). The data distribution business, which distributes personal information such as people's attributes and actions, and utilizes it in third-party services and systems, is becoming widespread. On the other hand, from the perspective of protecting individual privacy, when using information about people, such as camera images, it is necessary to process the data in a way that prevents individuals from being identified.

For example, as a system for analyzing human flow lines, a flow line analysis system is known that generates a flow line analysis image in which flow line information such as residence information or passing information of a moving object included in a camera image is superimposed (Patent Document (see 1). Patent Document 1 discloses that a background image and flow line information are generated from a camera image, and the flow line information is superimposed on the background image. To protect privacy, people included in the camera image are excluded when the background image is generated. The flow line information is linked to the attributes of the person included in the camera image, such as gender, age, and generation, and the user can specify conditions for the person's attributes and the amount of people passing through. It is possible to extract flow line information.

JP 2017-123024 (pages 6-35, Figure 2)

However, in the system described in Patent Document 1 mentioned above, flow line information is extracted even when the number of people included in the camera image is one. If there is only one man over 100 years old in the town where the system is installed, it may be possible to identify the person, which may pose a problem from a privacy perspective, and a third party may There was a problem that it became difficult to use.

The present disclosure has been made in order to solve the above problems, and aims to provide an information processing device that can generate privacy-friendly data from information about a person included in a camera image.

An information processing device according to the present disclosure includes a data acquisition unit that acquires image data, a setting unit that sets a threshold, and an estimation unit that recognizes a person from the image data and estimates estimated information for each of the recognized persons. , a data generation unit that generates provided information from the estimated information based on a comparison between the aggregation result of the estimated information and the threshold value, and a data output unit that outputs the provided information, the estimated information is The data generation unit includes at least one item of information indicating attributes and states of the person recognized from the data, and the data generation unit is configured to determine that, as a result of aggregating the estimated information, the number of people whose items match the estimated information is equal to the threshold. If the number is less than 1, the provided information is generated by excluding items of the estimated information regarding the person .

According to the present disclosure, person recognition and estimated information for each person are estimated from the acquired image data, and provided information is output after comparing the aggregated result of the estimated information with a set threshold, so privacy is considered. data can be generated.

1 is a block diagram schematically showing the configuration of an information processing device according to Embodiment 1. FIG. 1 is a diagram illustrating an example configuration of a system equipped with an information processing device according to Embodiment 1. FIG. FIG. 3 is a diagram showing an example of a camera image taken with a camera. 3 is a diagram illustrating an example of a history information table stored in an estimated information storage unit 140. FIG. This is a table in which the number of people included in historical information is aggregated for each predetermined time (time period). It is a figure showing an example of an environment information table stored in a data acquisition environment storage part. 5 is a flowchart showing a process of generating provided information in a data generation unit according to the first embodiment. 1 is a diagram illustrating an information processing apparatus according to a first embodiment using a processor; FIG.

Embodiment 1.
Embodiments will be described below with reference to the drawings. In the description of the drawings below, the same or similar parts are designated by the same or similar symbols. However, it should be noted that the drawings are schematic and the ratio of each dimension may differ from the actual one. Therefore, specific dimensions etc. should be determined with reference to the following explanation. Furthermore, it goes without saying that the drawings include portions with different dimensional relationships and ratios.

FIG. 1 is a block diagram schematically showing the configuration of an information processing apparatus according to this embodiment. Further, FIG. 2 is a diagram showing an example of the configuration of a system equipped with an information processing device.
Information processing device 100 according to this embodiment includes a data acquisition section 120, a setting section 110, an estimation section 130, a data generation section 160, and a data output section 170. Further, in FIG. 1, the information processing apparatus 100 includes an estimated information storage section 140 and a data acquisition environment storage section 150.

