JP2016208408A - Detection method, detection device and control method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a detection method etc. capable of stably detecting an object in an infrared ray image.SOLUTION: The detection method includes the steps of: acquiring a piece of information relevant to an environmental temperature as a temperature in a space where an infrared ray image is acquired and a piece of specific information for identifying a temperature of the object (S22, S23); identifying a reference model from plural reference models based on the acquired specific information and environmental temperature (S24); and detecting the object in the infrared ray image using the specified one reference model (S25).SELECTED DRAWING: Figure 11

Description

  The present invention relates to a detection method and a detection device for detecting an object in an infrared image, and a control method using the same.

  In recent years, research on machine learning has progressed, and it has reached a stage where it can be applied to image recognition of images taken with a visible camera. Specifically, contour features are statistically calculated by machine learning of a large number of images taken by a visible camera. By using this contour feature, an object in the image can be detected with high accuracy.

  However, there are objects that cannot be directly recognized and recognized in an image taken with a visible camera.

  Therefore, a technique for recognizing an object that cannot be recognized by an image taken by a visible camera using an infrared image has been proposed (for example, Patent Document 1). Patent Document 1 discloses an image processing apparatus that detects an object (object) in an infrared image. Patent Document 1 discloses a technique for detecting and processing an object in an infrared image after giving a predetermined temperature to the object, such as a vehicle muffler region of 100 ° C. and a person region of 30 ° C.

JP 2013-0240404 A

"Pedestrian Detection with Unsupervised Multi-Stage Feature Learning", Pierre Sermanet, Koray Kavukcuoglu, Soumith Chintala, Yann LeCun, CVPR2013

  However, for example, an object such as a person has a problem that when the ambient temperature (environmental temperature) changes and the surface temperature changes, the image processing method described in Patent Document 1 makes it difficult to detect the object.

  The present invention focuses on the above-described problems, and an object of the present invention is to provide a detection method, a detection device, and a control method using the same that can stably detect an object in an infrared image. .

  In order to achieve the above object, a detection method according to an aspect of the present invention is a detection method in which a computer detects a predetermined object in an infrared image, and (i) a space in which the infrared image is acquired. Based on the acquisition step of acquiring information on the environmental temperature, which is a temperature, and (ii) specific information for specifying the temperature of the object, and among the plurality of reference models based on the specific information and the environmental temperature And a detecting step of detecting the object in the infrared image using the specified reference model.

  These general or specific aspects may be realized by a system, a method, an integrated circuit, a computer program, or a computer-readable recording medium such as a CD-ROM. The system, method, integrated circuit, computer You may implement | achieve with arbitrary combinations of a program and a recording medium.

  ADVANTAGE OF THE INVENTION According to this invention, the detection method etc. which can detect the object in an infrared image stably can be provided.

FIG. 1 is a diagram illustrating an example of an infrared image. FIG. 2 is a diagram for explaining the problem. FIG. 3 is a diagram illustrating an example of a configuration of the detection device according to the first embodiment. FIG. 4 is a diagram illustrating an example of the configuration of the storage unit in the first embodiment. FIG. 5 is a diagram illustrating an example of a detailed configuration of the acquisition unit according to the first embodiment. FIG. 6 is a diagram illustrating an example of a detailed configuration of the detection processing unit in the first embodiment. FIG. 7 is a diagram illustrating an example of a detailed configuration of the specifying unit illustrated in FIG. 5. FIG. 8 is a diagram for explaining an example of the time-temperature conversion table in the disaster case in the first embodiment. FIG. 9 is a diagram illustrating an example of a plurality of reference models in the first embodiment. FIG. 10 is a flowchart showing an outline of the operation of the detection apparatus according to the first embodiment. FIG. 11 is a flowchart illustrating an example of details of the process of S22 of FIG. FIG. 12 is a flowchart illustrating an example of details of the processing in S24 of FIG. FIG. 13 is a flowchart showing an example of details of the processing in S25 of FIG. FIG. 14 is a diagram for explaining an example of the effect of the first embodiment. FIG. 15 is a diagram for explaining an example of the effect of the first embodiment. FIG. 16 is a diagram illustrating an example of the configuration of the machine learning device. FIG. 17 is a diagram illustrating an example of a data set of learning results of the machine learning device illustrated in FIG. FIG. 18 is a diagram illustrating an example of processing of the machine learning unit 21 illustrated in FIG. FIG. 19 is a diagram illustrating an example of the configuration of the detection device according to the second embodiment. FIG. 20 is a diagram illustrating an example of a configuration of a storage unit in the second embodiment. FIG. 21 is a diagram illustrating an example of a detailed configuration of the acquisition unit according to the second embodiment. FIG. 22 is a diagram illustrating an example of a detailed configuration of the specifying unit according to the second embodiment. FIG. 23 is a diagram for explaining an example of a time-temperature conversion table in the bath-up example in the second embodiment. FIG. 24 is a diagram illustrating an example of a plurality of reference models in the second embodiment. FIG. 25 is a flowchart showing an outline of the operation of the detection apparatus according to the first embodiment. FIG. 26 is a flowchart illustrating an example of details of the processing of S23 in the second embodiment. FIG. 27 is a diagram illustrating an example of the configuration of the detection device according to the third embodiment. FIG. 28 is a diagram illustrating an example of a configuration of a storage unit in the third embodiment. FIG. 29 is a diagram illustrating an example of a detailed configuration of an acquisition unit according to Embodiment 3. FIG. 30 is a diagram illustrating an example of a detailed configuration of the detection processing unit in the third embodiment. FIG. 31 is a diagram illustrating an example of a detailed configuration of the specifying unit illustrated in FIG. 30. FIG. 32 is a diagram illustrating an example of a relationship between time and body temperature during exercise of a person in the third embodiment. FIG. 33 is a flowchart illustrating an outline of the operation of the detection device according to the third embodiment. FIG. 34 is a diagram illustrating an example of a configuration of an air conditioner according to Embodiment 4. FIG. 35 is a diagram illustrating an example of a state in which the air conditioner of FIG. 34 is installed. FIG. 36 is a flowchart showing an outline of the operation of the air conditioner according to the fourth embodiment.

(Knowledge that became the basis of the present invention)
When image recognition is performed using an image taken by a visible camera, an object that cannot be directly seen by a shield or the like cannot be recognized. On the other hand, when performing image recognition using an infrared image, there is a possibility that even an object that has a shielding object and cannot be directly seen can be recognized. Here, for example, as shown in FIG. 1, the infrared image is an image composed of a plurality of pixels indicating the temperature portion distribution of the target space. FIG. 1 is a diagram illustrating an example of an infrared image.

