Determination device, determination method, and program

Abstract

The present invention provides an assessing device comprising: a reference image acquiring unit for acquiring a reference image; a reference image analyzing unit for identifying a worn item worn by a worker appearing in the reference image; a template information generating unit for generating worn item template information indicating the worn item that the worker should be wearing, on the basis of an analysis result of the reference image, and storing the same in a storage unit; a target image acquiring unit for acquiring a target image in which a worker to be assessed appears; a target image analyzing unit for identifying a worn item worn by the worker appearing in the target image; and an assessing unit for assessing whether the worker to be assessed is wearing the worn item that should be worn, on the basis of the worn item identified by analyzing the target image and the worn item indicated in the worn item template information.

Description

Determination device, determination method, and recording medium

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

[0002] Techniques related to the present invention are disclosed in Patent Documents 1 to 3.

[0003] The technology disclosed in Patent Document 1 detects a person's face as well as a detection target such as clothing or equipment, and then checks the possession of the detection target based on the positional relationship between the detected face and the detection target, and determines whether the detection target is properly equipped.

[0004] The technology disclosed in Patent Document 2 uses image analysis to monitor whether a subject's outfit is appropriate for the work environment or the work content. This technology determines whether the subject's outfit is appropriate based on pre-prepared suitability information. The suitability information defines the appropriate outfit required of the worker.

[0005] The technology disclosed in Patent Document 3 uses image analysis to determine whether a target person is wearing the equipment they should be wearing. This technology identifies the attributes of the target person through image analysis and determines whether the target person is wearing the equipment corresponding to the identified attributes. The attributes include the group, department, gender, etc., defined for each work content.

[0006] JP 2012-237569 A JP 2015-95209 A JP 2014-154129 A

[0007] At work sites such as building sites and construction sites, workers may be required to wear specific attachments to ensure their safety, etc. The technologies described in Patent Documents 1 to 3 make it possible to determine by image analysis whether a worker is wearing the specific attachments.

[0008] In the technology for determining whether a worker is wearing a predetermined accessory through image analysis, it is necessary to input information indicating the "predetermined accessory that the worker should wear" into the system in advance. The system determines whether the worker is wearing the predetermined accessory by comparing the accessory identified by image analysis with the predetermined accessory that the worker should wear, which is identified by the previously input information.

[0009] Therefore, there is a need for a technology that allows an operator to efficiently input information indicating the specific attachments that the operator should wear into the system. Patent Documents 1 to 3 do not disclose the above problem or a means for solving the problem.

[0010] In view of the above-mentioned problems, one example of the object of the present invention is to provide a determination device, a determination method, and a program that solve the problem of efficiently inputting information indicating the specified accessory that the worker should wear into a system in a technology that determines whether the worker is wearing the specified accessory through image analysis.

[0011] According to one aspect of the present invention, there is provided a determination device having: reference image acquisition means for acquiring a reference image; reference image analysis means for analyzing the reference image and identifying equipment worn by a worker appearing in the reference image; template information generation means for generating equipment template information indicating equipment to be worn by the worker based on a result of the analysis of the reference image and storing the template information in storage means; target image acquisition means for acquiring a target image in which a worker to be determined is appearing; target image analysis means for analyzing the target image and identifying equipment worn by the worker appearing in the target image; and determination means for determining whether the worker to be determined is wearing the equipment to be worn, based on the equipment identified by the analysis of the target image and the equipment indicated by the equipment template information.

[0012] According to one aspect of the present invention, there is provided a method for determining whether the worker to be determined is wearing the appropriate equipment, the method including: one or more computers acquiring a reference image, analyzing the reference image, and identifying equipment worn by a worker appearing in the reference image; generating equipment template information indicating the equipment that should be worn by the worker based on the analysis of the reference image, and storing the information in a storage means; acquiring a target image in which the worker to be determined is appearing; analyzing the target image, and identifying the equipment worn by the worker appearing in the target image; and determining whether the worker to be determined is wearing the appropriate equipment, based on the equipment identified in the analysis of the target image and the equipment indicated by the equipment template information.

[0013] According to one aspect of the present invention, there is provided a program that causes a computer to function as: reference image acquisition means for acquiring a reference image; reference image analysis means for analyzing the reference image and identifying equipment worn by a worker appearing in the reference image; template information generation means for generating equipment template information indicating equipment that should be worn by the worker based on the analysis result of the reference image and storing the information in storage means; target image acquisition means for acquiring a target image in which a worker to be judged is appearing; target image analysis means for analyzing the target image and identifying equipment worn by the worker appearing in the target image; and determination means for determining whether the worker to be judged is wearing the equipment that should be worn, based on the equipment identified by the analysis of the target image and the equipment indicated by the equipment template information.

[0014] According to one aspect of the present invention, in a technology for determining whether a worker is wearing a specified accessory through image analysis, a determination device, a determination method, and a program are realized that solve the problem of efficiently inputting information indicating the specified accessory that the worker should wear into a system.

[0015] The above-mentioned objects and other objects, features and advantages will become more apparent from the following description of the preferred embodiments and the accompanying drawings.

[0016] FIG. 1 is a diagram showing an example of a functional block diagram of a determination device; FIG. 2 is a diagram showing an example of a hardware configuration of a determination device; FIG. 3 is a diagram showing an example of a reference image; FIG. 4 is a flowchart showing an example of a processing flow of the determination device; FIG. 5 is a diagram showing another example of a reference image; FIG. 6 is a diagram showing an example of attachment template information; FIG. 7 is a diagram showing an output example of a determination result; FIG. 8 is a diagram showing another output example of a determination result; FIG. 9 is a flowchart showing another example of a processing flow of the determination device; FIG. 10 is a diagram showing an example of history information.

[0017] Hereinafter, embodiments of the present invention will be described with reference to the drawings. In all the drawings, like components are designated by like reference numerals, and the description thereof will be omitted as appropriate.

[0018] 1 is a functional block diagram showing an overview of a determination device 10 according to a first embodiment. The determination device 10 includes a reference image acquisition unit 11, a reference image analysis unit 12, a template information generation unit 13, a target image acquisition unit 14, a target image analysis unit 15, a determination unit 16, and a storage unit 17. Note that the determination device 10 does not necessarily have to include the storage unit 17. In this case, an external device configured to be able to communicate with the determination device 10 includes the storage unit 17.

[0019] Preparations for the determination are made by the reference image acquisition unit 11, the reference image analysis unit 12, and the template information generation unit 13. Then, the determination is made by the target image acquisition unit 14, the target image analysis unit 15, and the determination unit 16.

[0020] The determination is a determination as to whether the worker to be evaluated is wearing the appropriate accessory. The advance preparation is a registration of information indicating the appropriate accessory that the worker should wear.

[0021] The reference image acquisition unit 11 acquires a reference image. The reference image analysis unit 12 analyzes the reference image and identifies the equipment worn by the worker appearing in the reference image. The template information generation unit 13 generates equipment template information indicating the equipment to be worn by the worker based on the analysis result of the reference image, and stores the generated equipment template information in the storage unit 17.

[0022] The target image acquisition unit 14 acquires a target image that shows the worker to be judged. The target image analysis unit 15 analyzes the target image and identifies the equipment worn by the worker appearing in the target image. The determination unit 16 determines whether the worker to be judged is wearing the equipment that should be worn, based on the equipment identified by the analysis of the target image and the equipment indicated in the template information.