The setting section 110 outputs the set threshold value to the data generation section 160. The data acquisition unit 120 outputs image data acquired from the camera to the estimation unit 130. The estimation unit 130 outputs estimation information for each person recognized from the acquired image data to the data generation unit 160. The data generation unit 160 generates provided information from the estimated information based on a comparison between the total result of the estimated information and the threshold value, and outputs the generated information to the data output unit 170. The data output unit 170 outputs provided information. Further, the estimated information storage unit 140 stores the estimated information from the estimating unit 130, and outputs the stored estimated information to the setting unit 110 and the data generating unit 160 as history information. The data acquisition environment storage unit 150 stores history information acquisition conditions that can be used to generate provided information as environment information.

FIG. 2 shows a configuration example in which the information processing device 100 is installed inside a train vehicle. Although FIG. 2 shows a pattern in which the information processing device 100 is incorporated into the camera 10, there is also a pattern in which the information processing device 100 is incorporated in an edge terminal 20 that connects to the camera 10 by wire or wirelessly, or a pattern in which the information processing device 100 is connected to the camera 10 by wire or wirelessly. Alternatively, a pattern in which the information processing device 100 is incorporated into a wirelessly connected server (not shown) is also conceivable. In this embodiment, a situation will be described in which a camera 10 incorporating an information processing device 100 is installed in a train car, and an edge terminal 20 is connected to the camera 10 and the monitor 30 by wire. . In FIG. 2, it is assumed that at least one camera 10 per vehicle is installed at a position (near the ceiling, etc.) where it can photograph the interior of the vehicle. The camera 10 generates provided information including the passenger's attributes and status from the passenger information included in the camera image, and transmits it to the edge terminal 20. The edge terminal 20 switches the content displayed on the monitor 30 based on the provided information and train operation information received from the camera 10.

Each component of the information processing device 100 will be explained.
The setting unit 110 sets a threshold value that is a condition for outputting the provided information, and sends the threshold value to the data generation unit 160. For example, a threshold value setting screen is displayed on a display (not shown), and the user sets an arbitrary threshold value through an input device (not shown) such as a keyboard or mouse. The display and input device used to set the threshold value may be connected to the edge terminal 20 shown in FIG. In this embodiment, the threshold value is set to the number of people who have the same attribute as the person included in the camera image, but any other value may be set as the threshold value as long as it is a value that satisfies the output condition of the provided information. Alternatively, only the number of people may be set as the threshold value itself, and the data generation unit 160 described below may generate the provided information by determining whether the number of people having the same information is equal to or greater than the threshold value. Further, the threshold value to be set is set to be larger than a predetermined provision permission threshold value. The provision permission threshold is, for example, 1.

Since the threshold value in the setting unit 110 is set by the user, if an appropriate threshold value is not set, there is a possibility that a situation in which the provision information cannot be generated may occur frequently. For example, if a value that is clearly larger than the number of passengers is set as the threshold value, the provided information cannot be generated because the provided information output conditions are not satisfied. Therefore, when setting the threshold in the setting unit 110, if a certain number of pieces of historical information are stored in the estimated information storage unit 140, the average value of the number of people included in the historical information acquired from the estimated information storage unit 140 is calculated. Then, the average value for each time period will be presented to the user. The user can set an appropriate threshold based on this average value result. Details of the method for calculating the average value for each time period will be described later.

The data acquisition unit 120 acquires a camera image from an output unit (not shown) of an external terminal such as a surveillance camera, and sends it to the estimation unit 130. When the information processing device 100 is incorporated into the camera 10, the data acquisition unit 120 acquires a camera image taken by an imaging unit (not shown) of the camera 10. The data acquisition unit 120 acquires camera images according to a predetermined frame rate. Alternatively, the timing for photographing a camera image may be acquired from an external terminal connected to the camera 10, and the photographed camera image may be sent to the data acquisition unit 120. For example, the camera 10 acquires train operation information from the edge terminal 20, photographs a camera image when the train's running state changes (stopped, running between stations, before the next station stop, etc.), and the data acquisition unit You may also send it to 120.

