JP2024049181A - Information processor and information processing method - Google Patents

Abstract

To provide an information processor and information processing method that minimize erroneous determination in abnormality determination processing.SOLUTION: In an information processing system 1, an information processor 300 includes a detection unit, estimation unit, determination unit, and learning unit. The detection unit detects S3 a person from a shot image S2 taken S1 by a camera unit 30. The estimation unit estimates S4 a moving speed of the person detected by the detection unit. The determination unit determines S5 occurrence of an abnormality in a case where the moving speed estimated by the estimation unit exceeds a threshold. The learning unit learns S8 the threshold according to a result of determination by the determination unit and a result of verification of the shot image S2 by a user.SELECTED DRAWING: Figure 1

Description

Embodiments of the present invention relate to an information processing device and an information processing method.

Conventionally, there is known a technique for detecting a person from a captured image and detecting an abnormality based on the movement speed of the detected person. For example, there is known a technique for detecting an abnormality in an area captured in surveillance video when the movement speed of a person (e.g., a security guard) captured in the surveillance video exceeds a preset threshold.

JP 2021-87031 A

However, the conventional technology still had room for improvement in terms of reducing erroneous judgments in the anomaly judgment process.

The problem that the present invention aims to solve is to provide an information processing device and an information processing method that can reduce erroneous judgments in anomaly judgment processing.

The information processing device according to the embodiment includes a detection unit, an estimation unit, a determination unit, and a learning unit. The detection unit detects a person from an image captured by the imaging unit. The estimation unit estimates the movement speed of the person detected by the detection unit. The determination unit determines that an abnormality has occurred when the movement speed estimated by the estimation unit exceeds a threshold. The learning unit learns the threshold according to the determination result by the determination unit and the result of the user's confirmation of the captured image.

FIG. 1 is a diagram showing an overview of a process executed by an information processing apparatus according to an embodiment. FIG. 2 is a block diagram illustrating an example of the configuration of an information processing system according to the embodiment. FIG. 3 is a side view of the lighting device according to the embodiment. FIG. 4 is a bottom view of the lighting device according to the embodiment. FIG. 5 is a block diagram showing a configuration of a lighting device according to an embodiment. FIG. 6 is a block diagram showing the configuration of the information processing device according to the embodiment. FIG. 7 is a diagram illustrating an example of a tracking information storage unit according to the embodiment. FIG. 8 is a flowchart showing the procedure of the abnormality determination process executed by the information processing device according to the embodiment. FIG. 9 is a flowchart showing the procedure of the learning process executed by the information processing device according to the embodiment.

The information processing device 300 according to the embodiment described below includes a detection unit 122 that detects a person from an image captured by the imaging unit 31, an estimation unit 123 that estimates the movement speed of the person detected by the detection unit 122, a determination unit 126 that determines that an abnormality has occurred when the movement speed estimated by the estimation unit 123 exceeds a threshold, and a learning unit 128 that learns the threshold according to the determination result by the determination unit 126 and the result of the user checking the captured image.

The information processing device 300 according to the embodiment described below also includes a notification unit 127 that notifies the terminal device 400 used by the user of the determination result by the determination unit 126, and an acquisition unit 121 that acquires the confirmation result from the terminal device 400.

In addition, in the information processing device 300 according to the embodiment described below, the learning unit 128 learns a model that indicates the relationship between a situation, including at least one of the time period in which a person was detected, the number of people detected at the same time, and the moving speed estimated by the estimation unit, and a threshold value, based on the determination result and the confirmation result.

The information processing device 300 according to the embodiment described below also includes a situation identification unit 124 that identifies a situation, and a threshold determination unit 125 that determines a threshold from the situation identified by the situation identification unit 124 using a model learned by the learning unit 128.

The information processing method according to the embodiment described below includes a detection process for detecting a person from an image captured by the imaging unit 31, an estimation process for estimating the movement speed of the person detected by the detection process, a determination process for determining that an abnormality has occurred when the movement speed estimated by the estimation process exceeds a threshold, and a learning process for learning the threshold based on the determination result by the determination process and the result of the user checking the captured image.