  In the technique disclosed in Patent Document 1, an object in an infrared image is detected and processed after a predetermined temperature is applied to the object, such as a vehicle muffler region of 100 ° C. and a person region of 30 ° C. Is disclosed. Thereby, there is an effect that the person can be recognized even when the person cannot be directly seen at night.

  However, for example, an object such as a person has a problem that when the ambient temperature (environmental temperature) changes and the surface temperature changes, the image processing method described in Patent Document 1 makes it difficult to detect the object.

  For example, in order to detect an object in an infrared image, a contour feature calculated from the temperature difference between the ambient temperature (environment temperature) and the object is used. However, even if the ambient temperature is constant, if the human body temperature (the surface temperature of the object) changes, the region of the region where the difference from the environmental temperature is small increases. That is, as shown in FIG. 2, there is a problem that the contour feature changes in accordance with the change in the human body temperature. Therefore, the image processing method described in Patent Document 1 has a problem that an object cannot be stably detected unless the condition that the environment temperature and the object temperature are constant is satisfied.

  Therefore, an object of the present invention is to provide a detection method that can stably detect an object in an infrared image even when the environmental temperature and the temperature of the object fluctuate.

  A detection method according to an aspect of the present invention is a detection method in which a computer detects a predetermined object in an infrared image, and (i) information on environmental temperature that is a temperature of a space from which the infrared image is acquired; (Ii) specifying one reference model from among a plurality of reference models based on the acquisition step for acquiring specific information for specifying the temperature of the object, and the specifying information and the environmental temperature And a detection step of detecting the object in the infrared image using the specified one reference model.

  According to this, even when the environmental temperature and the temperature of the object fluctuate, it is possible to realize a detection method that can stably detect the object in the infrared image.

  Here, for example, each of the plurality of reference models defines a shape including pixels indicating the object in a combination of the environmental temperature and the temperature of the object, and the shape includes the environmental temperature and the temperature of the object. The shape may be smaller as the environmental temperature is closer to the object temperature.

  Further, for example, in the detection step, when there is a region similar to the shape and temperature indicated by the one reference model in the infrared image, it is determined that the object exists in the region, and the infrared image The object may be detected by acquiring information on the coordinate position of the region as the position of the object.

  Further, for example, the specific information is an elapsed time from a specific event, and in the specific step, the elapsed time is converted into a temperature of the object, and based on the environmental temperature and the converted temperature of the object. Thus, the one reference model may be specified.

  Here, for example, the specific event is a disaster, the elapsed time is a time elapsed from the time when the disaster occurred, and the time when the disaster occurred is included in the computer or in an external device. The disaster detection device is the time when the disaster was detected, and in the acquisition step, the computer acquires the time when the disaster was detected from the disaster detection device via a network or a physical connection cable. The elapsed time may be acquired.

  Further, for example, the disaster detection device may detect the disaster when detecting a vibration of a predetermined threshold value or more.

  Here, for example, the object is a person, the specific event is that the person has gone out of the bath, and the elapsed time is a time that has elapsed since the time at which the person got out of the bath, and the obtaining step Then, the computer may obtain the elapsed time by obtaining the time when the person has taken the bath from the bath control system via a network or a physical connection cable.

  Further, for example, the specific information is a temperature of the object, and the temperature of the object is measured by a temperature measuring device attached to the object. In the obtaining step, the computer calculates the temperature of the object, It may be acquired from the temperature measurement device via a network or a physical connection cable.

  The control method for an air conditioner according to an aspect of the present invention is a control method for controlling at least one of a set temperature, a wind direction, and an air volume of the air conditioner, and the detection according to claim 7. In the detection step, the computer executing the method detects the object in the infrared image acquired in the space that the air conditioner is subject to air conditioning, and the control method is detected by the computer. And a control step of controlling the air conditioner according to the position and temperature of the object.

  Moreover, the detection apparatus of the air conditioner according to one aspect of the present invention includes an infrared image acquisition unit that acquires an infrared image, an environmental temperature acquisition unit that acquires an environmental temperature that is the temperature of the space from which the infrared image is acquired, A specific information acquisition unit that acquires specific information for specifying a temperature of a predetermined object in the infrared image; and a detection processing unit that detects the object in the infrared image, the detection processing unit Identifies one reference model from among the plurality of reference models based on the identification information and the environmental temperature, and uses the identified one reference model to identify the object in the infrared image. To detect.

  These general or specific aspects may be realized by a system, a method, an integrated circuit, a computer program, or a computer-readable recording medium such as a CD-ROM. The system, method, integrated circuit, computer You may implement | achieve with arbitrary combinations of a program or a recording medium.

  Hereinafter, an infrared detection device and the like according to one embodiment of the present invention will be specifically described with reference to the drawings. Note that each of the embodiments described below shows a specific example of the present invention. Numerical values, shapes, materials, constituent elements, arrangement positions of constituent elements, and the like shown in the following embodiments are merely examples, and are not intended to limit the present invention. In addition, among the constituent elements in the following embodiments, constituent elements that are not described in the independent claims indicating the highest concept are described as optional constituent elements.

(Embodiment 1)
[Configuration of detection device]
Hereinafter, the detection apparatus according to Embodiment 1 will be described with reference to the drawings.

  FIG. 3 is a diagram illustrating an example of a configuration of the detection device according to the first embodiment. FIG. 4 is a diagram illustrating an example of the configuration of the storage unit in the first embodiment.

  As shown in FIG. 3, the detection device 100 includes a detection unit 1, an infrared sensor 2, a temperature sensor 3, a time measuring device 4, a storage unit 5, and an output unit 6.

  The infrared sensor 2 is, for example, a thermography attached to the detection device 100, and acquires an infrared image (thermal image) indicating a temperature distribution in the target space. The infrared sensor 2 transmits the acquired infrared image to the detection unit 1.

  Here, the infrared sensor 2 includes, for example, infrared detection elements (pixels) arranged two-dimensionally in a matrix, and acquires a two-dimensional infrared image at a time. Note that the infrared sensor 2 may be constituted by, for example, a line sensor, that is, an infrared detection element (pixel) arranged one-dimensionally, and may acquire a two-dimensional infrared image by scanning one-dimensionally. That is, the infrared sensor 2 may have any configuration as long as it can acquire a two-dimensional infrared image.

  The temperature sensor 3 is a thermistor or a thermocouple, for example, and can measure the temperature at one point in the target space or one point on the member surface. In the present embodiment, when the infrared sensor 2 acquires an infrared image, the temperature sensor 3 measures the environmental temperature that is the temperature of the space (target space) from which the infrared image is acquired, and transmits it to the detection unit 1. .

  The time measuring device 4 measures time. In the present embodiment, the time measuring device 4 measures the time when the infrared sensor 2 acquires the infrared image, and transmits it to the detection unit 1.

