JP2020087091A - Skill degree determination system - Google Patents

Abstract

To provide a technique for determining accurately, a skill degree of an operator on the basis of a view point of the operator.SOLUTION: A skill degree determination system comprises: an external imaging part for imaging a visual field video of an operator; and a view point detection part for detecting a view point position of the operator. The skill degree determination system cuts out a frame image from the visual field video, recognizes an operation object for every cut image, identifies for every cut image, a drawing region of the operation object, determines whether or not the view point is positioned in the identified drawing region, and on the basis of the determination result, derives a gaze period in which the operator gazed the operation object over a regulated period and a number at which the view point passed the drawing region of the operation object without gaze in the regulated period, determines whether or not an operation by the operator is proper, and displays the determination result.SELECTED DRAWING: Figure 1

Description

The present invention relates to a technique for determining a skill level regarding maintenance work of equipment by detecting and analyzing a viewpoint of a worker.

Conventionally, for example, education and training on work methods for maintenance work such as elevators is performed using actual machines at limited places such as customer's work site and education and training equipment, or work saved on paper or network It is carried out according to the procedure manual.

As for elevators, new models are being produced one after another and installed at each facility. Although the number of operating machines is not large, some old models installed more than 30 years ago still exist, and there are many types of maintenance targets. Cross over. As a result, the number of educational items is increasing daily, while the number of skilled workers who are familiar with old models is decreasing. In addition to this, since the models for which the production of products has already been completed are also subject to maintenance, there is a problem that it is difficult for a skilled worker who uses a real machine to pass on the technology because of a shortage of the real machine for education.

In addition, there are some items of work that visually judge the wear state and rust state of machine parts, and such work items depend on the skill level. This can lead to variability.

The following are disclosed as related technologies.

JP, 2017-204094, A

It is necessary to pay attention to the target object in the work of visually determining the wear state and rusting state of the mechanical parts. On the other hand, in the technique of Patent Document 1, even when the viewpoint passes over the target object due to fluctuations in the line of sight caused by the operator's inexperienced hesitation, for example, it is counted as the time at which the target object is focused (stop time). Therefore, in the technique of Patent Document 1, it may be determined as an excellent skilled worker without gazing at the object.

The present invention has been made in view of such circumstances of the prior art, and an object thereof is to provide a technique for detecting the viewpoint of an operator and determining the skill level regarding maintenance work with high accuracy based on the viewpoint. It is in.

The skill level determination system of the present invention is a skill level determination system that determines the skill level of a worker, and an external image capturing unit that captures a visual field image of the worker and a viewpoint detection that detects the position of the viewpoint of the worker. Unit and the image captured by the external image capturing unit, the frame images forming the image are respectively cut out, and for each cut out image, the projected work target object is recognized, and the drawing area of the work target object is displayed. The object recognition processing unit specified for each cutout image, and whether the viewpoint detected by the viewpoint detection unit is located in the drawing area specified by the object recognition processing unit is cut out. A matching processing unit that determines each image, based on the determination result of the matching processing unit, the gaze time during which the worker has gazed within the drawing area for more than a specified time, and without gazing for the specified time. Determining the number of times the viewpoint has passed through the drawing area, and a determination unit that determines whether or not the work by the worker has been appropriately performed based on the gaze time and the number of times that the viewpoint has passed, and determination by the determination unit And a display unit for displaying the result.

According to the present invention, the skill level of a worker can be determined with high accuracy based on the viewpoint of the worker.
Problems, configurations, and effects other than those described above will be clarified by the following description of the embodiments.

It is a schematic block diagram which shows an example of a structure of the skill determination system which concerns on embodiment. It is a schematic diagram which shows an example of the use aspect of the skill determination system which concerns on embodiment. It is a figure which shows the hardware structural example of the viewpoint detection device which concerns on embodiment. 6 is a flowchart showing an operation example of the skill level determination system according to the embodiment. It is explanatory drawing which shows an example of the gaze state and passing state (non-gaze state) which concern on embodiment. It is a figure which shows the data structural example of the determination result which concerns on embodiment. 6 is a flowchart illustrating an example of determination processing according to the embodiment. It is a figure which shows an example of the display screen of the determination result which concerns on embodiment.