The estimation unit 130 recognizes the person included in the camera image and estimates the attributes and state of the recognized person as estimated information. Items related to the person's attributes include the person's age, gender, and height, and items related to the person's condition include the person's location, actions, vehicle, and belongings. The estimation unit 130 then sends the estimation information to the estimation information storage unit 140 and the estimation information to the data generation unit 160.

Here, the attributes and status of the person included in the camera image are estimated using a known image recognition technique. In this embodiment, the position of the person is defined as the position of the person in the camera image (x, y), with the lower left of the camera image as the starting point (x0, y0), the horizontal axis is the x axis, and the vertical axis is the y axis. do.

For example, "standing" or "sitting" is estimated as the person's behavior.
Furthermore, it may be possible to estimate what the user is looking at. It estimates what people are looking at, such as ``in-car advertisements,'' ``destination information,'' ``mobile devices,''``books,''``people,'' and ``eyes closed.''
The vehicle is estimated to be, for example, "chair,""stroller," or "not riding." If the image is taken from a camera installed outdoors, a "bicycle" may also be a candidate.
Possible items are estimated to be, for example, ``suitcase,''``stroller,''``bag,'' or ``no belongings to be classified.''
These estimated information may be estimated by image recognition technology by allowing the user to additionally set attributes and status items that they wish to classify.

The estimated information may be limited to what the user needs; for example, if the camera image is captured within a vehicle, the state of the vehicle may not be estimated. The same applies when the shooting target area is inside a vehicle such as a taxi or a bus. However, when the camera 10 is installed in a city, it is also assumed that the person included in the camera image is riding a bicycle. In such a scene, the vehicle may also be estimated as the state of the person included in the camera image.

FIG. 3 shows an example of a camera image used for estimation by the estimation unit 130.
Here, person 1a included in camera image 1 is a 40-year-old man who is sitting on a seat with a suitcase, and person 1b is a 30-year-old man who is standing without carrying any belongings. do.
The estimation unit 130 first detects a person from the acquired camera image, and then performs image recognition to estimate the attributes and state of the detected person. As a result, for example, for the person 1a, it is estimated that (age, gender, location, behavior, belongings) = (40, male, (x1, y1), sitting, suitcase). For person 1b, it is estimated that (age, gender, location, behavior, belongings) = (30, male, (x2, y2), standing, no belongings). In this manner, the estimation unit 130 generates estimation information for each person included in the camera image, and by summing up the estimation information, information on the number of recognized persons can be obtained.

Further, the estimating unit 130 may obtain the number of people included in the camera image using a known person detection technique in the same way as estimating the attributes and states of the people included in the camera image. The number of people detected here may be the number of people included in the camera image, or may be the number of people whose attributes or states match. The number of people whose attributes or states match is, for example, when there are three people included in a camera image and the estimated gender of the attributes of the people included in the camera image is "male, male, female." , the number of people is (male, female) = (2, 1). The number of people in this embodiment will be explained below as the number of people whose attributes or states match.

The camera image acquired by the estimation unit 130 is discarded after the estimation information is acquired. However, the data may be stored in any storage unit or transferred to the edge terminal without being discarded. In this case, it is desirable to encrypt the camera image before storing or transmitting it.

Estimated information storage section 140 stores estimated information acquired from estimation section 130 as history information. Further, the estimated information storage section 140 sends history information to the setting section 110 or the data generating section 160 in response to a request from the setting section 110 or the data generating section 160.

FIG. 4 is a diagram showing an example of a history information table stored in the estimated information storage unit 140. In FIG. 4, the information stored in the history information table includes the acquisition time of the image data used for estimation, the age and gender as information on the attributes estimated for each person recognized as a person, and the information on the image data as information on the estimated state. It shows the location information, actions, and belongings of the recognized person. The history information table stores estimated information in the order of acquisition time of camera images by the data acquisition unit 120. For example, when the estimation unit 130 acquires the camera image of FIG. 3 at "19/12/17 8:00", it is stored in the history information table as shown in FIG. The format of the acquisition time in the history information table is, for example, "year (last two digits of the Western calendar)/month/day, hour: minute", but is not limited to this format and may be in other formats.