  As shown in FIG. 4, the storage unit 5 stores a time-temperature conversion table 51 and a learning result related information DB 52. The time-temperature conversion table 51 is a table used for converting time information into temperature information of an object when the specific information is information about time (time information). The learning result related information DB 52 is a database in which a collection of learning result data by machine learning is held. In the present embodiment, the learning result related information DB 52 holds a plurality of reference models that define the shape composed of pixels indicating the object in each combination of the environmental temperature and the temperature of the object. The shapes defined by the plurality of reference models differ depending on the combination of the environmental temperature and the object temperature. The shape becomes smaller as the environmental temperature and the temperature of the object are closer.

  Note that the plurality of reference models stored in the learning result related information DB 52 may not be data of learning results by machine learning. For example, an existing template owned by an intellectual property department or the like using the detection apparatus 100 may be used, or a template generated by a method disclosed in Non-Patent Document 1 may be used.

  The detection unit 1 detects an object in the infrared image. In the present embodiment, the detection unit 1 includes an acquisition unit 11 and a detection processing unit 12. The acquisition unit 11 acquires an infrared image from the infrared sensor 2, and the detection processing unit 12 detects an object such as a person from the acquired infrared image.

  Hereinafter, a detailed configuration of the detection unit 1 in the present embodiment will be described.

[Detailed configuration of acquisition unit]
FIG. 5 is a diagram illustrating an example of a detailed configuration of the acquisition unit according to the first embodiment.

  As illustrated in FIG. 5, the acquisition unit 11 includes an infrared image acquisition unit 111, a specific information acquisition unit 112, and an environmental temperature acquisition unit 113.

  The infrared image acquisition unit 111 acquires an infrared image. In the present embodiment, an infrared image of the target space is acquired from the infrared sensor 2.

  The environmental temperature acquisition unit 113 acquires information on the environmental temperature that is the temperature of the space from which the infrared image is acquired. In the present embodiment, the environmental temperature acquisition unit 113 acquires the environmental temperature of the target space measured by the temperature sensor 3. The environmental temperature acquisition unit 113 is not limited to the one measured by the temperature sensor 3, and information obtained by measuring the environmental temperature of the target space with a thermometer or the like is input to the detection apparatus 100 via a network or the like. You may acquire by.

  The specific information acquisition unit 112 acquires specific information for specifying the temperature of a predetermined object in the infrared image. In the present embodiment, the specific information is an elapsed time from a specific event. In the following description, it is assumed that the specific event is a disaster and the elapsed time is a time elapsed from the time when the disaster occurred.

  The specific information acquisition unit 112 calculates and acquires the elapsed time from the time when the disaster occurs and the time measured by the time measuring device 4. For example, the specific information acquisition unit 112 may acquire the time when the disaster detection device detects the disaster as the time when the disaster occurred. Here, the disaster detection device detects a disaster when detecting a vibration of a predetermined threshold value or more. The disaster detection device is provided in the detection device 100 or in an external device different from the detection device 100, and the disaster detection device and the specific information acquisition unit 112 are connected via a physical connection cable or a network. The In this way, the specific information acquisition unit 112 can acquire the time when the disaster occurred from the disaster detection device.

[Detailed configuration of detection processing unit]
FIG. 6 is a diagram illustrating an example of a detailed configuration of the detection processing unit in the first embodiment. FIG. 7 is a diagram illustrating an example of a detailed configuration of the specifying unit illustrated in FIG. 5.

  As illustrated in FIG. 6, the detection processing unit 12 includes a specifying unit 121 and an object detection unit 122.

  The specifying unit 121 includes a time-temperature converting unit 1211 and a reference model specifying unit 1212, and specifies one reference model from among a plurality of reference models based on the acquired specific information and environmental temperature.

  The time temperature conversion unit 1211 converts the specific information into the temperature of the object. In the present embodiment, the time temperature conversion unit 1211 converts the elapsed time, which is the specific information acquired by the specific information acquisition unit 112, into the temperature of the object to be detected. More specifically, the time-temperature conversion unit 1211 refers to the time-temperature conversion table 51 stored in the storage unit 5 and determines the elapsed time that is the specific information acquired by the specific information acquisition unit 112 of the detection target object. Convert to temperature.

  Here, an example of the time-temperature conversion table 51 will be described.

  FIG. 8 is a diagram for explaining an example of the time-temperature conversion table in the disaster case in the first embodiment. In the case of the disaster shown in FIG. 8, it is assumed that the affected person is covered with some object such as a wall of the house and the affected person is unable to act, and the environmental temperature Tc is lower than the human body temperature. ing. In this case, as shown in FIG. 8, it is considered that the human body temperature decreases linearly to the environmental temperature Tc and gradually decreases after the disaster time (elapsed time from the disaster occurrence time).

  Therefore, the time-temperature conversion table 51 describes the relationship between the elapsed time from the disaster occurrence time and the body temperature as shown in FIG. 8 (lower axis of FIG. 9 described later). By using this time-temperature conversion table 51, the time-temperature conversion unit 1211 can convert the elapsed time, which is the specific information acquired by the specific information acquisition unit 112, into a human temperature. Although the disaster time and the disaster occurrence time are handled at the same time, the present invention is not limited to this case. Further, the relationship between the elapsed time from the disaster occurrence time and the body temperature is not limited to the case shown in FIG. Any relationship that can be acquired in advance and has a certain degree of accuracy is acceptable.

  The reference model specifying unit 1212 specifies one reference model from among a plurality of reference models based on the environmental temperature acquired by the environmental temperature acquisition unit 113 and the temperature of the object converted by the time temperature conversion unit 1211. To do. In the present embodiment, the reference model specifying unit 1212 specifies one reference model with reference to a plurality of reference models held in the learning result related information DB 52 of the storage unit 5.

  Here, an example of a plurality of reference models held in the learning result related information DB 52 will be described. FIG. 9 is a diagram illustrating an example of a plurality of reference models in the first embodiment. FIG. 9 shows a plurality of reference models that define a shape composed of pixels representing a person in an infrared image when the environmental temperature is 15 degrees to 30 degrees assuming that the object to be detected is a person. . Since the human body temperature is around 37 degrees, the shape is smaller as the environmental temperature is closer to the human body temperature.

  Based on the environmental temperature acquired by the environmental temperature acquisition unit 113 and the temperature of an object such as a person converted by the time temperature conversion unit 1211, the reference model specifying unit 1212 includes a plurality of reference models as illustrated in FIG. Identify one reference model.

  The object detection unit 122 detects an object in the infrared image using the identified one reference model.