Below, an information processing device and an information processing method according to an embodiment will be described with reference to the drawings. Note that in the embodiments described below, the same components are given the same reference numerals, and duplicated descriptions will be omitted.

<Information processing by information processing device 300>
First, an overview of the process executed by the information processing device 300 according to the embodiment will be described with reference to Fig. 1. Fig. 1 is a diagram showing an overview of the process executed by the information processing device 300 according to the embodiment.

The information processing system 1 according to the embodiment includes, for example, a lighting device 100, an information processing device 300, and a terminal device 400.

The lighting device 100 is, for example, a lighting device attached to the ceiling of a building such as a factory. The information processing device 300 is, for example, a cloud server. The terminal device 400 is, for example, a desktop PC (Personal Computer), a notebook PC, a smartphone, a tablet terminal, or a PDA (Personal Digital Assistant). The terminal device 400 is used, for example, by a user who manages the building in which the lighting device 100 is installed.

The lighting device 100 has a lighting unit 20 and a camera unit 30. The lighting device 100 can capture an image of an area illuminated by the lighting unit 20 using the camera unit 30. The lighting device 100 captures an image of the illumination area using the camera unit 30 (step S1), and transmits the captured image to the information processing device 300 (step S2).

The information processing device 300 detects the person TG from the captured image by performing a predetermined image processing on the captured image obtained from the lighting device 100 (step S3).

Next, the information processing device 300 estimates the movement speed of the detected person TG (step S4). For example, the information processing device 300 can track the person TG between multiple frames by identifying the person TG based on the feature amount of the person TG extracted from the captured image. The information processing device 300 can estimate the movement speed of the person TG based on, for example, the frame rate of the camera unit 30 and the movement distance of the person TG between multiple consecutive frames.

Next, the information processing device 300 performs an abnormality determination process using the moving speed of the person TG and a threshold value (step S5). Specifically, the information processing device 300 determines that an abnormality has occurred when the moving speed of the person TG exceeds the threshold value. Note that an abnormality is, for example, an abnormality that has occurred in a building in which the lighting device 100 is installed.

Here, depending on the situation, a person appearing in the captured image (e.g., a worker working in a factory) may move faster than normal even when no abnormality is occurring. For this reason, if a fixed single threshold is used when performing an abnormality determination based on the speed at which a person moves, there is a risk of an erroneous determination that an abnormality exists when in fact it is normal. In order to reduce such erroneous determinations, it is desirable to optimize the threshold used in the abnormality determination process according to the circumstances, such as the number of people detected and the time of day. Therefore, in the information processing device 300 according to the embodiment, the threshold is learned according to the determination result of the abnormality determination process and the result of the user's confirmation of the captured image. By optimizing the threshold in this way, it is possible to reduce erroneous determinations in the abnormality determination process.

The information processing device 300 transmits a confirmation request including the determination result of the abnormality determination process to the terminal device 400 (step S6). The confirmation request includes, for example, the determination result of the abnormality determination process, as well as information about the situation, such as the captured image used in the abnormality determination process, the date and time when the abnormality was detected, the number of people TG detected, and the movement speed of the people TG detected.

When the terminal device 400 obtains the result of the abnormality determination process from the information processing device 300, it transmits the result of the user's confirmation of the captured image to the information processing device 300 (step S7).

For example, when the terminal device 400 acquires a judgment result, it causes the display unit to display a confirmation screen including the judgment result (information that an abnormality has occurred), the captured image used in the abnormality judgment process, the date and time when the abnormality was detected, the number of people detected, the estimated moving speed, a selection button for the confirmation result, and the like. The user refers to the confirmation screen to check whether the judgment result of the abnormality judgment process matches the actual situation. In other words, the user checks whether the situation shown in the captured image is abnormal or not.

If the user determines that the determination result of the anomaly determination process matches the actual situation, the user selects, for example, the "Yes" button displayed on the display unit of the terminal device 400 using an input operation unit such as a mouse. Also, if the user determines that the determination result of the anomaly determination process does not match the actual situation, the user selects the "No" button displayed on the display unit of the terminal device 400 using the input operation unit. The terminal device 400 transmits the confirmation result according to the input operation by the user to the information processing device 300. The confirmation result includes information on whether the determination result of the anomaly determination process matches the actual situation, that is, information on whether the determination result of the anomaly determination process is correct or not.