[0023] In this way, the determination device 10 of the first embodiment generates accessory template information indicating the accessory that the worker should wear based on the analysis results of the reference image, making it possible to efficiently input information indicating the predetermined accessory that the worker should wear into the system.

[0024] Second Embodiment "Overview" The determination device 10 of the second embodiment is a specific implementation of the determination device 10 of the first embodiment. More specifically, the determination device 10 of the second embodiment is a specific implementation of a process for generating accessory template information. The determination device 10 of the second embodiment analyzes a reference image of a worker wearing an accessory to be worn, and generates accessory template information based on the accessory worn by the worker. The configuration of the determination device 10 will be described in detail below.

[0025] "Hardware Configuration" An example of the hardware configuration of the determination device 10 will be described. Each functional unit of the determination device 10 is realized by any combination of hardware and software. Those skilled in the art will understand that there are various variations in the realization method and device. The software includes programs that are pre-stored in the device before shipping, programs downloaded from recording media such as CDs (Compact Discs), and programs downloaded from servers on the Internet.

[0026] FIG. 2 is a block diagram illustrating an example of the hardware configuration of the determination device 10. As shown in FIG. 2, the determination device 10 has a processor 1A, a memory 2A, an input / output interface 3A, a peripheral circuit 4A, and a bus 5A. The peripheral circuit 4A includes various modules. The determination device 10 does not necessarily have to have the peripheral circuit 4A. Note that the determination device 10 may be composed of multiple devices that are physically and / or logically separated. In this case, each of the multiple devices can have the above hardware configuration.

[0027] The bus 5A is a data transmission path for the processor 1A, memory 2A, peripheral circuit 4A, and input / output interface 3A to mutually transmit and receive data. The processor 1A is, for example, a processing unit such as a CPU or a graphics processing unit (GPU). The memory 2A is, for example, a random access memory (RAM) or a read-only memory (ROM). The input / output interface 3A includes interfaces for acquiring information from input devices, external devices, external servers, external sensors, cameras, etc., and interfaces for outputting information to output devices, external devices, external servers, etc. The input / output interface 3A also includes an interface for connecting to a communication network such as the Internet. Examples of input devices include a keyboard, mouse, microphone, physical buttons, touch panel, etc. Examples of output devices include a display, speaker, printer, mailer, etc. The processor 1A can issue commands to each module and perform calculations based on the results of those calculations.

[0028] "Functional Configuration" Next, the functional configuration of the determination device 10 of this embodiment will be described in detail. FIG. 1 shows an example of a functional block diagram of the determination device 10. As shown in the figure, the determination device 10 has a reference image acquisition unit 11, a reference image analysis unit 12, a template information generation unit 13, a target image acquisition unit 14, a target image analysis unit 15, a determination unit 16, and a storage unit 17. Note that the determination device 10 does not necessarily have to have the storage unit 17. In this case, an external device configured to be able to communicate with the determination device 10 has the storage unit 17.

[0029] The configurations of the target image acquisition unit 14, the target image analysis unit 15, and the determination unit 16 are the same as those in the first embodiment. The configurations of the reference image acquisition unit 11, the reference image analysis unit 12, the template information generation unit 13, and the storage unit 17 will be described below.

[0030] The reference image acquisition unit 11 acquires a reference image. The reference image is an image in which a worker is captured.

[0031] The reference image may be, for example, an image captured by having a worker stand in front of a camera, as shown in FIG. 3 . The reference image in this embodiment is an image generated by capturing an image of a worker wearing the appropriate attachment. A person may operate the camera to capture an image of the worker, and the image may be input to the determination device 10 by any means. Alternatively, the worker may be captured by a camera installed in any location and communicably connected to the determination device 10. This camera may then transmit the image to the determination device 10 using a communication function. The reference image may be a moving image or a still image.

[0032] 1 , the reference image analysis unit 12 analyzes the reference image and identifies the equipment worn by the worker appearing in the reference image. That is, the reference image analysis unit 12 identifies what equipment is worn by the worker appearing in the reference image.

[0033] The articles of wear include work clothes, a helmet, glasses, gloves, a mask, work boots, personal belongings, etc. These articles of wear may be further subdivided. Then, the reference image analysis unit 12 may identify which of the subdivided articles of wear is being worn.

[0034] The process of identifying the items worn by the worker in the reference image through image analysis can be realized using any technology. For example, a learning model that learns the characteristics of each item of clothing through machine learning may be used.

[0035] The template information generating unit 13 generates attachment template information indicating attachments that should be worn by the worker based on the analysis result of the reference image, and stores the template information in the storage unit 17 .

[0036] In this embodiment, the reference image acquisition unit 11 acquires a reference image of a "worker wearing an attachment that should be worn." The reference image analysis unit 12 then analyzes the reference image and identifies the attachment of the "worker wearing an attachment that should be worn."

[0037] The template information generating unit 13 generates clothing template information that indicates the "wear identified by the reference image analyzing unit 12" as "wear that should be worn by the worker."

[0038] Next, an example of the processing flow of the determination device 10 will be described with reference to the flowchart of FIG.

[0039] First, the determination device 10 acquires a reference image of "a worker wearing the appropriate garment" (S10). Next, the determination device 10 analyzes the reference image and identifies the garment of the "worker wearing the appropriate garment" that appears in the reference image (S11). Then, the determination device 10 generates garment template information that indicates the "garment identified in S11" as "garment that should be worn by the worker" and registers it (stores it in the storage unit 17) (S12).

[0040] "Operational Effects" According to the determination device 10 of the second embodiment, operational effects similar to those of the determination device 10 of the first embodiment are achieved.

[0041] Furthermore, the determination device 10 of the second embodiment analyzes a reference image of a "worker wearing the appropriate attachment" and generates attachment template information, making it possible to generate attachment template information that appropriately indicates the appropriate attachment.

[0042] Third Embodiment "Overview" The determination device 10 of the third embodiment is a specific implementation of the determination device 10 of the first embodiment. More specifically, the determination device 10 of the third embodiment is a specific implementation of a process for generating accessory template information. The determination device 10 of the third embodiment detects a predetermined worker who is thought to be wearing an accessory to be worn from an image (reference image) captured at a work site, and generates accessory template information based on the accessory worn by that worker. The configuration of the determination device 10 will be described in detail below.

[0043] "Hardware Configuration" This is the same as the determination device 10 of the second embodiment, so a description thereof will be omitted here.

[0044] "Functional Configuration" Next, the functional configuration of the determination device 10 of this embodiment will be described in detail. FIG. 1 shows an example of a functional block diagram of the determination device 10. As shown in the figure, the determination device 10 has a reference image acquisition unit 11, a reference image analysis unit 12, a template information generation unit 13, a target image acquisition unit 14, a target image analysis unit 15, a determination unit 16, and a storage unit 17. Note that the determination device 10 does not necessarily have to have the storage unit 17. In this case, an external device configured to be able to communicate with the determination device 10 has the storage unit 17.

[0045] The configurations of the target image acquisition unit 14, the target image analysis unit 15, and the determination unit 16 are the same as those in the first embodiment. The configurations of the reference image acquisition unit 11, the reference image analysis unit 12, the template information generation unit 13, and the storage unit 17 will be described below.