A case will be described in which the estimated information storage section 140 sends history information to the setting section 110. The estimated information storage unit 140 stores a request from the setting unit 110 when a certain number or more of history information is stored at the time of threshold setting in the setting unit 110 (for example, when one week's worth of history information is stored). The history information is sent to the setting unit 110 in accordance with the above.

Next, we briefly return to the explanation of the setting section 110. The setting unit 110 calculates the average number of people from the history information. First, the number of people included in the history information acquired from the estimated information storage unit 140 is totaled every predetermined time. Here, in the following description, the predetermined time will be explained as one hour, but it may be possible to set it to another time instead of one hour.

FIG. 5 shows a table in which the number of people included in the history information is totaled for each predetermined time (time period). FIG. 5 shows the number of people information for each day during the same time period from 5:00 to 10:00. For example, from 5:00 to 6:00 on December 10th, no person was recognized from the image data, that is, no person was shown in the camera image. Furthermore, since only one person was recognized from the image data between 5:00 and 6:00 on December 16th, one person was shown in the camera image.

Next, an average value is calculated for the aggregation results for each time period. For example, if the total result for each day in the time period from 7:00 to 8:00 from December 10th to December 16th is "8, 4, 7, 7, 6, 5, 5", then from 7:00 to 8:00 The average value at the time is (8+4+7+7+6+5+5)/7=6 people. By showing the average value result to the user, it becomes a reference material when the user sets the threshold value, for example, by setting the threshold value to the average value itself or setting it slightly less than the average value.

Also, if the aggregate result for each day from 5:00 to 6:00 from December 10th to December 16th is "0, 0, 1, 0, 0, 1, 1", the average value is 3 /7, which is less than one person. By showing this average value result to the user, the user can generate the information provided for that time period, since there are many cases where a person is not visible in the camera image on average each day, and it may be possible to identify the person by one person. You can decide not to. Note that the average value result may be an integer value with the decimal point rounded down.

In order to calculate the average value, we have explained an example in which the aggregated results of consecutive days (in the case of Figure 5, 7 days from 12/10 to 12/16) are used to calculate the recommended value. The data may be aggregated over the same time period and used to calculate the average value. For example, when applied to trains, on weekdays there are many commuters and students, but on holidays, there are fewer commuters and students, and there are more shoppers and passengers heading to leisure facilities. Since some routes have a trend, it is possible to calculate the average value for the same time period on the same day of the week, or calculate and output the average value for the same time period on weekdays and holidays, so that users can see the trend. It is possible to prompt threshold setting based on the combined average value results.

The data acquisition environment storage unit 150 stores history information acquisition conditions that can be used to generate provided information as environment information. In this embodiment, the history information acquisition condition indicates a time period of history information that can be acquired from the estimated information storage unit 140.

FIG. 6 is a diagram showing an example of an environment information table stored in the data acquisition environment storage unit 150. In FIG. 6, the history information acquisition condition of the camera 10a is shown as "1 hour". For example, if the current time is 10 o'clock, the history information used for the provided information generated by the data generation unit 160 is from the history information from 9 o'clock one hour before the current time to 10 o'clock the current time. Provided information can be obtained. In other words, it means that the provided information can be created from the history information from 9 o'clock to 10 o'clock.
Furthermore, like the camera 10d in FIG. 6, it is also possible to set the history information acquisition condition to "no condition". In FIG. 6, "no condition" is indicated by "-". In this case, it is assumed that there are no restrictions when acquiring historical information, and that historical information that exists from the current time can be acquired.

In this embodiment, the environment information table in the data acquisition environment storage unit 150 stores environment information in advance. For example, the historical information acquisition condition, which is environmental information, may be the travel time between specific stations. By doing so, information on passengers boarding trains running between specific stations can be acquired as provided information. Here, if there are no historical information acquisition conditions, for example, if you want to obtain information on passengers boarding from station B to station D, the provided information can be retrieved if the passenger is traveling near station B and the historical information is used. When generating the information, information from a camera image taken inside the vehicle before arriving at Station B will be used.