Hereinafter, embodiments according to the present invention will be described in detail with reference to the drawings.

FIG. 1 is a schematic configuration diagram showing a skill level determination system according to the present embodiment. Further, FIG. 2 is a schematic diagram showing an example of a usage state of the skill level determination system according to the present embodiment. The skill level determination system 1 has a viewpoint detection device 10 and a field terminal 20, which are connected to each other wirelessly or by wire.

In the present embodiment, the viewpoint detection device 10 is made of smart glasses, which are glasses-type wearable terminals, but a known eye tracker that detects the viewpoint of the wearer may be applied. The viewpoint detection device 10 has a viewpoint detection unit 2 and an external imaging unit 3. The viewpoint detection unit 2 acquires an image of the movement of the eyeball of the worker 11 by using, for example, a CCD (Charge Coupled Device) camera, and detects position information of the viewpoint of the worker 11 (hereinafter, referred to as viewpoint coordinates). .. The external imaging unit 3 acquires the in-field image 12 that is a field-of-view image including the work target 13 by using, for example, a CCD camera. The viewpoint coordinates obtained by the viewpoint detection unit 2 and the video data obtained by the external image pickup unit 3 are sequentially transmitted to the on-site terminal 20.

The site terminal 20 is a notebook or tablet computer or smartphone, and has an object recognition processing unit 4, a matching processing unit 5, a determination unit 6, an index value registration unit 7, a display unit 8, and a storage unit. 9 and.

The target object recognition processing unit 4 cuts out frame images constituting the video data captured by the external imaging unit 3 for every several frames, and the work target 13 in the cut-out image and each peripheral component displayed around it. The contours of are extracted respectively. The "peripheral parts" include a part of the body such as a worker's hand and tools used for the work. Then, the target object recognition processing unit 4 specifies what the work target 13 and each peripheral component are, for example, by using an existing automatic learning program, and the work target 13 and the peripheral component occupy in the cutout image. Get the coordinates of the drawing area. The target object recognition processing unit 4 associates the identification information of each identified object with each drawing area coordinate and outputs it to the matching processing unit 5 and the determination unit 6.

The matching processing unit 5 acquires the viewpoint coordinates detected by the viewpoint detection unit 2, the identification information of each projected object (work target 13 and each peripheral component), and each drawing area coordinate. Then, the matching processing unit 5 performs matching processing for determining what the operator 11 was looking by determining in which area the viewpoint coordinates are located. The matching processing unit 5 outputs the identification information of the object viewed by the worker 11 as a determination result.

The determination unit 6 acquires the determination result of the matching processing unit 5 for each cut-out image, and uses the time interval between the cut-out images (here, a specified value) such as the frame rate, on the work target 13 or the work target 13. The time when there is a viewpoint on the peripheral parts or the time when there is no viewpoint is calculated. Further, the determination unit 6 determines whether or not the work by the worker 11 is properly performed by using the index value of the skilled worker, which is stored in advance in the index value registration unit 7. This determination method will be described later.

In addition to the above, the determination unit 6 determines which one of the objects is based on the arrangement information in the image of each object based on the identification information of the work object 13 and each peripheral component and each area coordinate obtained from the object recognition processing unit 4. Determine if you are doing a work item.

The display unit 8 displays the determination result of the determination unit 6 and presents it to the worker 11. The accumulation unit 9 stores and accumulates the determination result of the determination unit 6.

The viewpoint detection device 10 may further include a 3D display unit (not shown) in order to reproduce and display the work target 13 as a 3D model on a VR (virtual reality) when an actual machine is not used for education and training. Good. The skill level determination system 1 may further include an input device (not shown) for operating the 3D model on the VR. There is no limitation on the 3D display unit and the input device, and for example, a known head mounted display may be applied to implement the 3D display of VR.

The functions of the object recognition processing unit 4, the matching processing unit 5, the determination unit 6, the index value registration unit 7, the display unit 8 and the storage unit 9 of the on-site terminal 20 are implemented in the viewpoint detection device 10 (smart glasses). May be. Further, the storage unit 9 may be configured such that a processing device such as a server is separately provided and the processing unit stores the processing device.