Then, the information processing device 300 learns the threshold value to be used in the abnormality determination process based on the result of the abnormality determination process and the confirmation result obtained from the terminal device 400 (step S8).

For example, when the information processing device 300 obtains a confirmation result indicating that the result of the abnormality determination process does not match the actual situation, the information processing device 300 increases the value of the threshold value used in the abnormality determination process. This optimizes the threshold value, thereby reducing erroneous determinations in the abnormality determination process.

The information processing device 300 maintains the threshold value when it obtains a confirmation result from the terminal device 400 indicating that the determination result of the anomaly determination process matches the actual situation. Alternatively, the information processing device 300 may lower the threshold value when it obtains a confirmation result from the terminal device 400 indicating that the determination result of the anomaly determination process matches the actual situation.

Here, an example has been described in which the information processing device 300 judges that an abnormality has occurred, but the information processing device 300 may transmit the judgment result of the abnormality judgment process to the terminal device 400 even if it judges that there is no abnormality. In this case, the information processing device 300 lowers the threshold value when it acquires a confirmation result from the terminal device 400 indicating that the judgment result of the abnormality judgment process does not match the actual situation. Furthermore, the information processing device 300 maintains the threshold value when it acquires a confirmation result from the terminal device 400 indicating that the judgment result of the abnormality judgment process matches the actual situation. Alternatively, the information processing device 300 may raise the threshold value when it acquires a confirmation result from the terminal device 400 indicating that the judgment result of the abnormality judgment process matches the actual situation.

The information processing device 300 may store a model showing the relationship between a situation including at least one of the time period when a person was detected, the number of people detected at the same time, and the estimated moving speed, and a threshold value used in the abnormality determination process. In this case, the information processing device 300 may identify a situation including at least one of the time period when a person was detected, the number of people detected at the same time, and the moving speed estimated by the estimation unit, and output a threshold value using the above model from the identified situation. Then, the information processing device 300 may learn the above model according to the judgment result of the abnormality determination process and the confirmation result of the captured image by the user. For example, when the judgment result is "abnormal" and the confirmation result is "not abnormal," the information processing device 300 updates the parameters of the model so that the threshold value is higher. By learning the model in this way, the abnormality determination process can be performed using a threshold value that changes to an appropriate value depending on the situation, and this can reduce erroneous judgments in the abnormality determination process.

<Configuration example of information processing system 1>
Next, a configuration example of the information processing system 1 according to the embodiment will be described with reference to Fig. 2. Fig. 2 is a block diagram showing a configuration example of the information processing system 1 according to the embodiment.

As shown in FIG. 2, the information processing system 1 has a plurality of lighting devices 100, an information processing device 300, and a terminal device 400. The plurality of lighting devices 100, the information processing device 300, and the terminal device 400 are connected to each other so as to be able to communicate with each other via a network N such as the Internet.

The lighting device 100 is attached to the ceiling of a building such as a factory. As described above, the lighting device 100 has a camera unit 30 in addition to the lighting unit 20, and the area illuminated by the lighting unit 20 can be imaged by the camera unit 30. The configuration of the lighting device 100 will be described with reference to Figures 3 and 4.

Here, an example is shown in which the information processing system 1 includes multiple lighting devices 100, but the information processing system 1 only needs to include at least one lighting device 100.

The information processing device 300 acquires a captured image from the lighting device 100 via the network N, and performs a process of estimating the moving speed of a person and a process of determining an abnormality based on the acquired captured image. The information processing device 300 also transmits a determination result of the abnormality determination process to the terminal device 400 via the network N. The configuration of the information processing device 300 will be described with reference to FIG. 6.

When the terminal device 400 acquires the result of the abnormality determination process from the information processing device 300 via the network N, the terminal device 400 transmits the user's confirmation result of the image corresponding to the determination result to the information processing device 300 via the network N.