In FIG. 6, as an example of the environmental information table, the time of the acquisition condition associated with the camera is set. However, the present invention is not limited to this, and the condition may be set such that the history information can be acquired in a section represented by multiple pieces of position information, for example. do not have. For example, if the information processing device is mounted on a moving object, information about the location of the moving object at the time when the image data was acquired is also stored in history information, and if the moving object is a train, environmental information can be stored. If it is set in the table that the section from the position of A station to the position of C station can be obtained, it is possible to generate the provided information from the historical information from A station to C station. As a result, even if there is a train operation error, it is possible to apply the conditions as long as they are set for the section to be obtained.

The data generation unit 160 generates provided information by acquiring a threshold value from the setting unit 110, estimation information from the estimation unit 130, history information from the estimation information storage unit 140, and environment information from the data acquisition environment storage unit 150. , and sent to the data output section 170. Note that when generating provided information for each acquired image data, the provided information can be generated without the estimated information storage unit 140.

A method of generating provided information in the data generation unit 160 will be described assuming that the threshold value obtained from the setting unit 110 is “2”. If the number of people acquired or totaled by the data generation unit 160 is (male, female) = (2, 0), the estimated information is further adjusted so that the number of people who are the same for the age in the estimated information is 2 or more. Process.

The estimated information obtained by the data generation unit 160 indicates that the person 1a is estimated to be (age, gender, location, behavior, belongings) = (40, male, (x1, y1), sitting, suitcase), and (age, gender, , position, action, belongings) = (30, male, (x2, y2), standing, no belongings) and the estimated person 1b is recognized, by setting the age to "30 to 40 years old". This means that there are two people with the same age range, which is the same number as the threshold value "2". In this case, the provided information is (age, gender, location, behavior, belongings) = (30-40, male, (x1, y1), sitting, suitcase) and (age, gender, location, behavior, belongings) = (30-40, male, (x2, y2), standing, no belongings) is generated. Here, the expression 30-40 indicates an estimated age range of 30 to 40 years.

Also, if the current time is 7 o'clock, the history information from 6 o'clock to 7 o'clock is (age, gender, location, behavior, belongings) = (20, female, (x6, y6), sitting, no belongings). Obtain the estimated person 1c and the estimated person 1d (age, gender, location, behavior, belongings) = (60, male, (x7, y7), sitting, no belongings), and further generate data at the current time. The estimated information acquired in the section 160 indicates the person 1e estimated as (age, gender, location, behavior, belongings) = (20, female, (x3, y3), sitting, no belongings) and (age, gender, location). , behavior, belongings) = (40, female, (x4, y4), sitting, no belongings) and the person 1f estimated as age, gender, location, behavior, belongings) = (40, male, (x5, y5) , standing, suitcase) is estimated to be 1g. Further, the set threshold value is set to 2.

In this case, if the estimated gender is "male" and the age range is "40-60", then the number of people who are "male" and whose estimated age range is between 40 and 60 is 2 of person 1c and person 1g. Since it is a person, it is above the threshold. Therefore, the data generation unit 160 generates (age, gender, location, behavior, belongings) = (40-60, male, (x5, y5), standing, suitcase) and (age, gender, position, behavior, belongings) = (40-60, male, (x7, y7), sitting, no belongings) can be generated as the provided information.

Similarly, if the estimated gender is "female" and the age range is "20-40", then the number of people who are "female" and whose estimated age range is 20 to 40 is Person 1c, Person 1e, and Person 1e. Since there are 3 people in 1f, the threshold is exceeded. Therefore, the data generation unit 160 generates (20-40, female, (x3, y3), sitting, no belongings) and (age, gender, position, behavior, belongings) = (20-40, female, (x4, y4) ), sitting, no belongings) and (age, gender, location, behavior, belongings) = (20-40, female, (x6, y6), sitting, no belongings) can be generated as the provided information.