FIG. 3 is a diagram illustrating a hardware configuration example of the viewpoint detection device 10. The viewpoint detection device 10 has a controller 101, an input device 110, and an output device 111. The controller 101 controls each hardware that operates inside the viewpoint detection device 10. The controller 101 has the following configuration.

A CPU 102 (CPU: Central Processing Unit) is a processing device that expands a program stored in a flash memory 104 into a RAM 103 (RAM: Random access memory) and executes an operation. The CPU 102 comprehensively controls each hardware inside the controller 101 by executing a program. The RAM 103 is a volatile memory and is a work memory when the CPU 102 processes. The RAM 103 temporarily stores necessary data while the CPU 102 is executing a program.

The flash memory 104 is a data rewritable auxiliary storage device that stores data in a nonvolatile manner. The flash memory 104 stores programs executed by the CPU 102 and control data. In the present embodiment, programs and various data that provide each function described below are installed in the flash memory 104 in advance.

The network I/F 106 is a device that controls data communication performed with an external device such as the field terminal 20. The data communication by the network I/F 106 is performed by wireless communication conforming to a communication standard such as IEEE802.11 system or IEEE802.15.1 system, but may be wired communication by cable connection.

The input I/F 107 is an interface that controls input/output of signals with the input device 110. The output I/F 108 receives an instruction from the CPU 102 and causes the output device 111 to draw an image or output a sound.

The input device 110 is a device group or sensor such as an inward CCD sensor for detecting a viewpoint, an outward CCD sensor for capturing the outside (in-field image), a microphone for collecting sound, and a physical push switch. It is a group. The output device 111 is a speaker for outputting sound, a small projector for displaying an image on the lens portion of smart glasses, or the like.

Since the field terminal 20 is also a conventional computer, the CPU, RAM, flash memory, network I/F, I/F for input/output, various input devices, as well as the hardware configuration shown in FIG. Has an output device. A hard disk drive, for example, may be provided as the auxiliary storage device. Further, the input device may be a mouse, a keyboard or a touch panel. The output device may be a liquid crystal monitor or the like.

Hereinafter, an operation example of the skill level determination system 1 will be described. In addition, here, the training training of the brake maintenance in the maintenance work of the elevator which is the elevator will be described as an example. Further, the work target 13 will be described as a core that is one of sliding parts of the brake.

The elevator brake is one of the important parts because it is responsible for the stopping operation of the elevator. Therefore, regular maintenance work is performed, and the brake is composed of four work items: disassembly, cleaning, inspection, and assembly. The work items include items visually determined, such as important points (vital points) shown in S1 to S6 below. In this embodiment, the skill level determination process is performed for each of these items S1 to S6.
S1. Did you check the main lever for dirt, rust, scratches, wear, and astringency?
S2. Did you check the pins for rust, scratches, wear, and astringency?
S3. Did you confirm the wear of the coil bobbin?
S4. Did you check the core for scratches?
S5. Did you check the push rod for scratches, wear, rust, or looseness?
S6. Did you check the wear of the DU bush?

FIG. 4 is a flowchart showing an operation example of the skill level determination system 1 according to the embodiment. Before the work, the worker 11 first mounts the viewpoint detection device 10 on his/her head and performs a system startup operation. Then, in the in-field-of-view image acquisition step (S001), the external imaging unit 3 sequentially acquires the field-of-view image of the worker 11 and transmits this to the on-site terminal 20.

On the other hand, in the line-of-sight detection step (S002), the viewpoint detection unit 2 detects the viewpoint coordinates of the worker 11 and sequentially transmits them to the on-site terminal 20.

In the target object recognition step (S003), the target object recognition processing unit 4 cuts out image data from the video data captured by the external imaging unit 3 every several frames. Then, the target object recognition processing unit 4 recognizes the work target object 13 and peripheral components from the cut-out in-field image 12. The automatic recognition method is performed, for example, by the object recognition processing unit 4 executing an automatic learning program. The target object recognition processing unit 4 has a plurality of pieces of image data in which the conditions of the viewpoint direction, angle, illuminance, and background information of the work target object 13 are changed in advance, and the work target object 13 projected on the image data, By annotating peripheral parts including the body of the worker 11 and automatically learning the annotation result, what is captured in the in-field image 12 is recognized.