  More specifically, when there is a region similar to the shape and temperature indicated by the identified one reference model in the infrared image acquired by the acquisition unit 11, the object detection unit 122 detects the detection target in the region. It is determined that the object exists. And the object detection part 122 detects the object of a detection target by acquiring the information regarding the coordinate position of the said area | region in the infrared image which the acquisition part 11 acquired as a position of the said object.

  In the present embodiment, the object detection unit 122 determines that there is a person in a region similar to the shape and temperature indicated by one reference model identified from a plurality of reference models as shown in FIG. A person in the infrared image is detected by acquiring the coordinate position of the region.

[Detection device operation]
Next, the operation of the detection apparatus 100 configured as described above will be described.

  FIG. 10 is a flowchart showing an outline of the operation of the detection apparatus according to the first embodiment. FIG. 11 is a flowchart illustrating an example of details of the process of S23 of FIG. FIG. 12 is a flowchart illustrating an example of details of the processing in S24 of FIG. FIG. 13 is a flowchart showing an example of details of the processing in S25 of FIG.

  First, the detection apparatus 100 acquires an infrared image (S21). Specifically, the detection unit 1 acquires an infrared image from the infrared sensor 2.

  Next, the detection apparatus 100 acquires the environmental temperature (S22). Specifically, the detection unit 1 acquires an environmental temperature that is measured by the temperature sensor 3 and is the temperature of the space from which the infrared image is acquired.

  Next, the detection apparatus 100 acquires specific information (S23). Specifically, the detection unit 1 acquires specific information for specifying the temperature of a predetermined object in the infrared image. In the present embodiment, the specific information is an elapsed time that is an elapsed time from the time when the disaster occurred. Therefore, first, as shown in FIG. 11, the detection unit 1 acquires the time when the disaster was detected from the disaster detection device as the time when the disaster occurred (S231), and acquires the current time from the time measuring device 4 (S231). S232). Next, the detection unit 1 calculates the elapsed time from the time when the disaster occurred and the time measured by the time measuring device 4 (S233).

  Next, the detection apparatus 100 specifies a reference model (S24). Specifically, the detection unit 1 identifies one reference model from among a plurality of reference models based on the acquired environmental temperature and identification information. In the present embodiment, as shown in FIG. 12, the detection unit 1 first converts the elapsed time as the specific information into the temperature of the object to be detected (S241). Next, the detection unit 1 refers to the learning result related information DB 52 and acquires a plurality of reference models (S242). Next, the detection unit 1 identifies one reference model from among the plurality of reference models from the acquired environmental temperature and the converted object temperature (S243).

  Next, the detection apparatus 100 detects an object in the infrared image (S25). Specifically, the detection unit 1 detects an object in the infrared image using the specified one reference model. In the present embodiment, as shown in FIG. 11, the detection unit 1 first determines whether or not there is a region similar to the shape and temperature indicated by the specified one reference model in the acquired infrared image. (S251). When the region is present in the infrared image acquired in S251 (Yes in S251), the detection unit 1 detects that the region is an object to be detected in the infrared image (S252). On the other hand, when the region does not exist in the infrared image acquired in S251 (No in S251), the detection unit 1 detects that there is no object to be detected in the infrared image (S253).

  In addition, the process of S21-S23 is not restricted to this order. If it is before the process of S24, the order of the processes of S21 to S23 may be switched.

[Effects of First Embodiment, etc.]
As described above, according to the first embodiment, it is possible to realize a detection method that can stably detect an object in an infrared image even when the environmental temperature and the temperature of the object fluctuate.

  For example, when the object to be detected is a person, when the environmental temperature and the temperature of the object fluctuate, the human contour feature fluctuates due to lack of a human body part (contour) with a small difference between the environmental temperature and the like. . Therefore, even if a person in an infrared image is detected using one reference model that defines one contour feature as in the prior art, it cannot be detected stably.

  On the other hand, according to the detection method of the present embodiment, since a plurality of reference models that define the contour characteristics in the combination of the environmental temperature and the object temperature are used, the infrared rays can be detected even if the environmental temperature and the object temperature vary Objects in the image can be detected stably.

  For example, in the image taken by the visible camera shown in FIG. Since the area under the shield is not directly visible, it is not known whether there is a person in the area X. However, by using the detection method of the present embodiment, there is a possibility that it is possible to detect the presence of a person (region Y) in the region X as shown in FIG. Here, FIG. 14 and FIG. 15 are diagrams for explaining an example of the effect of the first embodiment. In this way, by using sensor fusion using not only a visible camera but also a temperature sensor or the like, there is a possibility that a robust recognition process robust to environmental changes may be realized.

  In the present embodiment, the learning result related information DB 52 stored in the storage unit 5 is described as a database in which a set of data of learning results by machine learning is held. Hereinafter, a process of obtaining a plurality of reference models as learning result data (learning products) by machine learning will be described. In the following description, the object is described as a person.

  FIG. 16 is a diagram illustrating an example of the configuration of the machine learning device. FIG. 17 is a diagram illustrating an example of a data set of learning results of the machine learning device illustrated in FIG.

  A machine learning device 200 illustrated in FIG. 16 includes a machine learning unit 21, a learning data related information DB 22, and a learning result related information DB 23.

  The learning data related information DB 22 holds data (learning data) for the machine learning unit 21 to learn. In the present embodiment, the learning data related information DB 22 holds a large number of infrared images including people defined by nine patterns of contour features. Each infrared image is associated with information on the environmental temperature at the time of shooting.

  The learning result related information DB 23 holds a set of learning result data after the machine learning unit 21 performs the learning process. In the present embodiment, the learning result related information DB 23 defines a shape composed of pixels indicating the person in a combination of the environmental temperature and the surface temperature (object temperature) of the person in the infrared image as shown in FIG. Nine reference models are held. The learning result related information DB 23 may be the same as the learning result related information DB 52. That is, a set of learning result data after the machine learning unit 21 performs the learning process may be directly stored in the learning result related information DB 52.

  The machine learning unit 21 learns using the learning data held in the learning data related information DB 22 and holds a set of learning result data in the learning result related information DB 23.

  Next, the process of the machine learning device 200 configured as described above will be described. In the following description, it is assumed that the object is a person.

  FIG. 18 is a diagram illustrating an example of processing of the machine learning unit 21 illustrated in FIG.

  First, the machine learning unit 21 selects learning data held in the learning result related information DB 23 (S31).

  Next, the machine learning unit 21 acquires the surface temperature of the person included in each of the infrared images constituting the learning data from the infrared image (S32).

  Next, the machine learning unit 21 acquires an environmental temperature corresponding to each of the infrared images constituting the learning data (S33).

  Next, the machine learning unit 21 performs a learning process using the infrared image constituting the selected learning data, the acquired human surface temperature, and the environmental temperature (S34). Note that the details of the learning process by machine learning are different from the gist of the present embodiment, and thus description thereof is omitted.