In this way, if the number of people for whom the same estimated age item matches the estimated gender is less than the threshold value from the setting unit 110, the age range of the estimated age is expanded and the number of people is aggregated, so that the number of people matches the threshold value. By generating provided information by securing the above number of people, it is possible to reduce the possibility that an individual will be identified and to generate provided information data that takes privacy into consideration.

In the above, by widening the age range of the estimated age and counting the number of people, we secured the number of people whose number is equal to or greater than the threshold value, but the present invention is not limited to this. For example, if there is no time constraint in the history information acquisition conditions, the provided information data may be generated by extending the time taken to trace the history information to ensure a number of people whose number is equal to or greater than a threshold value. In this case, this can be done by including information on the time period in which the provided information was generated.
In addition, even if there is a time restriction in the history information acquisition condition, if there is no restriction on the number of days or days of the week, you can increase the number of days that history information goes back and extend the time to ensure that the number of people exceeds the threshold and provide information. It doesn't matter if you generate data. In this case, it is possible to do this by including information on the number of days in which the provided information was generated.

The data output unit 170 acquires the provided information from the data generation unit 160 and outputs it to the outside. When the information processing device 100 is incorporated into the camera 10, the data output unit 170 outputs the provided information to a terminal connected to the camera 10. In this embodiment, since it is assumed that the camera 10 is connected to the edge terminal 20 by wire, the camera 10 transmits the provided information to the edge terminal 20.

In this embodiment, the edge terminal 20 uses the provided information acquired from the camera 10 to switch the content to be displayed on the monitor 30. For example, out of multiple contents, if there is content with a clear target, such as advertising content for cosmetics or advertising content for travel tours aimed at middle-aged people, advertising can be made more efficient by selecting and displaying the content based on the provided information. This shows that it is possible to improve

Next, the flow of operations of the information processing apparatus 100 according to this embodiment will be explained.
First, after powering on the information processing device 100, a threshold value is set in the setting unit 110. At this time, the threshold value may be read from a storage unit that stores threshold values set by the user in advance, or the threshold value may be set by the user after the information processing apparatus 100 is powered on. Also, based on the average value from the history information, a predetermined number is subtracted from the average value, and for example, if the average value is 20, a predetermined number of 3 is subtracted to set 17. However, it is desirable to set the threshold value to at least a value greater than 1 in order to reduce the probability of being identified. Here, the number of history information to be acquired from the estimated information storage section 140 is defined in advance.
As a result, when the user sets, the average value from historical information is displayed to the user and the user inputs the threshold value, thereby prompting the user to set the threshold value based on the average value. .

After the threshold value is set by the setting unit 110, the data acquisition unit 120 starts acquiring the camera image taken by the camera 10. The data acquisition unit 120 sends the acquired camera image to the estimation unit 130.

Next, the estimation unit 130 recognizes the person included in the camera image, and obtains estimation information that is the attributes and status of the recognized person. Upon acquiring the estimated information, the estimating section 130 sends it to the estimated information storage section 140 and the data generating section 160.

Next, the estimated information storage unit 140 stores the estimated information acquired from the estimating unit 130 into the history information table. In the history information table of FIG. 4, the history information is stored in descending order, but the history information is not limited to this, and may be stored in ascending order.

Next, when the data generation unit 160 acquires the estimated information from the estimation unit 130, it generates provided information. FIG. 7 is a flowchart showing a process of generating provided information in data generating section 160 according to the present embodiment. The data generating unit 160 acquires the estimated information and the number of people from the estimating unit 130 (step S101).However, the number of people may not be acquired from the estimating unit 130, but the number of people information may be obtained by totaling the estimated information. . After that, a threshold value is acquired from the setting unit 110 (step S102). Then, environment information is acquired from the data acquisition environment storage unit 150 (step S103). The order of step S101, step S102, and step S103 is not limited to this, and they may be acquired at the same time.