[0046] The reference image acquisition unit 11 acquires a reference image. The reference image is an image in which a worker is captured. The reference image may be a moving image or a still image.

[0047] The reference image in this embodiment is an image of a worker working at a work site, as shown in FIG. 5 , for example. A person may operate a camera to photograph the work site and input the image to the determination device 10 by any means. Alternatively, a camera installed at any location at the work site and communicably connected to the determination device 10 may photograph the work site. The camera may then use its communication function to transmit the image to the determination device 10. Alternatively, an apparatus equipped with a moving means and a camera and further communicably connected to the determination device 10 may photograph the work site while moving within the work site. The apparatus may then use its communication function to transmit the image to the determination device 10. The moving means may be a means for moving on land or a means for moving in the air.

[0048] The reference image analysis unit 12 detects a predetermined worker from an image (reference image) taken at a work site based on the appearance features of the predetermined worker registered in advance. Then, the reference image analysis unit 12 analyzes the reference image and identifies the equipment worn by the detected predetermined worker.

[0049] The predetermined worker is a worker who is considered to be wearing the intended wearable equipment. For example, the predetermined worker may be, but is not limited to, an expert, a leader, a model worker, etc.

[0050] The feature quantities of the appearance of a predetermined worker are information for detecting the worker by image analysis, such as facial information, physique, height, gait, etc. The process of detecting the predetermined worker from an image based on these feature quantities can be realized by any means.

[0051] The process of identifying the detected garment worn by the specified worker is the same as the "process of identifying the garment worn by the worker shown in the reference image" described in the second embodiment.

[0052] The template information generating unit 13 generates attachment template information indicating attachments that should be worn by the worker based on the analysis result of the reference image, and stores the template information in the storage unit 17 .

[0053] In this embodiment, the reference image acquisition unit 11 acquires an image taken at a work site as a reference image. Then, the reference image analysis unit 12 detects a "predetermined worker who is thought to be wearing an attachment that should be worn" from the image taken at the work site and identifies the attachment of that worker. In this way, the reference image analysis unit 12 identifies the attachment of the predetermined worker that appears in the image taken at the work site.

[0054] Based on the analysis results of the images taken at the work site, the template information generation unit 13 generates clothing template information that indicates the “wear identified by the reference image analysis unit 12 (wear worn by a specified worker)” as “wear that should be worn by the worker.”

[0055] The determination device 10 of this embodiment may have the same configuration as the determination device 10 of the second embodiment. That is, the determination device 10 of this embodiment may include both the means for generating attachment template information described in the second embodiment and the means for generating attachment template information described in this embodiment.

[0056] Next, an example of the processing flow of the determination device 10 will be described with reference to the flowchart of FIG.

[0057] First, the determination device 10 acquires an image taken at the work site as a reference image (S10). Next, the determination device 10 detects a predetermined worker from the image taken at the work site (reference image) based on the appearance features of the predetermined worker registered in advance. The determination device 10 then analyzes the reference image and identifies the clothing worn by the detected predetermined worker (S11). The determination device 10 then generates and registers (stores in the storage unit 17) clothing template information that indicates the "wear identified in S11" as "wear that should be worn by the worker" (S12).

[0058] "Operations and Effects" According to the determination device 10 of the third embodiment, the same operations and effects as those of the determination device 10 of the first embodiment are realized.

[0059] Furthermore, the determination device 10 of the third embodiment detects "a specific worker who is thought to be wearing the appropriate garment" from images captured at the work site, and generates garment template information based on the results of identifying the garment worn by that worker. The determination device 10 of the third embodiment can generate garment template information that appropriately indicates the appropriate garment. Furthermore, with the determination device 10 of the third embodiment, there is no need to stop or call the specific worker to capture a reference image. This reduces the burden on the specific worker.

[0060] Furthermore, when the determination device 10 of the third embodiment has the same configuration as the determination device 10 of the second embodiment, it can achieve the same effects.

[0061] Fourth Embodiment "Overview" The determination device 10 of the fourth embodiment is a specific implementation of the determination device 10 of the first embodiment. More specifically, the determination device 10 of the fourth embodiment is a specific implementation of a process for generating equipment template information. The determination device 10 of the fourth embodiment analyzes images (reference images) captured at a work site, identifies the equipment worn by multiple workers, and generates equipment template information based on the results of this identification. The configuration of the determination device 10 will be described in detail below.

[0062] "Hardware Configuration" This is the same as the determination device 10 of the second embodiment, so a description thereof will be omitted here.

[0063] "Functional Configuration" Next, the functional configuration of the determination device 10 of this embodiment will be described in detail. FIG. 1 shows an example of a functional block diagram of the determination device 10. As shown in the figure, the determination device 10 has a reference image acquisition unit 11, a reference image analysis unit 12, a template information generation unit 13, a target image acquisition unit 14, a target image analysis unit 15, a determination unit 16, and a storage unit 17. Note that the determination device 10 does not necessarily have to have the storage unit 17. In this case, an external device configured to be able to communicate with the determination device 10 has the storage unit 17.

[0064] The configurations of the target image acquisition unit 14, the target image analysis unit 15, and the determination unit 16 are the same as those in the first embodiment. The configurations of the reference image acquisition unit 11, the reference image analysis unit 12, the template information generation unit 13, and the storage unit 17 will be described below.

[0065] The reference image acquisition unit 11 acquires a reference image. The reference image is an image in which a worker is photographed. For example, as shown in Fig. 5, the reference image acquisition unit 11 acquires an image of a worker working at a work site as the reference image. The means for acquiring such a reference image has been described in the third embodiment, and therefore will not be described here.

[0066] The reference image analysis unit 12 identifies the items worn by each of the workers who appear in the images (reference images) taken at the work site. The process of identifying the items worn by the workers is the same as the "process of identifying the items worn by the workers who appear in the reference images" described in the second embodiment.

[0067] The template information generating unit 13 generates attachment template information based on the results of identification by the reference image analyzing unit 12 and stores it in the storage unit 17 .

[0068] The template information generating unit 13 can generate clothing template information indicating “wear worn by a predetermined number or more of the workers among the multiple workers appearing in the reference image” as clothing that should be worn by the workers.

[0069] Alternatively, the template information generation unit 13 can generate, as the attachment that should be worn by the worker, attachment template information that indicates “attachment worn by a predetermined percentage or more of the workers appearing in the reference image.”

[0070] It is believed that the majority of workers at a work site are wearing the appropriate accessories, and therefore the template information generating unit 13 generates appropriate accessory template information through the above process.

[0071] The determination device 10 of this embodiment may have the configuration of at least one of the determination devices 10 of the second and third embodiments. That is, the determination device 10 of this embodiment may have at least one of the means for generating outfit template information described in the second embodiment and the means for generating outfit template information described in the third embodiment, and the means for generating outfit template information described in this embodiment.

[0072] Next, an example of the processing flow of the determination device 10 will be described with reference to the flowchart of FIG.