  Next, the machine learning unit 21 causes the learning result related information DB 23 to hold a learning result that is a result of performing the learning process (S35).

(Embodiment 2)
In the first embodiment, a disaster example is given as an example when the environmental temperature and the temperature of the object fluctuate. However, the present invention is not limited thereto. In the present embodiment, a human bathing will be described as an example when the temperature of an object varies.

  Hereinafter, the detection apparatus 100A according to the second embodiment will be described focusing on differences from the first embodiment.

  FIG. 19 is a diagram illustrating an example of the configuration of the detection device according to the second embodiment. Elements similar to those in FIG. 3 are denoted by the same reference numerals, and detailed description thereof is omitted.

  A detection device 100A shown in FIG. 19 differs from the detection device 100 of the first embodiment in the configuration of the detection unit 1A and the storage unit 5A. The detection unit 1A is different from the detection unit 1 of the first embodiment in the configuration of the acquisition unit 11A and the detection processing unit 12A.

[Configuration of storage unit]
FIG. 20 is a diagram illustrating an example of a configuration of a storage unit in the second embodiment. Elements similar to those in FIG. 4 are denoted by the same reference numerals, and detailed description thereof is omitted.

  As shown in FIG. 20, the storage unit 5A stores a time-temperature conversion table 51A and a learning result related information DB 52.

  The time-temperature conversion table 51A is a table used to convert time information into temperature information of an object when the specific information is information related to time (time information). In the present embodiment, the time-temperature conversion table 51A is a table in the bath-up example. Since details will be described later, description thereof is omitted here.

[Detailed configuration of acquisition unit]
FIG. 21 is a diagram illustrating an example of a detailed configuration of the acquisition unit according to the second embodiment. The same elements as those in FIG. 5 are denoted by the same reference numerals, and detailed description thereof is omitted.

  As illustrated in FIG. 21, the acquisition unit 11A includes an infrared image acquisition unit 111, a specific information acquisition unit 112A, and an environmental temperature acquisition unit 113. The acquisition unit 11A illustrated in FIG. 21 is different from the acquisition unit 11 illustrated in FIG. 5 in the configuration of the specific information acquisition unit 112A.

  The specific information acquisition unit 112A acquires specific information for specifying the temperature of a predetermined object in the infrared image. In the present embodiment, the object to be detected is a person, and the specific information is an elapsed time from a specific event. The specific event is that the person has taken the bath, and the elapsed time is the time that has elapsed since the person has taken the bath.

  The specific information acquisition unit 112A calculates and acquires the elapsed time from the time when the person gets out of the bath and the time measured by the time measuring device 4. The specific information acquisition unit 112A acquires the time when the person detected by the bath control system leaves the bath.

  Here, the bath control system is connected to the detection apparatus 100A via a network or a physical connection cable. The said control system detects the time when the said person got out of the bath by another sensors, such as the heat sensor of a dressing room, and the weight sensor attached to the bathroom. Note that the method for detecting the time when the control system leaves the person's bath is not limited to this example. Any method may be used as long as the control system can detect the time when the person gets out of the bath and can transmit it to the specific information acquisition unit 112A. In this manner, the specific information acquisition unit 112A can acquire the time when the person has taken the bath from the bath control system.

  Note that the control system may further sense the surface temperature of the person at the time when he / she gets out of the person's bath and transmit the sensed temperature to the specific information acquisition unit 112A.

[Detailed configuration of detection processing unit]
19 is different from the detection processing unit 12 shown in FIG. 6 in the configuration of the specifying unit 121A. Hereinafter, the specifying unit 121A will be described.

  FIG. 22 is a diagram illustrating an example of a detailed configuration of the specifying unit according to the second embodiment. The same elements as those in FIG. 7 are denoted by the same reference numerals, and detailed description thereof is omitted. 22 differs from the specifying unit 121 shown in FIG. 7 in the configuration of the time-temperature conversion unit 1211A.

  The time-temperature conversion unit 1211A converts the specific information into the temperature of the object. In the present embodiment, the time temperature conversion unit 1211A converts the elapsed time, which is the specific information acquired by the specific information acquisition unit 112A, into the temperature of the object to be detected. More specifically, the time-temperature conversion unit 1211A refers to the time-temperature conversion table 51A stored in the storage unit 5A, and determines the elapsed time that is the specific information acquired by the specific information acquisition unit 112A of the detection target object. Convert to temperature.

  Here, an example of the time-temperature conversion table 51A will be described.

  FIG. 23 is a diagram for explaining an example of a time-temperature conversion table in the bath-up example in the second embodiment. FIG. 23 shows an example of the relationship between time and body temperature (deep body temperature) when a person enters a bath filled with 41-degree hot water as a bath-up example. FIG. 24 is a diagram illustrating an example of a plurality of reference models in the second embodiment. FIG. 24 shows a plurality of reference models that define a shape composed of pixels indicating a person in an infrared image when the environmental temperature is 15 degrees to 30 degrees when the object to be detected is a person. . The shape becomes smaller as the environmental temperature and the human surface temperature are closer.

  As shown in FIG. 23, the human body temperature is considered to increase linearly with the time from the time of taking a bath. From there, it is thought that the surface temperature of a person falls linearly from the time of getting out of the bath. For example, as shown in FIG. 23, if a person takes a bath for 30 minutes and the body temperature reaches 39 degrees, it is considered that the temperature drops similarly twice in 30 minutes from the time of getting out of the bath. Therefore, the time-temperature conversion table 51A describes the relationship between the elapsed time from the time of getting out of the bath as shown in the lower axis of FIG. 24 and the surface temperature of the person. By using this time temperature conversion table 51A, the time temperature conversion unit 1211A can convert the elapsed time, which is the specific information acquired by the specific information acquisition unit 112A, into a human temperature (surface temperature).

  It should be noted that the relationship between the time when the person leaves the bath and the body temperature (deep body temperature) is not limited to the example shown on the lower axis of FIG. Any relationship that can be acquired in advance and has a certain degree of accuracy is acceptable. Further, the position of the origin of the lower axis in FIG. 24 may be appropriately moved by separately acquiring the surface temperature at the time when the person gets out of the bath with a control system or the like.

[Detection device operation]
Next, the operation of the detection apparatus 100A configured as described above will be described.

  FIG. 25 is a flowchart showing an outline of the operation of the detection apparatus according to the first embodiment. FIG. 26 is a flowchart illustrating an example of details of the processing of S23 in the second embodiment.

  Since the detection apparatus 100A differs from the operation of the detection apparatus 100 shown in FIG. 10 in the process of S23A, the process of S23A will be described below. Since the processing of S21, S22, S24, and S25 is as described in the first embodiment, the description thereof is omitted here.