In the matching processing step (S004), the matching processing unit 5 superimposes the viewpoint coordinates on the in-field image 12. As a result, it is possible to recognize where the worker 11 is looking in the field-of-view image 12 and what he/she is looking at. For example, when the region of the work target 13 and the viewpoint coordinates overlap, the matching processing unit 5 determines that the worker 11 is looking at the work target 13, and when they do not overlap, the worker 11 looks at the work target 13. It is determined that it was not. The same applies to the peripheral parts, and when the area of the peripheral parts displayed in the in-field image 12 and the viewpoint coordinates overlap, the worker 11 determines that the peripheral parts are viewed. It should be noted that the word “looking at” here simply means that the viewpoint is located on the work target 13 and the peripheral parts, and does not ask “whether it is gazing”.

In the work recognition step (S005), the determination unit 6 determines, in the brake maintenance work procedure, the order of the brake maintenance work procedure, the high gaze rate at which the object serving as the main body of the work is observed, and the image within the visual field. Under the condition that an operator's hand (not shown) or a tool (not shown) is recognized in the area of 12, the operator 11 performs which step (here, S1 to S6) of the brake maintenance work procedure. To recognize

In the skill level determination step (S006), the determination unit 6 determines, based on various data obtained and the viewpoint analysis result (index value) of the skilled worker (experienced person) registered in the index value registration unit 7, The skill level of the worker 11 is determined.

The skill level is determined by, for example, the time when the viewpoint of the worker 11 overlaps the work target 13 and other peripheral parts, and the number of times the viewpoint of the worker 11 passes the work target 13 and other peripheral parts. It is carried out using.

FIG. 5 is a diagram illustrating a locus of viewpoints according to the embodiment, and FIG. 5A illustrates that the user is paying attention to visually judge wear or rust on the surface of the work target 13. When the worker 11 is gazing, the trajectory 24a of the viewpoint 23 traces within the drawing area of the work target 13. Alternatively, the locus 24a of the viewpoint 23 traces within the range of the drawing area of the hand of the worker 11, the gripped tool, and other peripheral components as necessary. On the other hand, FIG. 5B illustrates that the locus 24 b of the viewpoint 23 simply passes through the work target 13 and the work target 13 is not being watched.

In the present embodiment, the time when the viewpoint of the worker 11 overlaps the work target 13 and the peripheral parts and the number of times the viewpoint of the worker 11 passes over the work target 13 and the peripheral parts are respectively derived and visually determined. For each work item accompanied by, the quality of the work is judged by scoring.

In this skill level determination step (S006), the determination unit 6 obtains the index value of the skilled worker (index value of “good” determination or “not” determination) from the index value registration unit 7. For example, the determination is made by using the following equations (1A), (1B), and (1C). When any one of the determinations is “No”, the determination unit 6 sets the comprehensive determination to “No” and fails.
“Good” judgment: Δt _M +Δt _S >Δt _B1 ×80 (%) (1A)
“Good” judgment: (Δt _M +Δt _S )/Δt _A >Δt _B2 ×80(%) (1B)
“Good” judgment: N _M <(N _BM +cnst) (1C)
However, Δt _M : time during which the work target 13 is focused Δt _S : time during which peripheral parts of the work target 13 are focused Δt _B1 , Δt _B2 : work analysis result (index value) of a skilled worker
Delta] t _A: total working time of each work item N _M: number viewpoint has passed through the workpiece 13 N _BM: skilled operator work analysis (index values)
cnst: a constant Δt _B1 is a total value of the time during which the skilled worker performs the same work and the time during which the work target is focused and the peripheral parts are focused. Further, Δt _B2 is a value obtained by dividing the total value by the total work time when the skilled worker performs the same work. _NBM is the number of times the viewpoint has passed the work target when a skilled worker performs the same work. In the present embodiment, the index value is set as described above, but the present invention is not limited to this.