The configuration of the information processing system 1 is not limited to the example shown in FIG. 2. For example, the information processing system 1 may include a repeater in addition to the lighting device 100, the information processing device 300, and the terminal device 400. The repeater is a device that has a function of performing relay processing. For example, the repeater is an information processing device (computer) that functions as a so-called gateway. In this case, the lighting device 100 is connected to the repeater via a LAN (Local Area Network) or the like so as to be able to communicate with the information processing device 300 via the repeater in a wired or wireless manner, and communicates with the information processing device 300 via the repeater.

<Configuration example of lighting device 100>
Next, an example of the configuration of the lighting device 100 will be described with reference to Fig. 3 and Fig. 4. Fig. 3 is a side view of the lighting device 100 according to the embodiment. Fig. 4 is a bottom view of the lighting device 100 according to the embodiment.

As shown in FIG. 3, the lighting device 100 according to the embodiment includes a lighting unit 20, a camera unit 30, and a power supply unit 40. The lighting unit 20 and the camera unit 30 are screwed together via an L-shaped metal fitting, and the lighting unit 20 and the power supply unit 40 are connected together via a connecting portion 12.

The connecting portion 12 is further joined to the mounting portion 11, and the lighting device 100 is attached to an installation surface such as a ceiling via the mounting portion 11.

The lighting unit 20 includes a light source section 21 and fins 25. The fins 25 are provided on the upper part of the light source section 21 and cool the heat generated by the light emission of the light source section 21. As shown in FIG. 4, the light source section 21 includes a light emitting element 21b such as an LED (Light Emitting Diode) provided on a substrate 21a.

The light output surface of the light source unit 21 is covered by a cover (not shown). The cover is made of, for example, transparent glass or resin. In the example shown in FIG. 4, the light source unit 21 is composed of four substrates 21a, but is not limited to this.

The camera unit 30 includes an imaging section 31 and a lens 32. The imaging section 31 includes an imaging sensor such as a charge coupled device (CCD) or a complementary metal oxide semiconductor (CMOS), as well as various components required to capture images or videos. The bottom surface of the imaging section 31 is covered by the lens 32.

As shown in FIG. 4, the light source unit 21 and the imaging unit 31 face in the same direction when attached. This allows the imaging unit 31 to capture an image of the area illuminated by the light source unit 21, ensuring sufficient illuminance for the subject, allowing a clear image to be captured.

Since the lighting device 100 is installed on the ceiling, it is possible to capture an image of the interior of the space looking down from the ceiling. Note that the external shape of the lighting device 100 shown in Figures 3 and 4 is an example, and the lighting device 100 may have other shapes, such as a base light or a ceiling light.

<Example of functional configuration of lighting device 100>
Next, an example of a functional configuration of the lighting device 100 according to the embodiment will be described with reference to Fig. 5. Fig. 5 is a block diagram showing the configuration of the lighting device according to the embodiment. As shown in Fig. 5, the lighting device 100 has a communication unit 10, a lighting unit 20, a camera unit 30, and a control unit 50.

The communication unit 10 transmits and receives information to and from other devices such as the information processing device 300. The communication unit 10 is realized, for example, by a predetermined communication circuit or a NIC (Network Interface Card). The communication method used by the communication unit 10 may be, for example, wireless communication such as a local area network (LAN). Note that the communication method used by the communication unit 10 may be any communication method, for example, wired communication, etc.

The camera unit 30 has an imaging section 31, an imaging control section 33, and a storage section 34. The imaging section 31 captures an image of a specific angle of view (hereinafter referred to as the "imaging area") to generate image data (also referred to as the "captured image"). The image data may be video data or still image data. The image data may include the imaging date and time (date, day of the week, time, etc.), frame rate, identification information of the imaging section 31, etc.

The imaging control unit 33 controls the imaging by the imaging unit 31. The imaging control unit 33 also outputs image data generated by the imaging unit 31 to the control unit 50.

The storage unit 34 stores programs and data for implementing various controls of the camera unit 30. The storage unit 34 may also store image data generated by the imaging unit 31. The storage unit 34 is implemented, for example, by a semiconductor memory element such as a random access memory (RAM) or a flash memory, or a storage device such as a hard disk or an optical disk.