[0073] First, the determination device 10 acquires an image taken at the work site as a reference image (S10). Next, the determination device 10 identifies the clothing worn by each of the multiple workers appearing in the image taken at the work site (the reference image) (S11). Then, the determination device 10 generates and registers (stores in the storage unit 17) clothing template information based on the results of the identification in S11 (S12). Specifically, the determination device 10 generates clothing template information indicating clothing worn by at least a predetermined number of workers among the multiple workers appearing in the reference image, or clothing worn by at least a predetermined percentage of workers, as clothing that should be worn by the workers.

[0074] "Operational Effects" According to the determination device 10 of the fourth embodiment, operational effects similar to those of the determination device 10 of the first embodiment are achieved.

[0075] Furthermore, the determination device 10 of the fourth embodiment identifies the clothing worn by each of multiple workers based on images captured at the work site and generates clothing template information based on the measurement results. It is believed that the majority of workers at a work site are wearing the appropriate clothing. Therefore, the determination device 10 of the fourth embodiment configured as described above can generate clothing template information that appropriately indicates the appropriate clothing. Furthermore, with the determination device 10 of the fourth embodiment, there is no need to stop or call a specific worker to capture a reference image. This reduces the burden on the specific worker.

[0076] Furthermore, if the determination device 10 of the fourth embodiment has a configuration similar to that of at least one of the determination device 10 of the second embodiment and the determination device 10 of the third embodiment, it can achieve the same effects.

[0077] Fifth Embodiment "Overview" The accessories that a worker should wear may differ depending on the type of work. The determination device 10 of the fifth embodiment includes a means for generating accessory template information for each type of work, in addition to the means for generating accessory template information described in the second embodiment. This will be described in detail below.

[0078] "Hardware Configuration" This is the same as the determination device 10 of the second embodiment, so a description thereof will be omitted here.

[0079] "Functional Configuration" Next, the functional configuration of the determination device 10 of this embodiment will be described in detail. FIG. 1 shows an example of a functional block diagram of the determination device 10. As shown in the figure, the determination device 10 has a reference image acquisition unit 11, a reference image analysis unit 12, a template information generation unit 13, a target image acquisition unit 14, a target image analysis unit 15, a determination unit 16, and a storage unit 17. Note that the determination device 10 does not necessarily have to have the storage unit 17. In this case, an external device configured to be able to communicate with the determination device 10 has the storage unit 17.

[0080] The configurations of the target image acquisition unit 14, the target image analysis unit 15, and the determination unit 16 are the same as those in the first embodiment. The configurations of the reference image acquisition unit 11, the reference image analysis unit 12, the template information generation unit 13, and the storage unit 17 will be described below.

[0081] The reference image acquisition unit 11 acquires a reference image for each work type. The reference image associated with each work type is an image of a worker wearing an attachment that should be worn when performing the work of that work type.

[0082] For example, a person may input a reference image into the determination device 10, linking it to a work type. Alternatively, a person may link a work type to a reference image acquired by the reference image acquisition unit 11. The "work type" is the type of work, and examples include high-altitude work, loading work, and machine operation work. There are various ways to classify work types, and no particular restrictions are imposed. The other configurations of the reference image acquisition unit 11 are as described in the second embodiment, so description thereof will be omitted here.

[0083] Here are examples of the equipment to be worn for each type of work. For work at heights, a safety belt is an example. For welding work, a face mask is an example. For work involving the handling of specific chemicals (special chemicals), a lab coat or protective mask is an example. For work involving contact with specific objects (hot objects, blades, etc.), gloves are an example. For work involving electrical systems, rubber gloves are an example. For work in a clean room, dustproof clothing is an example. For work involving the handling of heavy objects, safety shoes and a helmet are an example.

[0084] The configuration of the reference image analysis unit 12 is the same as that explained in the second embodiment, and therefore, explanation thereof will be omitted here.

[0085] The template information generating unit 13 generates, for each work type, attachment template information indicating attachments that should be worn by the worker, and stores the template information in the storage unit 17. In the attachment template information associated with each work type, attachments identified by analyzing the reference image associated with each work type are indicated as attachments that should be worn.

[0086] An example of the accessory template information is shown in Fig. 6. The accessory template information shown in Fig. 6 includes items such as template identification information, work type, and accessory.

[0087] The template identification information field contains information that distinguishes between multiple pieces of equipment template information. The work type field contains the details of the work type. The equipment field contains the equipment that should be worn by a worker performing each work type.

[0088] The other configuration of the template information generating unit 13 is the same as that described in the second embodiment, and therefore description thereof will be omitted here.

[0089] Next, an example of the processing flow of the determination device 10 will be described with reference to the flowchart of FIG.

[0090] First, the determination device 10 acquires a reference image of a worker wearing the appropriate garment for each work type (S10). Next, the determination device 10 analyzes the reference image for each work type and identifies the garment worn by the "worker wearing the appropriate garment" appearing in each reference image (S11). Then, the determination device 10 generates garment template information for each work type, indicating the "garment identified in S11" as the "garment to be worn by the worker," and registers the information (stored in the storage unit 17) (S12).

[0091] "Operations and Effects" According to the determination device 10 of the fifth embodiment, the same operations and effects as those of the determination devices 10 of the first and second embodiments are realized.

[0092] Furthermore, the accessories that a worker should wear may differ depending on the type of work. According to the determination device 10 of the fifth embodiment, it is possible to generate accessory template information that appropriately indicates the accessories that a worker should wear for each type of work.

[0093] Sixth Embodiment Overview The accessories that a worker should wear may differ depending on the type of work. The determination device 10 of the sixth embodiment includes a means for generating accessory template information for each type of work, in addition to the means for generating accessory template information described in the third embodiment. This will be described in detail below.

[0094] "Hardware Configuration" This is the same as the determination device 10 of the second embodiment, so a description thereof will be omitted here.

[0095] "Functional Configuration" Next, the functional configuration of the determination device 10 of this embodiment will be described in detail. FIG. 1 shows an example of a functional block diagram of the determination device 10. As shown in the figure, the determination device 10 has a reference image acquisition unit 11, a reference image analysis unit 12, a template information generation unit 13, a target image acquisition unit 14, a target image analysis unit 15, a determination unit 16, and a storage unit 17. Note that the determination device 10 does not necessarily have to have the storage unit 17. In this case, an external device configured to be able to communicate with the determination device 10 has the storage unit 17.

[0096] The configurations of the target image acquisition unit 14, the target image analysis unit 15, and the determination unit 16 are the same as those in the first embodiment. The configurations of the reference image acquisition unit 11, the reference image analysis unit 12, the template information generation unit 13, and the storage unit 17 will be described below.

[0097] The configuration of the reference image acquisition unit 11 is the same as that explained in the third embodiment, and therefore, explanation thereof will be omitted here.

[0098] As described in the third embodiment, the reference image analysis unit 12 detects a predetermined worker from an image (reference image) taken at a work site based on the appearance feature amounts of the predetermined worker registered in advance, and then identifies the worker's clothing using the means described in the third embodiment.

[0099] Furthermore, the reference image analysis unit 12 analyzes the reference image and identifies the type of work being performed by the detected specific worker. The type of work can be identified based on the worker's posture, the tools and machines used by the worker, the object being worked on, the number of workers, etc. The process of identifying the type of work being performed by the worker appearing in the reference image through image analysis can be realized using any technology. For example, a learning model that has learned the above characteristics of various types of work through machine learning may be used.

[0100] The template information generating unit 13 generates, for each work type, attachment template information indicating attachments that should be worn by the worker, and stores the template information in the storage unit 17 .