For example, when the viewpoint crosses the work target 13 within a specified time, it is considered that the work object 13 has passed without gazing. For example, the time difference between the cut-out image at the moment when the viewpoint enters the area of the work target 13 and the cut-out image at the moment when the viewpoint deviates from the area of the work target 13 is equal to the specified time. When it does not exceed, the determination unit 6 determines that the viewpoint simply crosses the work target 13 without gazing. The same determination method is applied to the peripheral parts of the work target 13. Then, the determination unit 6 does not add the time across which is simply crossed to the above Δt _M and Δt _S, and increments N _M by one.

On the contrary, when the time difference between the cut-out image at the moment when the viewpoint enters the area of the work target 13 and the cut-out image at the moment when the viewpoint deviates exceeds the specified time, the worker 11 determines that the work target 13 is treated as being watched, and this time is cumulatively added to Δt _M (same for peripheral parts). In the case of determining that the gazing thus, not to count up the N _M.

The determination unit 6 makes a determination for each work item (S1 to S6 described above) using the formulas (1A), (1B), and (1C). Then, when all of the calculation results of the above formulas (1A), (1B), and (1C) are “good”, the judging unit 6 regards the comprehensive judgment as “good” and passes, and even one of them If the result is "No", the comprehensive judgment is "No" and the test is rejected.

FIG. 6 is a diagram showing a data configuration example of the determination result by the determination unit 6. The determination table 25 includes a management number 26 assigned to each work item, a work process 27 describing the contents of the work item (above S1 to S6), and the formulas (1A), (1B), (1C) The operator measurement result 28 that is the calculated value on the left side is included. Further, the determination table 25 has a determination value 29 (a value on the right side of the previous expressions (1A), (1B), and (1C)) obtained by calculating a coefficient such as 80 (%) on the index value read from the index value registration unit 7. ) And a score value 30 in which a ratio (upper limit 100) with the worker measurement result 28 when the determination value 29 is used as a reference (=100) is calculated. The determination table 25 determines whether the determination result 31 according to the above formulas (1A), (1B), and (1C) and the determination result 31 are all “good”. If the determination is “no”, the overall determination 32 is “no”.

In addition, the score value 30 may be used as the determination method of the determination result 31, and a certain value or more may be determined as “good”. Further, the determination method of the comprehensive determination 32 may be determined as “good” determination when the number of “good” determinations of the determination result 31 is equal to or more than a certain number.

Further, the work process 27 shown in FIG. 6 is stored separately for each of the important points (vital points) of S1 to S6. If it is necessary to perform the maintenance work in the order of S1 to S6 described above, in the present embodiment, the management numbers 26 are also numbered in order from the smallest number according to this order. Then, the determination unit 6 also determines whether the work is performed in the order of the management number 26 (the order of S1 to S6) using the work recognition result obtained in step S005.

FIG. 7 is a flowchart showing an example of the determination method by the determination unit 6 according to the embodiment. In each value calculation step (S101), the determination unit 6 calculates each value (=Δt _M , Δt _S , Δt _A , N _M ) required for determination from the output result of the matching processing step (S004). Then, in the score calculation step (S102), the determination unit 6 calculates the score value 30 from the ratio (upper limit 100) to the worker measurement result 28 when the determination value 29 is used as the reference (=100).

In the determination step (S103) based on the determination formula, the determination unit 6 performs the quality determination using the above formulas (1A), (1B), and (1C), and in the comprehensive determination process (S104, S105, S106), the determination is performed. The unit 6 makes a “good” determination when all the determination results 31 are “good” determinations, and makes a “failure” determination when any of the determination results 31 is a “failure” determination.

After the above steps are completed, in the determination result display step (S007) of FIG. 4, the display unit 8 displays each determination result. Thereby, the skill level and the determination result for each work item can be shown to the worker 11.

FIG. 8 is a diagram illustrating an example of the determination result display screen according to the embodiment. The judgment result display screen 37 includes a judgment table display area 38 for displaying the work process 27, the score value 30, and the judgment result 31 which are the same information as the information stored in the judgment table 25. In this way, by displaying not only the quality judgment result 31 but also the score value 30, the worker 11 can grasp how much (or not) the performance of his/her own work has reached by a numerical value. it can. Further, the determination result display screen 37 includes a comprehensive determination result display area 38a for displaying the comprehensive determination result 32 and a work reflection display area 39 for displaying the in-field-of-view image 12 of the worker 11 corresponding to the work process 27, for example. And are displayed.