The control unit 50 is realized, for example, by a CPU (Central Processing Unit) or MPU (Micro Processing Unit) executing various programs stored in the internal memory using the RAM as a working area. The control unit 50 may be realized, for example, by an integrated circuit such as an ASIC (Application Specific Integrated Circuit) or an FPGA (Field Programmable Gate Array). The control unit 50 has an acquisition unit 51 and a transmission processing unit 52, and realizes or executes the functions and actions of the information processing described below. Note that the internal configuration of the control unit 50 is not limited to the configuration shown in FIG. 5, and may be other configurations as long as they perform the information processing described below.

The acquisition unit 51 acquires various information. For example, the acquisition unit 51 acquires various information from the camera unit 30. The transmission processing unit 52 performs a process of transmitting various information to an external device via the communication unit 10. For example, the transmission processing unit 52 performs a process of transmitting image data acquired by the acquisition unit 51 to the information processing device 300 via the communication unit 10 and the network N.

<Example of functional configuration of information processing device 300>
Next, a configuration example of the information processing device 300 according to the embodiment will be described with reference to Fig. 6. Fig. 6 is a block diagram showing the configuration of the information processing device 300 according to the embodiment. As shown in Fig. 6, the information processing device 300 includes a communication unit 110, a control unit 120, and a storage unit 130.

The communication unit 110 is realized, for example, by a predetermined communication circuit or a NIC. The communication method by the communication unit 110 may be, for example, wireless communication such as a local area network (LAN).

The storage unit 130 is realized, for example, by a semiconductor memory element such as a RAM or a flash memory, or a storage device such as a hard disk or an optical disk. The storage unit 130 has an image storage unit 131, a tracking information storage unit 132, and a model storage unit 133.

The image storage unit 131 is a database that stores images captured by the imaging unit 31. The image storage unit 131 stores, for example, pixel data, image capture date and time (date, day of the week, time, etc.), frame rate, and identification information of the imaging unit 31, which are associated with each other.

The tracking information storage unit 132 is a database that stores tracking information of a person generated by the detection unit 122, which will be described later. FIG. 7 is a diagram showing an example of the tracking information storage unit 132 according to an embodiment.

In the example shown in FIG. 7, the tracking information storage unit 132 stores, for example, a "person ID" and "tracking information" in association with each other. The "person ID" is identification information that identifies a person detected by the detection unit 122, which will be described later.

"Tracking information" has items such as "date", "day of the week", "time", "location", and "movement speed". The "date" item stores the date on which the person was detected, for example, the year, month, and day. The "day of the week" item stores the day of the week on which the person was detected. The "time" item stores the time on which the person was detected. The "location" item stores information on the detected camera unit 30 (for example, information on the installation location of the camera unit 30 and identification information of the camera unit 30). The "movement speed" item stores the movement speed estimated by the estimation unit 123, which will be described later.

The model storage unit 133 stores a model that indicates the relationship between information about various conditions when the captured image was captured and a threshold value used by the determination unit 126, which will be described later. Specifically, this model is a model that calculates the value of the threshold value used by the determination unit 126, which will be described later, when information about various conditions when the captured image was captured is input.

Returning to the explanation of FIG. 6, the control unit 120 will now be described. The control unit 120 is realized, for example, by a CPU, an MPU, or the like, executing various programs stored in an internal memory using a RAM as a working area. The control unit 120 may also be realized, for example, by an integrated circuit such as an ASIC or an FPGA.

The control unit 120 includes an acquisition unit 121, a detection unit 122, an estimation unit 123, a situation identification unit 124, a threshold determination unit 125, a judgment unit 126, a notification unit 127, and a learning unit 128.

The acquisition unit 121 acquires an image from the lighting device 100. The acquisition unit 121 stores the acquired image in the image storage unit 131. The image storage unit 131 stores, for example, pixel data, image capture date and time (date, day of the week, time, etc.), frame rate, identification information of the image capture unit 31, etc., in association with one another.

In addition, the acquisition unit 121 acquires the result of the user's confirmation of the captured image from the terminal device 400. The acquisition unit 121 passes the acquired confirmation result to the learning unit 128.