[0101] The template information generator 13 generates clothing template information that indicates "the clothing worn by the predetermined worker identified by the reference image analyzer 12" as "the clothing to be worn by the worker performing the work type identified by the reference image analyzer 12." An example of the clothing template information is shown in FIG. 6. The clothing template information in FIG. 6 has been described in the fifth embodiment, so a description thereof will be omitted here.

[0102] The other configuration of the template information generating unit 13 is the same as that described in the third embodiment, and therefore description thereof will be omitted here.

[0103] Next, an example of the processing flow of the determination device 10 will be described with reference to the flowchart of FIG.

[0104] First, the determination device 10 acquires an image taken at a work site as a reference image (S10). Next, the determination device 10 detects a predetermined worker from the image taken at the work site (reference image) based on the appearance features of the predetermined worker registered in advance. The determination device 10 then analyzes the reference image and identifies the equipment worn by the detected predetermined worker (S11). Also in S11, the determination device 10 analyzes the reference image and identifies the type of work being performed by the detected predetermined worker.

[0105] Then, the determination device 10 generates and registers (stores in the memory unit 17) accessory template information that indicates the "accessory identified in S11" as "accessory that should be worn by a worker performing the work type identified in S11" (S12).

[0106] "Operations and Effects" According to the determination device 10 of the sixth embodiment, the same operations and effects as those of the determination devices 10 of the first and third embodiments are realized.

[0107] Furthermore, the accessories that a worker should wear may differ depending on the type of work. According to the determination device 10 of the sixth embodiment, it is possible to generate accessory template information that appropriately indicates the accessories that a worker should wear for each type of work.

[0108] Seventh Embodiment "Overview" The accessories that a worker should wear may differ depending on the type of work. The determination device 10 of the seventh embodiment includes a means for generating accessory template information for each type of work, in addition to the means for generating accessory template information described in the fourth embodiment. This will be described in detail below.

[0109] "Hardware Configuration" This is the same as the determination device 10 of the second embodiment, so a description thereof will be omitted here.

[0110] "Functional Configuration" Next, the functional configuration of the determination device 10 of this embodiment will be described in detail. FIG. 1 shows an example of a functional block diagram of the determination device 10. As shown in the figure, the determination device 10 has a reference image acquisition unit 11, a reference image analysis unit 12, a template information generation unit 13, a target image acquisition unit 14, a target image analysis unit 15, a determination unit 16, and a storage unit 17. Note that the determination device 10 does not necessarily have to have the storage unit 17. In this case, an external device configured to be able to communicate with the determination device 10 has the storage unit 17.

[0111] The configurations of the target image acquisition unit 14, the target image analysis unit 15, and the determination unit 16 are the same as those in the first embodiment. The configurations of the reference image acquisition unit 11, the reference image analysis unit 12, the template information generation unit 13, and the storage unit 17 will be described below.

[0112] The configuration of the reference image acquisition unit 11 is the same as that explained in the fourth embodiment, and therefore, explanation thereof will be omitted here.

[0113] The reference image analysis unit 12 identifies the equipment worn by each of the multiple workers who appear in the image (reference image) taken at the work site. This process is as described in the fourth embodiment. The reference image analysis unit 12 also identifies the type of work being performed by each of the multiple workers who appear in the image (reference image) taken at the work site. The process of identifying the type of work by image analysis is as described in the sixth embodiment.

[0114] The template information generating unit 13 generates, for each work type, attachment template information indicating attachments that should be worn by the worker, and stores the template information in the storage unit 17 .

[0115] The template information generation unit 13 identifies, for each work type, the attachments worn by a predetermined number or more of the workers among the "workers performing each work type who appear in the reference image," or the attachments worn by a predetermined percentage or more of the workers. Then, the template information generation unit 13 generates attachment template information indicating the attachments identified for each work type as attachments to be worn by workers performing each work type.

[0116] Next, an example of the processing flow of the determination device 10 will be described with reference to the flowchart of FIG.

[0117] First, the determination device 10 acquires an image taken at the work site as a reference image (S10). Next, the determination device 10 identifies the items worn by each of the multiple workers who appear in the image taken at the work site (reference image) (S11). In addition, in S11, the determination device 10 identifies the type of work being performed by each of the multiple workers who appear in the image taken at the work site (reference image).

[0118] Then, the determination device 10 generates and registers (stores in the storage unit 17) accessory template information based on the results of the identification in S11 (S12). Specifically, the determination device 10 generates accessory template information indicating accessory worn by a predetermined number or more of the "workers performing each type of work in the reference image," or accessory worn by a predetermined percentage or more of the "workers performing each type of work."

[0119] "Operational Effects" According to the determination device 10 of the seventh embodiment, operational effects similar to those of the determination devices 10 of the first and fourth embodiments are realized.

[0120] Furthermore, the accessories that a worker should wear may differ depending on the type of work. According to the determination device 10 of the seventh embodiment, it is possible to generate accessory template information that appropriately indicates the accessories that a worker should wear for each type of work.

[0121] Eighth Embodiment "Overview" The determination device 10 of the eighth embodiment is a specific implementation of the determination device 10 of the first embodiment. More specifically, the determination device 10 of the eighth embodiment is a specific implementation of a process for determining whether a worker to be determined is wearing an attachment that should be worn. The configuration of the determination device 10 will be described in detail below.

[0122] "Hardware Configuration" This is the same as the determination device 10 of the second embodiment, so a description thereof will be omitted here.

[0123] "Functional Configuration" Next, the functional configuration of the determination device 10 of this embodiment will be described in detail. FIG. 1 shows an example of a functional block diagram of the determination device 10. As shown in the figure, the determination device 10 has a reference image acquisition unit 11, a reference image analysis unit 12, a template information generation unit 13, a target image acquisition unit 14, a target image analysis unit 15, a determination unit 16, and a storage unit 17. Note that the determination device 10 does not necessarily have to have the storage unit 17. In this case, an external device configured to be able to communicate with the determination device 10 has the storage unit 17.

[0124] The configurations of the reference image acquisition unit 11, the reference image analysis unit 12, and the template information generation unit 13 are the same as those of any of the first to fourth embodiments. That is, the determination device 10 of this embodiment generates attachment template information using any of the means described in the first to fourth embodiments and stores it in the storage unit 17. The configurations of the target image acquisition unit 14, the target image analysis unit 15, and the determination unit 16 will be described below.

[0125] The target image acquisition unit 14 acquires a target image in which the worker to be judged appears.

[0126] The target image may be, for example, an image taken by having a worker stand in front of a camera. Such a target image can be acquired by the same means as for acquiring the reference image described in the second embodiment. This method is ideal for identifying each worker at any time, such as before the start of work.

[0127] Alternatively, the target image may be, for example, an image of a worker working at a work site. Such a target image can be acquired by the same means as the acquisition of the reference image described in the third embodiment. This method is ideal for monitoring a worker at work.

[0128] The target image analysis unit 15 analyzes the target image and identifies the items worn by the worker appearing in the target image. Identification of the items worn by the worker appearing in the target image is realized, for example, by a method similar to that used to identify the items worn by the worker appearing in the reference image described in the second embodiment.