Next, the obtained number of people is compared with a threshold value (step S104). If the acquired number of people or the number of people whose items match the estimated information is equal to or greater than the threshold (YES in step S104), the process proceeds to step S105, and if it is less than the threshold (NO in step S104), the process proceeds to step S106. Proceed to processing.

If YES is determined in step S104, the data generation unit 160 generates provided information from the estimated information (step S105), and ends the data generation process.

If it is determined NO in step S104, the data generation unit 160 determines whether or not the history information acquisition condition in the environmental information is satisfied (step S106). If it is determined that the history information acquisition condition in the environmental information is satisfied (YES in step S106), the process proceeds to step S107; if it is determined that the condition is not satisfied (NO in step S106), the data generation process is ended. .

In the process of step S107, the data generation unit 160 acquires history information for a certain period of time from the estimated information storage unit 140. Here, assuming that the time period of the history information to be acquired is one hour, if the acquisition time in the estimated information acquired from the estimation unit 130 is 7 o'clock, the history information from 6 o'clock to 7 o'clock is acquired.

Then, the number of people in the history information acquired from the estimated information storage unit 140 is added to the number of people acquired from the estimation unit 130 (step S108), and the number of people is compared with the threshold value again (step S109). If the number of people is equal to or greater than the threshold (YES in step S109), the process advances to step S110. If NO in step S109, the process advances to step S111.

In the process of step S110, the data generation unit 160 generates provided information from the estimation information obtained from the estimation unit 130 and the history information obtained from the estimation information storage unit 140 in the same manner as in step S105 (step S110). After that, the data generation process ends.

In the process of step S111, the data generation unit 160 changes the time period for acquiring history information so as to acquire past history information than the history information acquired last time (step S111). For example, if history information from 6:00 to 7:00 was acquired last time, the time zone of the history information to be acquired is changed so that history information from 5:00 to 6:00 can be acquired. The process then proceeds to step S106 again.

FIG. 8 is a diagram showing the information processing device 100 using a processor. The processor 201 is connected to a memory 202, a key input/output interface (hereinafter referred to as I/F) 203, a data input/output I/F 204, and a display output I/F 205.

The processor 201 is hardware that operates when using the memory 202 to execute a program for executing the processing of the present disclosure. The key input/output I/F 203 is connected to a keyboard, a touch key device, etc., and is used when a threshold value is set by a user. It is also conceivable that the key input/output I/F 203 is used when presenting average value information to a user who sets a threshold value. The data input/output I/F 204 is used when connecting to a camera and acquiring image data. Further, the data input/output I/F 204 may be connected to an external storage device to access and record threshold values, history information, and environmental information in the external storage device (not shown). The display output I/F 205 is used to display provided information. If the configuration is as shown in FIG. 2, the display output I/F 205 is connected to the edge terminal 20 and outputs provided information to the edge terminal 20.

As described above, according to the information processing apparatus according to the first embodiment, data that takes privacy into account can be generated from information about a person included in a camera image.
Furthermore, by presenting the recommended threshold value to the user, it is possible to allow the user to set an appropriate threshold value for generating the provided information.

Furthermore, by defining the conditions for generating the provided information, it is possible to generate the provided information according to the situation in which the camera 10 is installed.

Note that in this embodiment, in order to make the explanation easy to understand, the explanation has been limited to train cars, but this is effective when controlling an external terminal with cameras installed in places other than train cars. Therefore, similar effects can be expected.

Furthermore, although the embodiments of the present invention have been described above, the present invention is not limited to these embodiments.