The detection unit 122 detects a person from the captured image by performing image analysis on the captured image stored in the image storage unit 131. The image analysis may be performed, for example, by using a detection model. The detection model is obtained by performing learning using a machine learning algorithm such as a deep neural network (DNN). The detection model may be configured to use a machine learning algorithm such as a support vector machine (SVM) using a histogram of gradient (HOG) feature. The detection unit 122 may also be configured to perform various detections using, for example, template matching, without using a detection model that has been subjected to machine learning. The detection unit 122 performs detection processing, for example, for each frame.

The detection unit 122 also tracks the detected person. For example, the detection unit 122 generates feature information of the detected person by extracting features from the captured image and converting the extracted features into a multidimensional feature vector. The detection unit 122 can identify the person using this feature information and track the person across multiple frames.

The detection unit 122 stores tracking information including, for example, date, day of the week, time, and location, for each person ID, which is the identification information of a person, in the tracking information storage unit 132.

The estimation unit 123 estimates the moving speed of the person detected by the detection unit 122 based on the tracking information. For example, the estimation unit 123 calculates the moving distance of the person between multiple consecutive frames. Then, the estimation unit 123 estimates the moving speed of the person based on the calculated moving distance and the frame rate of the camera unit 30. The estimation unit 123 adds the estimated moving speed to the tracking information storage unit 132.

The situation identification unit 124 identifies the situation when a person is detected. For example, the situation identification unit 124 identifies the time period, day of the week, number of people detected at the same time, and moving speed when the person is detected from the tracking information storage unit 132.

The threshold determination unit 125 uses the model stored in the model storage unit 133 to output thresholds based on the various situations identified by the situation identification unit 124. The threshold determination unit 125 determines the thresholds output from the model as the thresholds to be used in the abnormality determination process by the determination unit 126.

For example, the model stored in the model storage unit 133 may be expressed as TH=f(x). Here, x is information about the situation when a person is detected, and is a feature of the model. The model may be a linear or nonlinear regression model. For example, the model may be a regression model with x as an explanatory variable and TH (threshold) as a target variable. As an example, the model may be a regression model such as TH=ax ₁ +bx ₂ +cx ₃ +dx ₄ .... Here, x _n (n is a positive integer) is a feature. For example, x ₁ , x ₂ , x ₃ , and x ₄ may be feature amounts corresponding to the time period, the day of the week, the number of people detected, and the moving speed of the people, respectively. a to d are coefficients of x ₁ , x ₂ , x ₃ , and x ₄ , and are parameters obtained by a learning process by the learning unit 128 described later.

The threshold determination unit 125 inputs the features of the various situations identified by the situation identification unit 124 into the model and outputs TH (threshold value). The threshold determination unit 125 passes the output threshold value to the judgment unit 126.

The determination unit 126 performs an abnormality determination process using the moving speed estimated by the estimation unit 123 and the threshold value acquired from the threshold value determination unit 125. Specifically, the determination unit 126 compares the moving speed with the threshold value, and determines that an abnormality has occurred when the value of the moving speed is greater than the threshold value.

The processing by the estimation unit 123, the situation identification unit 124, the threshold determination unit 125, and the judgment unit 126 may be performed, for example, every unit time, in other words, every multiple frames. For example, assuming that the frame rate of the camera unit 30 is 30 fps, the processing by the estimation unit 123, the situation identification unit 124, the threshold determination unit 125, and the judgment unit 126 may be performed, for example, every second, in other words, every 30 frames. Furthermore, without being limited to this, the processing by the estimation unit 123, the situation identification unit 124, the threshold determination unit 125, and the judgment unit 126 may be performed for each frame.

The notification unit 127 transmits the result of the determination by the determination unit 126 to the terminal device 400 via the communication unit 110 and the network N. For example, when the determination unit 126 determines that an abnormality has occurred, the notification unit 127 notifies a confirmation request including information that an abnormality has occurred, the captured image used in the abnormality determination process, and the situation when the abnormality was detected.

When the learning unit 128 acquires a confirmation result from the terminal device 400, the learning unit 128 learns a threshold value according to the acquired confirmation result and the judgment result by the judgment unit 126. Specifically, the learning unit 128 learns a model indicating the relationship between the threshold value and a situation including the time period in which a person was detected, the number of people detected at the same time, the moving speed estimated by the estimation unit 123, etc., according to the acquired confirmation result and the judgment result by the judgment unit 126.