[0129] The determination unit 16 determines whether the worker to be determined is wearing the appropriate wear based on the wear identified in the analysis of the target image by the target image analysis unit 15 and the wear indicated in the wear template information. Specifically, the determination unit 16 determines whether the wear indicated in the wear template information is included among the wear identified in the analysis of the target image by the target image analysis unit 15. If at least one of the wear indicated in the wear template information is not included, the determination unit 16 determines that the worker to be determined is not wearing the appropriate wear. On the other hand, if all of the wear indicated in the wear template information are included, the determination unit 16 determines that the worker to be determined is wearing the appropriate wear.

[0130] The determination unit 16 can output the determination result. Below, output examples will be described for the case where the determination unit 16 is used to determine workers one by one at any timing, and for the case where the determination unit 16 is used to monitor workers while they are working.

[0131] In the case of an application in which workers are judged one by one at any timing, as described above, the worker is asked to stand in front of the camera, and a target image is acquired. The judgment unit 16 can output an image in which information indicating the judgment result is superimposed on the target image, as shown in Figures 7 and 8.

[0132] 7 shows an image that is output when it is determined that the worker being assessed is wearing the appropriate accessory. In the example of Fig. 7, information "Assessment result OK" is superimposed on the target image.

[0133] Fig. 8 shows an image that is output when it is determined that the worker being assessed is not wearing the appropriate accessory. In the example of Fig. 8, information indicating "Assessment Result NG" is superimposed on the target image. Furthermore, in the example of Fig. 8, information indicating that the worker is not wearing an accessory that should be worn is superimposed on the target image. Specifically, in the example of Fig. 8, it is pointed out that the worker has forgotten to wear a "helmet."

[0134] For example, a display may be installed near a camera that captures an image of the worker to be assessed. The assessment unit 16 may then display the assessment results on this display, as shown in Fig. 7 or 8. In this case, the process from capturing an image of the worker to be assessed to displaying the assessment results is performed in real time.

[0135] Next, in the case of an application for monitoring a worker at work, as described above, a target image of a worker working at a work site is acquired. The determination unit 16 can output an image in which information indicating the determination result is superimposed on the target image, as shown in Fig. 9 .

[0136] 9, information identifying the worker who has been determined not to be wearing the appropriate accessory is superimposed on the target image. Furthermore, in the example of Fig. 9, information indicating the accessory that should be worn but is not being worn is superimposed on the target image.

[0137] In this example, the determination unit 16 constantly performs a determination process on the worker appearing in the target image. When a worker is detected not wearing the appropriate attachment, an image such as that shown in Fig. 9 is output. For example, an image (moving image) of a worker working at a work site may be displayed in real time on a monitor's display. The determination unit 16 may then superimpose information such as that shown in Fig. 9 on the image displayed on the monitor's display.

[0138] Alternatively, when the determination unit 16 detects a worker not wearing a required attachment, the determination unit 16 may generate a still image as shown in FIG. 9 . The still image is a captured image (moving image) of the worker working at the work site, with predetermined information superimposed on it. The determination unit 16 may then send the generated still image to a pre-registered destination. For example, the monitor's terminal, email address, or the like may be registered as the destination.

[0139] Next, an example of the flow of processing by the determination device 10 will be described with reference to the flowchart of FIG.

[0140] When the determination device 10 acquires a target image showing a worker to be determined (S20), it analyzes the target image and identifies the worker's accessories shown in the target image (S21). Next, the determination device 10 determines whether the worker to be determined is wearing the accessories that should be worn, based on the accessories identified by the analysis of the target image and the accessories indicated in the accessories template information (S22). The determination device 10 can then output the determination result.

[0141] "Effects" The determination device 10 of the eighth embodiment achieves the same effects as the determination devices 10 of the first and fourth embodiments. Furthermore, the determination device 10 can appropriately determine whether the worker to be determined is wearing the appropriate attachment.

[0142] Ninth Embodiment "Overview" The determination device 10 of the ninth embodiment is a specific implementation of the determination device 10 of the first embodiment. More specifically, the determination device 10 of the ninth embodiment is a specific implementation of a process for determining whether a worker to be determined is wearing an appropriate piece of clothing. The determination device 10 of the ninth embodiment differs from the determination device 10 of the eighth embodiment in that it identifies the type of work to be performed by the worker to be determined and makes a determination using clothing template information linked to the identified type of work. The configuration of the determination device 10 will be described in detail below.

[0143] "Hardware Configuration" This is the same as the determination device 10 of the second embodiment, so a description thereof will be omitted here.

[0144] "Functional Configuration" Next, the functional configuration of the determination device 10 of this embodiment will be described in detail. FIG. 1 shows an example of a functional block diagram of the determination device 10. As shown in the figure, the determination device 10 has a reference image acquisition unit 11, a reference image analysis unit 12, a template information generation unit 13, a target image acquisition unit 14, a target image analysis unit 15, a determination unit 16, and a storage unit 17. Note that the determination device 10 does not necessarily have to have the storage unit 17. In this case, an external device configured to be able to communicate with the determination device 10 has the storage unit 17.

[0145] The configurations of the reference image acquisition unit 11, the reference image analysis unit 12, and the template information generation unit 13 are the same as those of any of the fifth to seventh embodiments. That is, the determination device 10 of this embodiment generates attachment template information using any of the means described in the fifth to seventh embodiments and stores it in the storage unit 17. The configurations of the target image acquisition unit 14, the target image analysis unit 15, and the determination unit 16 will be described below.

[0146] The target image acquisition unit 14 acquires a target image in which the worker to be judged is photographed. The target image acquisition unit 14 can acquire the target image by the means described in the eighth embodiment.

[0147] The target image analysis unit 15 analyzes the target image and identifies the items worn by the worker appearing in the target image. Identification of the items worn by the worker appearing in the target image is realized, for example, by a method similar to that used to identify the items worn by the worker appearing in the reference image described in the second embodiment.

[0148] Furthermore, the target image analysis unit 15 identifies the type of work being performed by the worker appearing in the target image.

[0149] For example, when the target image acquisition unit 14 acquires a target image by having a worker stand in front of a camera, the worker is asked to input a task type at the time of capturing the image. The input is realized via an input device (touch panel, microphone, operation button, etc.) installed near the camera. Then, information indicating the specified task type is linked to the target image and input to the determination device 10. Based on this information, the target image analysis unit 15 identifies the task type of the task being performed by the worker appearing in the target image.

[0150] In addition, when the target image acquisition unit 14 acquires a target image of a worker working at a work site, the target image analysis unit 15 analyzes the target image and identifies the type of work being performed by the worker appearing in the target image. The process of identifying the type of work by image analysis is as described in the sixth embodiment.

[0151] The determination unit 16 determines whether the worker being evaluated is wearing the appropriate equipment, based on the equipment identified by the target image analysis unit 15 and the equipment indicated in the equipment template information linked to the type of work identified by the target image analysis unit 15.

[0152] Specifically, the determination unit 16 extracts the equipment template information linked to the work type identified by the target image analysis unit 15 from the pre-registered equipment template information for each work type (see FIG. 6 ). Then, the determination unit 16 determines whether the target worker is wearing the equipment that should be worn, based on the equipment of the target worker identified by the analysis of the target image by the target image analysis unit 15 and the equipment indicated by the extracted equipment template information. Details of the determination process are as described in the eighth embodiment.