  In S <b> 23 </ b> A, the detection apparatus 100 </ b> A acquires specific information for specifying the temperature of a person in the infrared image. Here, the specific information is an elapsed time that is an elapsed time from the time when the person leaves the bath. More specifically, as shown in FIG. 26, first, the detection apparatus 100A acquires the time when the person to be detected has gone out of the bath from the bath control system (S231A), and obtains the current time from the time measuring device 4. Obtain (S232A). Next, the detection device 100A calculates the elapsed time from the time when the person gets out of the bath and the time measured by the time measuring device 4 (S233A).

  Thereafter, the detection apparatus 100A performs the processes of S24 and S25 to identify one reference model from among the plurality of reference models based on the acquired environmental temperature and the specific information, and to identify the identified one reference model Is used to detect a person in an infrared image.

  In this way, a person in an infrared image can be detected stably even if the person's surface temperature differs immediately after bathing or afterwards.

[Effects of Second Embodiment, etc.]
As described above, according to the second embodiment, it is possible to realize a detection method and the like that can stably detect a person in an infrared image even when the environmental temperature and the temperature of a person as an object fluctuate.

  More specifically, in the detection method according to the second embodiment, the time information that can be acquired is converted into the surface temperature of the person, and a plurality of reference models that define the contour characteristics in the combination of the environmental temperature and the surface temperature of the person are used. . Thereby, in the detection method of the second embodiment, even if the environmental temperature and the temperature of the person fluctuate and the region that can be recognized as the person in the thermal image changes, the infrared image can be obtained by using the reference model corresponding to the fluctuation. The object in the can be detected stably.

(Embodiment 3)
In the third embodiment, a case different from the second embodiment will be described as an example in the case where the environmental temperature and the temperature of the object fluctuate.

  In the present embodiment, a case where a person is exercising will be described as an example of a case where the temperature of an object varies.

  Hereinafter, the detection apparatus 100B according to the third embodiment will be described focusing on differences from the first embodiment.

  FIG. 27 is a diagram illustrating an example of the configuration of the detection device according to the third embodiment. Elements similar to those in FIG. 3 are denoted by the same reference numerals, and detailed description thereof is omitted.

  The detection device 100B shown in FIG. 27 differs from the detection device 100 of the first embodiment in the configuration of the detection unit 1B and the storage unit 5B. The detection unit 1B is different from the detection unit 1 of the first embodiment in the configuration of the acquisition unit 11B and the detection processing unit 12B.

[Configuration of storage unit]
FIG. 28 is a diagram illustrating an example of a configuration of a storage unit in the third embodiment. Elements similar to those in FIG. 4 are denoted by the same reference numerals, and detailed description thereof is omitted.

  The storage unit 5B stores a learning result related information DB 52 as shown in FIG. That is, the storage unit 5B shown in FIG. 28 differs from the storage unit 5 shown in FIG. 4 in that the time-temperature conversion table 51 is not provided.

[Detailed configuration of acquisition unit]
FIG. 29 is a diagram illustrating an example of a detailed configuration of an acquisition unit according to Embodiment 3. The same elements as those in FIG. 5 are denoted by the same reference numerals, and detailed description thereof is omitted.

  As shown in FIG. 29, the acquisition unit 11B includes an infrared image acquisition unit 111, a specific information acquisition unit 112B, and an environmental temperature acquisition unit 113. The acquisition unit 11B illustrated in FIG. 29 differs from the acquisition unit 11 illustrated in FIG. 5 in the configuration of the specific information acquisition unit 112B.

  The specific information acquisition unit 112B acquires specific information for specifying the temperature of a predetermined object in the infrared image. In the present embodiment, the specific information is the temperature of the object to be detected. The temperature of this object is measured by a temperature measuring device attached to the object. In the following description, it is assumed that the detection target object is a person.

  The specific information acquisition unit 112B acquires the temperature of the object from the temperature measurement device via a network or a physical connection cable. More specifically, the specific information acquisition unit 112B acquires the temperature of the person to be detected by communicating with the temperature measurement device worn by the person. The temperature measuring device may be able to measure the person's pulse and the like. Thereby, since a person can exercise while visually recognizing his / her health condition, the possibility of wearing becomes high.

[Detailed configuration of detection processing unit]
FIG. 30 is a diagram illustrating an example of a detailed configuration of the detection processing unit in the third embodiment. FIG. 31 is a diagram illustrating an example of a detailed configuration of the specifying unit illustrated in FIG. 30. Elements similar to those in FIGS. 6 and 7 are denoted by the same reference numerals, and detailed description thereof is omitted.

  The detection processing unit 12B is different from the detection processing unit 12 illustrated in FIG. 6 in the configuration of the specifying unit 121B.

  The configuration of the specifying unit 121B is different from that of the specifying unit 121 illustrated in FIG. 7 in that the configuration of the time-temperature conversion unit 1211 is not provided. This is because in the present embodiment, the specific information acquisition unit 112B can acquire the temperature of the person to be detected from the temperature measurement device, and therefore it is not necessary to include the configuration of the time-temperature conversion unit 1211.

  Here, the relationship between time and body temperature during human exercise will be described.

  FIG. 32 is a diagram illustrating an example of a relationship between time and body temperature during exercise of a person in the third embodiment. For example, assuming that a person exercises while taking a break, a rise in body temperature during exercise and a decrease in body temperature during break are alternately repeated. And the degree of the fall of body temperature during a break decreases with time, and the body temperature during exercise tends to increase as a whole. That is, there is no linear relationship between time and body temperature. Therefore, in this embodiment, a human body temperature is measured using a temperature measurement device that can be worn by the person. By doing so, it is possible to use a plurality of reference models that define a shape composed of pixels indicating the person in a combination of the environmental temperature and the person's temperature.

[Detection device operation]
Next, the operation of the detection apparatus 100B configured as described above will be described.

  FIG. 33 is a flowchart illustrating an outline of the operation of the detection device according to the third embodiment.

  The detection device 100B differs from the operation of the detection device 100 shown in FIG. 10 in the process of S23B. Since the processing of S21, S22, S24, and S25 is as described in the first embodiment, the description thereof is omitted here.

  In S23B, the detection apparatus 100B acquires specific information for specifying the temperature of the person in the infrared image, and proceeds to the process of S24. Here, the specific information is the temperature of the person measured by the temperature measuring device attached to the person. That is, in S23B, the time-temperature conversion process as described in the first and second embodiments is not required, and therefore the process can proceed to S24 as it is.

  Next, in S24 and S25, the detection apparatus 100B specifies one reference model from among a plurality of reference models based on the acquired environmental temperature and the temperature of the person, and uses the specified one reference model. Detect a person in an infrared image.

  In this way, a person in an infrared image can be detected stably even if the surface temperature of the person is different during exercise.