For example, if the judgment result is "abnormal" and the confirmation result is "not abnormal", the learning unit 128 updates the model parameters so that the threshold value becomes higher. Also, if the judgment result is "abnormal" and the confirmation result is also "abnormal", the learning unit 128 does not update the parameters. Note that, without being limited to this, the learning unit 128 may update the model parameters so that the threshold value becomes lower when the judgment result is "abnormal" and the confirmation result is also "abnormal".

The confirmation result may include the correct value of the threshold. The correct value of the threshold is input, for example, by the user of the terminal device 400. In this case, the learning unit 128 may update the parameters of the model so that the difference between the threshold output by the model and the correct value of the threshold included in the confirmation result becomes smaller.

Here, when the judgment unit 126 judges an abnormality, the user is requested to confirm the judgment result, but the timing of the user's confirmation does not necessarily have to be the timing when the abnormality is judged. For example, the information processing device 300 may store the judgment result by the judgment unit 126 (including the judgment result of "not abnormal") in the storage unit 130 together with the situation when the judgment process was performed. In this case, the terminal device 400 may acquire the information stored in the storage unit 130, and the user may confirm whether the accumulated information is "abnormal" or "not abnormal", and the learning unit 128 may learn the threshold value based on the confirmation result. Note that if the judgment result is "not abnormal" and the confirmation result is "abnormal", the learning unit 128 updates the model parameters so that the threshold value is lower. Also, if the judgment result is "not abnormal" and the confirmation result is also "not abnormal", the learning unit 128 does not update the parameters. However, without being limited to this, the learning unit 128 may update the model parameters so that the threshold value becomes higher when the judgment result is "not abnormal" and the confirmation result is also "not abnormal."

<Specific Operation of Information Processing Device 300>
Next, a process procedure executed by the information processing device 300 according to the embodiment will be described with reference to Fig. 8 and Fig. 9. First, a process procedure of an abnormality determination process executed by the information processing device 300 according to the embodiment will be described with reference to Fig. 8. Fig. 8 is a flowchart showing a process procedure of an abnormality determination process executed by the information processing device 300 according to the embodiment.

As shown in FIG. 8, the information processing device 300 acquires a captured image from the lighting device 100 (step S101), detects a person from the acquired captured image, and tracks the detected person (step S102). Then, the information processing device 300 estimates the moving speed of the person based on the tracking information of the person (step S103).

Next, the information processing device 300 identifies the situation when the person was detected based on the tracking information (step S104), and inputs the identified situation into the model as a feature, thereby outputting a threshold value according to the situation (step S105).

The information processing device 300 then determines whether the movement speed estimated in step S103 exceeds the threshold value output from the model in step S105 (step S106). In this process, if it is determined that the movement speed exceeds the threshold value (step S106, Yes), the information processing device 300 notifies the terminal device 400 of an abnormality (step S107). Specifically, the information processing device 300 notifies a confirmation request including information that an abnormality has been detected, the captured image used in the abnormality determination process, and the situation when the abnormality was detected.

When the processing of step S107 is completed, or when the moving speed does not exceed the threshold value in step S106 (step S106, No), the information processing device 300 ends the processing.

Next, the procedure of the learning process executed by the information processing device 300 according to the embodiment will be described with reference to FIG. 9. FIG. 9 is a flowchart showing the procedure of the learning process executed by the information processing device 300 according to the embodiment.

As shown in FIG. 9, the information processing device 300 acquires the confirmation result from the terminal device 400 (step S201). Then, the information processing device 300 updates the model parameters based on the judgment result in step S106 and the confirmation result acquired in step S201 (step S202).

Other Embodiments
In the above-described embodiment, an example has been described in which the learning unit 128 learns a model indicating a relationship between a threshold and a situation including a time period when a person is detected, a day of the week, the number of people detected at the same time, and a moving speed estimated by the estimation unit 123. Without being limited to this, the learning unit 128 may learn a model indicating a relationship between a threshold and a situation including at least one of a time period when a person is detected, a day of the week, the number of people detected at the same time, and a moving speed estimated by the estimation unit 123. In other words, the model stored in the model storage unit 133 may be a model indicating a relationship between a threshold and a situation including at least one of a time period when a person is detected, the number of people detected at the same time, and a moving speed estimated by the estimation unit 123.