[0153] Furthermore, the determination unit 16 can output the determination result using the means described in the eighth embodiment.

[0154] Next, an example of the flow of processing by the determination device 10 will be described with reference to the flowchart of FIG.

[0155] When the determination device 10 acquires a target image showing a worker to be determined (S20), it analyzes the target image and identifies the worker's accessories shown in the target image (S21). In S21, the determination device 10 also identifies the type of work being performed by the worker shown in the target image. Next, the determination device 10 determines whether the worker to be determined is wearing the appropriate accessories based on the accessories identified by the analysis of the target image and the accessories indicated in the accessories template information associated with the identified type of work (S22). The determination device 10 can then output the determination result.

[0156] "Effects" The determination device 10 of the ninth embodiment achieves the same effects as the determination devices 10 of the fifth and seventh embodiments. Furthermore, the determination device 10 can appropriately determine whether the worker to be determined is wearing an attachment that should be worn.

[0157] <Modifications> Modifications applicable to the first to ninth embodiments will be described below.

[0158] "Variation 1" The target image analysis unit 15 distinguishes between multiple workers appearing in the target image based on the feature amounts of the workers' appearances. The feature amounts of the appearances of each of the multiple workers may be registered in advance. Then, the target image analysis unit 15 may distinguish between multiple workers appearing in the target image based on the feature amounts.

[0159] Then, as shown in FIG. 11, the determination unit 16 registers, for each worker, a history of determination results indicating that the worker is not wearing the attachment that should be worn.

[0160] In the information shown in FIG. 11, worker identification information and defect history are linked to each other.

[0161] The worker identification information field is used to input information identifying the worker. The defect history field is used to input a history of when it was determined that the appropriate attachment was not being worn. In the example shown in the figure, the date and time when it was determined that the appropriate attachment was not being worn, and the attachment that was not being worn are displayed.

[0162] Based on such a history, it is possible to determine which workers are not wearing their accessories properly, which items each worker tends to forget to wear, and so on.

[0163] "Variation 2" The determination device 10 can determine whether the worker to be determined is wearing the appropriate attachment, based on an image of the worker at the start of work (reference image) and an image of the worker during work (target image).

[0164] The reference image acquisition unit 11 acquires a reference image before or immediately after work begins on each work day. The reference image analysis unit 12 analyzes the reference image acquired at this timing. The template information generation unit 13 then generates attachment-related plate information based on the analysis results of the reference image acquired at this timing.

[0165] The target image acquisition unit 14 acquires target images taken at the work site during work on each work day. The target image analysis unit 15 analyzes the target images acquired at such timing. The determination unit 16 then determines whether the worker being evaluated is wearing the appropriate attachment based on the analysis results of the target images acquired at such timing.

[0166] "Variation 3" The target image acquisition unit 14 acquires a target image of a worker working, photographed at a work site. In this way, when making a judgment based on a target image of a worker working, an attachment that is actually being worn may not be detected due to the worker's orientation, posture, work content, etc., and it may be erroneously determined that the worker is not wearing the attachment.

[0167] Therefore, the target image analysis unit 15 uses a known tracking technique to track the worker detected in the target image (moving image), and then makes a judgment on the worker over a certain period of time.

[0168] The determination unit 16 determines whether the worker is wearing the appropriate accessory based on the determination results within a certain period of time. Specifically, if the determination unit 16 determines that the worker has not been wearing a certain accessory for the entire certain period of time, it determines that the worker is not wearing the accessory. On the other hand, if the determination unit 16 determines that the worker is not wearing a certain accessory at a certain time during the certain period of time, but determines that the worker is wearing the accessory at another time during the certain period of time, it determines that the worker is wearing the accessory. This configuration can reduce erroneous determinations.

[0169] "Modification 4" The target image acquisition unit 14 acquires target images of workers taken at the work site. In Modification 4, time periods that are not subject to judgment are registered in advance in the judgment device 10. For example, time periods when no work is being performed, such as break times and lunch times, are not subject to judgment.

[0170] Even if the determination unit 16 determines that the worker to be determined is not wearing the appropriate attachment based on the analysis result of the target image captured during a time period other than the target time period for determination, the determination unit 16 does not output a warning as shown in Fig. 9. This configuration can reduce the inconvenience of unnecessarily outputting a warning during a time period other than the target time period for determination when no work is being performed.

[0171] "Modification 5" The target image acquisition unit 14 acquires target images of workers photographed at the work site. In Modification 5, areas that are not subject to judgment are registered in advance in the judgment device 10. For example, areas where no work is performed, such as rest areas and toilets, are not subject to judgment.

[0172] The target image analysis unit 15 further determines whether the area in the target image where the worker is located is an "area outside the scope of determination." If the camera is a fixed-point camera, the area outside the scope of determination that exists in the image captured by the camera may be identified and registered in advance using the coordinates of the image, etc. In this example, the user can specify the area outside the scope of determination in the image.

[0173] "Variation 6" The target image acquisition unit 14 acquires a target image that shows a worker taken at a work site. The target image analysis unit 15 has a means for distinguishing between a worker and a person other than a worker. That is, the target image analysis unit 15 has a means for determining whether a person shown in the target image is a worker or not.

[0174] In the sixth modification, only for a person identified as a worker, the target image analysis unit 15 identifies an attachment and the determination unit 16 determines whether the person is wearing the attachment that should be worn. In other words, for a person determined not to be a worker, the target image analysis unit 15 does not identify an attachment and the determination unit 16 does not determine whether the person is wearing the attachment that should be worn.

[0175] Depending on the orientation and position of the camera capturing the work site, there is a possibility that passersby passing near the work site may appear in the target image. The above-mentioned judgment may be made regarding such passersby, and a warning may be output. This modification can prevent such inconvenience.

[0176] There are no particular limitations on the means for distinguishing between workers and other people. For example, the feature amounts of the appearance of each worker may be registered in advance. Then, the target image analysis unit 15 may use the feature amounts to realize the above-mentioned distinction.

[0177] Although the embodiments of the present invention have been described above with reference to the drawings, these are merely examples of the present invention, and various other configurations may be adopted. The configurations of the above-described embodiments may be combined with each other, or some of the configurations may be replaced with other configurations. Furthermore, various modifications may be made to the configurations of the above-described embodiments without departing from the spirit of the invention. Furthermore, the configurations and processes disclosed in the above-described embodiments and modified examples may be combined with each other.

[0178] In addition, in the flowcharts used in the above description, multiple steps (processes) are described in order. However, the order of the steps performed in each embodiment is not limited to the order described. In each embodiment, the order of the steps shown in the drawings can be changed to the extent that the content is not affected. Furthermore, the above-described embodiments can be combined to the extent that the content is not contradictory.