[Effects of Embodiment 3, etc.]
As described above, according to the third embodiment, it is possible to realize a detection method or the like that can stably detect a person in an infrared image even when the environmental temperature and the temperature of a person as an object fluctuate.

  More specifically, in the detection method according to the third embodiment, the temperature of the person measured from the temperature measurement device attached to the person is acquired, and a plurality of contour characteristics are defined for the combination of the environmental temperature and the temperature of the person. The reference model is used. As a result, even if the environmental temperature and the temperature of the person fluctuate and the area that can be recognized as a person in the thermal image changes, the object in the infrared image can be detected stably by using the reference model according to the fluctuation. can do.

  In the present embodiment, the detection target object is described as a person, but the present invention is not limited to this. The object to be detected may be a posture of a person such as standing, sitting or raising a hand. In this case, it is only necessary to have a plurality of reference models that define the contour feature in the combination of the environmental temperature and the temperature of the person for each posture of the person.

(Embodiment 4)
This Embodiment demonstrates the air conditioner which mounts the detection apparatus in Embodiment 1-3.

[Configuration of air conditioner]
FIG. 34 is a diagram illustrating an example of a configuration of an air conditioner according to Embodiment 4. FIG. 35 is a diagram illustrating an example of a state in which the air conditioner of FIG. 34 is installed. The same elements as those in FIG. 7 are denoted by the same reference numerals, and detailed description thereof is omitted.

  As shown in FIG. 34, the air conditioner 300 includes at least the detection unit 1, the infrared sensor 2, the temperature sensor 3, the time measuring device 4, and the control unit 7 in the first embodiment, for example. The air conditioner 300 includes a louver, a compressor, a fan, and the like, and at least one of a set temperature, a wind direction, and an air volume according to the position and temperature of the object detected in the target space 70 shown in FIG. The air conditioning control of the target space 70 is performed by controlling one.

  Here, the air conditioner 300 is an air conditioner, for example, and is installed at a predetermined height from the floor on an installation surface substantially perpendicular to the floor of the target space 70 as shown in FIG. 35, for example. The air conditioner 300 controls the louver, the compressor, and the fan according to the position and temperature of the person 71 detected by the detection unit 1 attached to the front surface of the air conditioner 300, for example, to set temperature and wind direction. Then, air conditioning control of the target space 70 is performed by controlling at least one of the air volumes.

  In this Embodiment, the detection part 1 detects the person 71 in the infrared image acquired in the object space 70 which is the space which the air conditioner 300 makes air conditioning object.

  The control unit 7 is comfortable for the person 71 by controlling the louver, the compressor, the fan, and the like according to the detection result of the detection unit 1, that is, the position and temperature of the person 71 in the infrared image acquired in the target space 70. The air conditioning control of the target space 70 is performed so that the ambient temperature becomes a proper temperature. For example, when the surface temperature of the person 71 in the infrared image acquired in the target space 70 is high (that is, it can be determined that the person 71 feels hot), the control unit 7 operates the compressor and the fan to Control the wind speed and generate cold air. The control unit 7 may perform control so that the louver is directed toward the position of the person 71. Thereby, since the surface temperature of the person 71 can be lowered, the person 71 does not feel hot and can spend comfortably. For example, when the surface temperature of the person 71 in the infrared image acquired in the target space 70 is low (that is, it can be determined that the person 71 feels cold), the control unit 7 operates the compressor and the fan to Controls temperature and wind speed to generate hot air. The control unit 7 may perform control so that the louver is directed toward the position of the person 71. Thereby, since the surface temperature of the person 71 can be raised, the person 71 does not feel cold and can spend comfortably.

[Air conditioner operation]
Next, the operation of the air conditioner 300 configured as described above will be described.

  FIG. 36 is a flowchart showing an outline of the operation of the air conditioner according to the fourth embodiment.

  As shown in FIG. 36, first, the air conditioner 300 acquires an infrared image of the target space 70, which is a target space for air conditioning, and detects an object in the infrared image (S101). Specifically, the detection unit 1 mounted on the air conditioner 300 detects the person 71 in the infrared image, but since details are as described above, description thereof is omitted here.

  Next, the air conditioner 300 performs air conditioning control of the target space 70 by controlling at least one of the set temperature, the wind direction, and the air volume in accordance with the detected position and temperature of the object (S102). ).

[Effects of Embodiment 4 and the like]
As described above, according to the fourth embodiment, an air conditioner using a detection device that can stably detect a person in an infrared image even when the environmental temperature and the temperature of an object fluctuate. A control method can be realized. Thereby, the air conditioning of the space where the said person exists can be performed so that a person may feel comfortable.

  In addition, Embodiment 4 demonstrated the air conditioner provided with the detection part 1, the infrared sensor 2, the temperature sensor 3, the time measuring device 4, and the control part 7. FIG. However, some components of the air conditioner such as the detection unit 1, the time measuring device 4, and the control unit 7 may be configured separately as software. In this case, the main body that processes the software may be an arithmetic unit of an air conditioner, an arithmetic unit included in a PC (personal computer), a smart phone, or the like. It may be a cloud server connected via a network. Moreover, it is good also considering some structures, such as the detection part 1, the time measuring device 4, and the control part 7, as a structure different from an air conditioner. That is, it is good also as a sensor system provided with the detection part 1 described in said embodiment at least.

  As described above, according to the present invention, it is possible to realize a detection method, a detection apparatus, and a control method using the same that can stably detect an object in an infrared image even when the environmental temperature and the temperature of the object fluctuate. .

  In this way, the product (a plurality of reference models) obtained by machine learning of infrared images is recognized and processed after switching to one suitable for each environment (environment temperature and object temperature) in disaster cases. Positive recognition can be improved while suppressing recognition. Thereby, the object in the infrared image can be stably detected even in an environment where the change in the environmental temperature or the object temperature is large.

  That is, by using these detection methods, detection devices, and control methods using the detection methods, recognition processing that is robust against environmental changes can be realized.

  As described above, the detection method, the detection apparatus, and the control method using the detection method according to one or more aspects of the present invention have been described based on the embodiment. However, the present invention is limited to this embodiment. is not. Unless it deviates from the gist of the present invention, one or more of the present invention may be applied to various modifications that can be conceived by those skilled in the art, or forms constructed by combining components in different embodiments. It may be included within the scope of the embodiments. For example, the present invention includes the following cases.

  (1) In the first to third embodiments described above, the detection device has been described as including a detection unit, an infrared sensor, a temperature sensor, a time measuring device, a storage unit, and an output unit. Not limited to that. The detection apparatus of the present invention only needs to include, for example, a detection unit (detection unit 1, 1A, or 1B) as a minimum configuration.