Reference Signs List 10 camera, 20 edge terminal, 30 monitor, 100 information processing device, 110 setting unit, 120 data acquisition unit, 130 estimation unit, 140 estimation information storage unit, 150 data acquisition environment storage unit, 160 data generation unit, 170 data output unit

Claims (10)

a data acquisition unit that acquires image data;
a setting section for setting a threshold;
an estimation unit that recognizes a person from the image data and estimates estimated information for each of the recognized persons;
a data generation unit that generates provided information from the estimated information based on a comparison between the aggregated result of the estimated information and the threshold;
and a data output unit that outputs the provided information,
The estimated information includes at least one item of information indicating attributes and conditions of the person recognized from the image data,
If, as a result of aggregating the estimated information, the number of people whose estimated information items match is less than the threshold, the data generation unit generates the provided information by excluding the estimated information items related to the person. An information processing device characterized by: a data acquisition unit that acquires image data;
a setting section for setting a threshold;
an estimation unit that recognizes a person from the image data and estimates estimated information for each of the recognized persons;
a data generation unit that generates provided information from the estimated information based on a comparison between the aggregated result of the estimated information and the threshold;
and a data output unit that outputs the provided information,
The estimated information includes at least an item related to age,
As a result of aggregating the estimated information, if the number of people estimated to be of the same age is less than the threshold, the data generation unit widens the age range and aggregates the number to make the number equal to or greater than the threshold. An information processing device that generates the provided information. comprising a storage unit that records the number of people newly recognized by the estimation unit, the estimated information for each estimated person, and acquisition time information of the image data as history information;
The information processing according to claim 1 or 2, wherein the data generation unit generates the provided information from a result of adding the history information to the estimated information of the newly recognized person and aggregating the result. Device. When the number of people obtained from the aggregation result of the estimated information is less than the threshold, the setting unit uses the history information to output the average value of the number of people in various time periods from the data output unit, so as to exceed the threshold. 4. The information processing apparatus according to claim 3 , wherein the information processing apparatus prompts for change. a data acquisition unit that acquires image data;
a setting section for setting a threshold;
an estimation unit that recognizes a person from the image data and estimates estimated information for each of the recognized persons;
a data generation unit that generates provided information from the estimated information based on a comparison between the aggregated result of the estimated information and the threshold;
and a data output unit that outputs the provided information,
The estimated information includes at least one attribute or state among one or more attributes and one or more states as information regarding the person recognized from the image data,
If the number of people whose estimated information items match is less than the threshold as a result of aggregating the estimated information, the data generation unit eliminates the estimated information items related to the person or adds the estimated information to the estimated information. An information processing apparatus characterized in that the information processing apparatus generates the provided information after expanding the items to make the number of people equal to or more than the threshold value. Obtaining information regarding the conditions that can be provided as the provided information,
The information processing apparatus according to claim 3 , wherein the data generation unit generates the provided information from among the estimated information that satisfies the condition among the history information. The conditions include time zone information,
The data generation unit generates the provided information from among the estimated information estimated within the time slot indicated by the time slot information among the history information,
If the number of people whose contents match the estimated information items in the time period is less than the threshold, the provision includes the estimated information estimated in the same time period on a different day from among the historical information. The information processing device according to claim 6 , wherein the information processing device generates information. The conditions include section information,
The storage unit records location information at which the newly recognized image data was acquired by the estimation unit, including it in history information;
The information processing apparatus according to claim 6 or 7 , wherein the data generation unit generates the provided information from among the estimated information that satisfies a condition of the section information among the history information. Information processing according to any one of claims 3 and 6 to 8 , characterized in that the average value of the number of people estimated for each time period is calculated from the historical information, and the result of the calculation is output. Device. An information processing method implemented by an information processing device, the method comprising:
a data acquisition step of acquiring image data;
a threshold setting step of setting a threshold;
an estimation step of recognizing a person from the image data and estimating estimated information for each of the recognized persons;
a data generation step of generating provided information from the estimated information based on a comparison between the aggregated result of the estimated information and the threshold;
an output step of outputting the provided information;
Equipped with
The estimated information includes at least one item of information indicating attributes and conditions of the person recognized from the image data,
In the data generation step, if the number of people whose estimated information items match is less than the threshold as a result of aggregating the estimated information, the provided information is generated by excluding the estimated information items related to the person. do
An information processing method characterized by:

Images