[0179] Some or all of the above embodiments may be described as, but are not limited to, the following notes: 1. A determination device comprising: a reference image acquisition means for acquiring a reference image; a reference image analysis means for analyzing the reference image and identifying equipment worn by a worker appearing in the reference image; a template information generation means for generating equipment template information indicating equipment to be worn by the worker based on the analysis results of the reference image and storing the information in a storage means; a target image acquisition means for acquiring a target image in which a worker to be determined is appearing; a target image analysis means for analyzing the target image and identifying equipment worn by the worker appearing in the target image; and a determination means for determining whether the worker to be determined is wearing the equipment to be worn, based on the equipment identified in the analysis of the target image and the equipment indicated in the equipment template information. 2. The determination device according to 1, wherein the reference image acquisition means acquires an image taken at a work site as the reference image, the reference image analysis means identifies clothing worn by a worker appearing in the image taken at the work site, and the template information generation means generates the clothing template information based on the analysis results of the image taken at the work site. 3. The determination device according to 2, wherein the reference image analysis means detects a predetermined worker from the image taken at the work site based on appearance features of the predetermined worker registered in advance and identifies the detected clothing worn by the predetermined worker, and the template information generation means generates the clothing template information indicating the clothing worn by the predetermined worker as clothing that should be worn by the worker. 3. The determination device according to 2, wherein the reference image analysis means identifies the equipment worn by each of a plurality of workers appearing in the image taken at the work site, and the template information generation means generates the equipment template information indicating, as the equipment to be worn by the workers, equipment worn by a predetermined number or more of the plurality of workers appearing in the image taken at the work site, or equipment worn by a predetermined percentage or more of the workers.5. The determination device according to any one of 1 to 4, wherein the reference image analysis means analyzes the reference image and identifies a type of work being performed by the worker appearing in the reference image, and the template information generation means generates the equipment template information for each type of work based on the analysis results of the reference image. 6. The determination device according to 5, wherein the target image analysis means analyzes the target image and identifies a type of work being performed by the worker appearing in the target image, and the determination means determines whether the worker to be determined is wearing appropriate equipment based on the equipment identified in the analysis of the target image and the equipment indicated in the equipment template information corresponding to the type of work identified in the analysis of the target image. 7. The determination device according to any one of 1 to 6, wherein the target image analysis means distinguishes between multiple workers appearing in the target image based on features of the worker's appearance, and the determination means registers a history of determination results for each worker that the worker is not wearing appropriate equipment. a determination method in which one or more computers acquire a reference image, analyze the reference image, and identify equipment worn by a worker appearing in the reference image; generate equipment template information indicating equipment that should be worn by the worker based on a result of the analysis of the reference image, and store the information in a storage means; acquire a target image in which the worker to be determined is appearing, analyze the target image, and identify equipment worn by the worker appearing in the target image; and determine whether the worker to be determined is wearing the equipment that should be worn based on the equipment identified in the analysis of the target image and the equipment indicated in the equipment template information.9. A program causing a computer to function as: reference image acquisition means for acquiring a reference image; reference image analysis means for analyzing the reference image and identifying equipment worn by a worker appearing in the reference image; template information generation means for generating equipment template information indicating equipment to be worn by the worker based on the analysis results of the reference image and storing the information in storage means; target image acquisition means for acquiring a target image in which the worker to be judged is appearing; target image analysis means for analyzing the target image and identifying equipment worn by the worker appearing in the target image; and determination means for determining whether the worker to be judged is wearing the equipment to be worn, based on the equipment identified in the analysis of the target image and the equipment indicated in the equipment template information.

[0180] This application claims priority based on Japanese Patent Application No. 2023-010955, filed on January 27, 2023, the disclosure of which is incorporated herein in its entirety by reference.

[0181] REFERENCE SIGNS LIST 10 Determination device 11 Reference image acquisition unit 12 Reference image analysis unit 13 Template information generation unit 14 Target image acquisition unit 15 Target image analysis unit 16 Determination unit 17 Storage unit 1A Processor 2A Memory 3A Input / output I / F 4A Peripheral circuit 5A Bus

Claims

1. a reference image acquisition means for acquiring a reference image; a reference image analysis means for analyzing the reference image and identifying the worker's clothing appearing in the reference image; a template information generating means for generating, based on the analysis result of the reference image, clothing template information indicating clothing to be worn by the worker, and storing the clothing template information in a storage means; a target image acquisition means for acquiring a target image in which a worker to be determined is photographed; a target image analysis means for analyzing the target image and identifying the items worn by the worker appearing in the target image; a determination means for determining whether the worker to be determined is wearing an accessory that should be worn, based on the accessory identified by the analysis of the target image and the accessory indicated by the accessory template information; A determination device having the above.

2. The reference image acquisition means An image taken at a work site is acquired as the reference image; The reference image analysis means Identifying items worn by workers in images taken at the work site; The template information generating means The determination device according to claim 1 , wherein the attachment template information is generated based on an analysis result of an image taken at the work site.

3. The reference image analysis means Detecting a predetermined worker from an image taken at the work site based on a feature amount of the appearance of the predetermined worker that has been registered in advance, and identifying an item worn by the detected predetermined worker; The template information generating means The determination device according to claim 2 , wherein the equipment template information indicates an equipment to be worn by the predetermined worker.

4. The reference image analysis means Identifying the items worn by each of the workers shown in the images taken at the work site; The template information generating means 3. The determination device according to claim 2, wherein the equipment template information indicates, as the equipment to be worn by the workers, equipment worn by a predetermined number or more of the workers appearing in the images taken at the work site, or equipment worn by a predetermined percentage or more of the workers.

5. the reference image analysis means analyzes the reference image and identifies the type of work being performed by the worker appearing in the reference image; The determination device according to claim 1 , wherein the template information generating means generates the attachment template information for each type of work based on an analysis result of the reference image.

6. The target image analysis means analyzes the target image and identifies the type of work being performed by the worker appearing in the target image; 6. The determination device according to claim 5, wherein the determination means determines whether the worker to be determined is wearing an appropriate piece of clothing, based on the clothing identified by the analysis of the target image and the clothing indicated in the clothing template information corresponding to the type of work identified by the analysis of the target image.

7. the target image analysis means distinguishes between the plurality of workers appearing in the target image based on the feature amounts of the workers' appearances; The determination device according to claim 1 , wherein the determination unit registers a history of determination results indicating that an attachment that should be worn is not worn for each worker.

8. One or more computers Obtain a reference image, Analyzing the reference image and identifying the worker's clothing shown in the reference image; generating, based on the analysis result of the reference image, an attachment template information indicating an attachment to be worn by the worker, and storing the generated attachment template information in a storage means; Acquire a target image that shows the worker to be judged, Analyzing the target image and identifying the worker's clothing appearing in the target image; determining whether the worker to be assessed is wearing the appropriate accessory based on the accessory identified by the analysis of the target image and the accessory indicated in the accessory template information; Judgment method.

9. the one or more computers: An image taken at a work site is acquired as the reference image; Identifying items worn by workers in images taken at the work site; The method according to claim 8 , further comprising generating the attachment template information based on an analysis result of an image taken at the work site.

10. Computer, a reference image acquisition means for acquiring a reference image; a reference image analysis means for analyzing the reference image and identifying the items worn by the worker appearing in the reference image; a template information generating means for generating, based on the analysis result of the reference image, clothing template information indicating clothing to be worn by the worker, and storing the clothing template information in a storage means; a target image acquisition means for acquiring a target image in which the worker to be determined is photographed; a target image analysis means for analyzing the target image and identifying the items worn by the worker appearing in the target image; a determination means for determining whether the worker to be determined is wearing an appropriate accessory based on the accessory identified by the analysis of the target image and the accessory indicated in the accessory template information; A program that functions as a