  (2) In the first to third embodiments described above, the detection device includes the storage unit, but the present invention is not limited thereto. The time-temperature conversion table and the learning result related information DB included in the storage unit may be stored in a storage device connected to the detection device via a network. Further, the time-temperature conversion table and the learning result related information DB may be provided by a cloud service such as a service provider connected to the detection device via a network.

  (3) In the above-described first to third embodiments, a person has been described as an example of an object to be detected. However, the present invention is not limited to this. Animals such as dogs and cats may be used as long as they are constant temperature animals.

  (4) Specifically, each of the above devices is a computer system including a microprocessor, a ROM, a RAM, a hard disk unit, a display unit, a keyboard, a mouse, and the like. A computer program is stored in the RAM or hard disk unit. Each device achieves its functions by the microprocessor operating according to the computer program. Here, the computer program is configured by combining a plurality of instruction codes indicating instructions for the computer in order to achieve a predetermined function.

  (5) A part or all of the constituent elements constituting each of the above-described devices may be configured by one system LSI (Large Scale Integration). The system LSI is an ultra-multifunctional LSI manufactured by integrating a plurality of components on a single chip, and specifically, a computer system including a microprocessor, ROM, RAM, and the like. . A computer program is stored in the RAM. The system LSI achieves its functions by the microprocessor operating according to the computer program.

  (6) A part or all of the components constituting each of the above devices may be configured as an IC card that can be attached to and detached from each device or a single module. The IC card or the module is a computer system including a microprocessor, a ROM, a RAM, and the like. The IC card or the module may include the super multifunctional LSI described above. The IC card or the module achieves its function by the microprocessor operating according to the computer program. This IC card or this module may have tamper resistance.

  (7) The present disclosure may be the method described above. Further, the present invention may be a computer program that realizes these methods by a computer, or may be a digital signal composed of the computer program.

  (8) Further, the present disclosure provides a computer-readable recording medium such as a flexible disk, a hard disk, a CD-ROM, an MO, a DVD, a DVD-ROM, a DVD-RAM, a BD ( It may be recorded on a Blu-ray (registered trademark) Disc), a semiconductor memory, or the like. The digital signal may be recorded on these recording media.

  (9) Further, the present disclosure may transmit the computer program or the digital signal via an electric communication line, a wireless or wired communication line, a network represented by the Internet, a data broadcast, or the like.

  (10) The present disclosure may be a computer system including a microprocessor and a memory, the memory storing the computer program, and the microprocessor operating according to the computer program. .

  (11) In addition, by recording the program or the digital signal on the recording medium and transferring it, or transferring the program or the digital signal via the network or the like, another independent computer It may be implemented by the system.

  (12) The above embodiment and the above modifications may be combined.

  The present invention can be used for a detection method, a detection device, and a control method using the same, and in particular, an air conditioner that is installed in a vehicle air conditioner or an individual air conditioner installed in an office and performs air conditioning control of a target space, It can be used for a monitoring device used in a parking lot or the like.

DESCRIPTION OF SYMBOLS 1, 1A, 1B detection part 2 Infrared sensor 3 Temperature sensor 4 Time measuring device 5, 5A, 5B Storage part 6 Output part 7 Control part 11, 11A, 11B Acquisition part 12, 12A, 12B Detection processing part 21 Machine learning part 22 Learning data related information DB
23, 52 Learning result related information DB
51, 51A Time temperature conversion table 70 Target space 71 Person 100, 100A, 100B Detection device 111 Infrared image acquisition unit 112, 112A, 112B Specific information acquisition unit 113 Environmental temperature acquisition unit 121, 121A, 121B Identification unit 122 Object detection unit 200 Machine learning device 300 Air conditioner 1211, 1211A Time temperature conversion unit 1212 Reference model specifying unit

Claims (10)

A detection method in which a computer detects a predetermined object in an infrared image,
(I) an acquisition step of acquiring information related to an environmental temperature that is a temperature of the space from which the infrared image is acquired; and (ii) specifying information for specifying the temperature of the object;
A specifying step of specifying one reference model from a plurality of reference models based on the specifying information and the environmental temperature;
Detecting the object in the infrared image using the identified one reference model,
Detection method. Each of the plurality of reference models defines a shape constituted by pixels indicating the object in a combination of the environmental temperature and the temperature of the object,
The shape differs depending on the combination of the environmental temperature and the temperature of the object,
The shape becomes smaller as the ambient temperature and the temperature of the object are closer.
The detection method according to claim 1. In the detection step,
When there is a region similar to the shape and temperature indicated by the one reference model in the infrared image, it is determined that the object exists in the region,
Detecting the object by obtaining information on the coordinate position of the region in the infrared image as the position of the object;
The detection method according to claim 2. The specific information is an elapsed time from a specific event,
In the specifying step, the elapsed time is converted into the temperature of the object,
Identifying the one reference model based on the environmental temperature and the converted temperature of the object;
The detection method of any one of Claims 1-3. The specific event is a disaster,
The elapsed time is the time elapsed from the time when the disaster occurred,
The time when the disaster occurred is the time when the disaster detection device provided in the computer or in an external device detected the disaster,
In the acquisition step,
The computer acquires the elapsed time by acquiring the time when the disaster was detected from the disaster detection device via a network or a physical connection cable.
The detection method according to claim 4. The disaster detection device detects the disaster when detecting a vibration of a predetermined threshold value or more.
The detection method according to claim 5. The object is a person,
The specific event is that the person has taken a bath,
The elapsed time is the time elapsed from the time when the person got out of the bath,
In the acquisition step,
The computer obtains the elapsed time by obtaining the time when the person has taken the bath from the bath control system via a network or a physical connection cable.
The detection method according to claim 4. The specific information is the temperature of the object,
The temperature of the object is measured by a temperature measuring device attached to the object,
In the obtaining step, the computer
Obtaining the temperature of the object from the temperature measuring device via a network or a physical connection cable;
The detection method according to claim 1. A control method for controlling at least one of set temperature, wind direction, and air volume of an air conditioner,
A computer that executes the detection method according to claim 7, in the detection step, detects the object in the infrared image acquired in a space that the air conditioner is subject to air conditioning,
The control method includes a control step of controlling the air conditioner according to the position and temperature of the object detected by the computer.
Control method. An infrared image acquisition unit for acquiring an infrared image;
An environmental temperature acquisition unit that acquires an environmental temperature that is the temperature of the space from which the infrared image was acquired;
A specific information acquisition unit for acquiring specific information for specifying the temperature of a predetermined object in the infrared image;
A detection processing unit for detecting the object in the infrared image,
The detection processing unit
One reference model is specified from the plurality of reference models based on the specification information and the environmental temperature, and the object in the infrared image is detected using the specified reference model. ,
Detection device.