As described above, the information processing device 300 according to the embodiment includes a detection unit 122, an estimation unit 123, a determination unit 126, and a learning unit 128. The detection unit 122 detects a person from an image captured by the imaging unit 31. The estimation unit 123 estimates the movement speed of the person detected by the detection unit 122. The determination unit 126 determines that an abnormality has occurred when the movement speed estimated by the estimation unit 123 exceeds a threshold. The learning unit 128 learns the threshold according to the determination result by the determination unit 126 and the result of the user checking the captured image.

Therefore, the information processing device 300 according to the embodiment can reduce erroneous determinations in the anomaly determination process.

The information processing device 300 according to the embodiment may include a notification unit 127 and an acquisition unit 121. The notification unit 127 notifies the terminal device 400 used by the user of the determination result by the determination unit 126. The acquisition unit 121 acquires the confirmation result from the terminal device 400. With this configuration, it is possible to collect the confirmation results necessary for learning the threshold value.

In the information processing device 300 according to the embodiment, the learning unit 128 may learn a model that indicates the relationship between a situation, including at least one of the time period in which a person was detected, the number of people detected at the same time, and the moving speed estimated by the estimation unit, and a threshold value, based on the determination result and the confirmation result. With this configuration, it is possible to generate a model that outputs a threshold value that is optimized according to the situation.

The information processing device 300 according to the embodiment may include a situation identification unit 124 and a threshold determination unit 125. The situation identification unit 124 identifies a situation. The threshold determination unit 125 determines a threshold from the situation identified by the situation identification unit 124 using a model learned by the learning unit 128. With this configuration, it is possible to perform an abnormality determination process using a threshold optimized according to the situation, thereby reducing erroneous determinations in the abnormality determination process.

Although an embodiment of the present invention has been described, this embodiment is presented as an example and is not intended to limit the scope of the invention. This embodiment can be implemented in various other forms, and various omissions, substitutions, and modifications can be made without departing from the gist of the invention. This embodiment and its variations are within the scope of the invention and its equivalents as described in the claims, as well as the scope and gist of the invention.

20 Lighting unit 30 Camera unit 31 Imaging unit 100 Lighting device 110 Communication unit 120 Control unit 121 Acquisition unit 122 Detection unit 123 Estimation unit 124 Situation identification unit 125 Threshold determination unit 126 Determination unit 127 Notification unit 128 Learning unit 130 Storage unit 131 Image storage unit 132 Tracking information storage unit 133 Model storage unit 300 Information processing device 400 Terminal device

Claims (5)

a detection unit that detects a person from an image captured by the imaging unit;
an estimation unit that estimates a moving speed of the person detected by the detection unit;
a determination unit that determines that an abnormality has occurred when the moving speed estimated by the estimation unit exceeds a threshold;
a learning unit that learns the threshold value in accordance with a result of the determination by the determination unit and a result of confirmation of the captured image by a user;
An information processing device comprising: a notification unit that notifies a terminal device used by the user of a result of the determination made by the determination unit;
an acquisition unit that acquires the confirmation result from the terminal device;
The information processing apparatus according to claim 1 , further comprising: The information processing device according to claim 1, wherein the learning unit learns a model that indicates a relationship between the threshold value and a situation including at least one of the time period in which the person was detected, the number of people detected at the same time, and the moving speed estimated by the estimation unit, in accordance with the determination result and the confirmation result. A situation identification unit that identifies the situation;
a threshold determination unit that determines the threshold from the situation identified by the situation identification unit using the model trained by the learning unit;
The information processing apparatus according to claim 3 , further comprising: a detection step of detecting a person from an image captured by the imaging unit;
an estimation step of estimating a moving speed of the person detected by the detection step;
a determination step of determining that an abnormality has occurred when the moving speed estimated by the estimation step exceeds a threshold value;
a learning step of learning the threshold value in accordance with a result of the determination step and a confirmation result of the captured image by a user;
An information processing method comprising:

Images