The work review display area 39 may also display an image within the field of view of a skilled worker (not shown) for learning by the worker 11. Further, the work review display area 39, as time-series data for the total work time (Δt _A ) for each work item, the watch time (Δt _M ) during which the work target 13 was watched and the time during which peripheral parts were watched. (Δt _S ), the number of times the viewpoint has passed the work target 13 (N _M ) may be displayed as a time series graph, for example, as in a Gantt chart. By displaying in this way, the worker 11 himself/herself can be made to recognize the work with low skill level, and it is possible to carry out the education efficiently.

In the accumulation process (S008) of the determination result of FIG. 4, together with the quality determination result in the skill determination process (S006), the attention time (Δt _M ) during which the work target 13 was observed and peripheral components are observed. The time (Δt _S ), the number of times the viewpoint has passed the work target 13 (N _M ), and the total work time (Δt _A ) for each work item are stored in the storage unit 9. The accumulated data may be used for quality determination of the work to be performed later. That is, the determination unit 6 reads out one or more “good” determination data from the accumulated past data, calculates an index value and a determination value (when multiple readings are performed, calculates an average value thereof), and presently It may have a function of comparing the result of the work being performed with the calculated index value or judgment value.

In the present embodiment, an object is identified using an artificial intelligence technique such as executing an automatic learning program. For example, a machine-readable mark such as a two-dimensional barcode is attached to a work target in advance. Alternatively, it may be mounted by reading this and specifying the work target.

Further, in the present embodiment, two determinations, “good” determination and “fail” determination, are used as the determination result, but by providing a plurality of determination values, the result of the intermediate layer between the “good” determination and the “fail” determination May be implemented.

Further, in the object recognition step (S003), the work manual is displayed in the cut-out image, and in the matching processing step (S004) and the skill level determination step (S006), it is determined that the work manual is being watched. In this case, the time taken to gaze at this work manual may be accumulated, and if this accumulated time exceeds the threshold value, a “no” determination may be made.

Note that the degree to which the “skilled worker” in the above embodiment is skilled is not particularly limited. That is, the “skilled worker” includes an experienced person who has performed the same work as the work performed by the worker 11 at least once.

In the above-described embodiment, the mode applied to the maintenance work and the education of the maintenance work for the elevator equipment such as the elevator has been described as an example, but the target equipment is not limited to this. In addition, the usage of the present embodiment is not limited to education and technology transfer, but is applied to all maintenance work.

Further, in the above-described embodiment, a mode has been described in which the "work target" is a component of the facility (more specifically, the core of the elevator), and it is determined whether or not the drawing area of the "work target" is watched. However, it is not limited to this. For example, a peculiar portion of the component such as a stain, rust, scratch, or worn portion attached to the component may be identified and treated as the “work target”. By handling in this way, it is possible to determine whether not only the entire part but also its peculiar part (dirt, rust, scratch, or worn part) is focused.

Further, as shown in FIG. 6, in the present embodiment, the determination value is calculated using the same coefficient (80% or the like) for any work item, but a different coefficient may be provided for each work item. Absent.

According to the present embodiment, it is possible to eliminate the influence of fluctuation of the viewpoint due to hesitation and improve the accuracy of determining the skill level.

It should be noted that the present invention is not limited to the above-described embodiment, and various modifications are included. For example, the above-described embodiments have been described in detail in order to explain the present invention in an easy-to-understand manner, and are not necessarily limited to those including all the configurations described. Further, a part of the configuration of one embodiment can be replaced with the configuration of another embodiment, and the configuration of another embodiment can be added to the configuration of one embodiment. Further, it is possible to add/delete/replace other configurations with respect to a part of the configurations of the respective embodiments. Further, the above-mentioned respective configurations, functions, processing units, processing means, etc. may be realized by hardware by designing a part or all of them with, for example, an integrated circuit. Further, each of the above-described configurations, functions, and the like may be realized by software by a processor interpreting and executing a program that realizes each function. Information such as a program, a table, and a file that realizes each function can be stored in a memory, a recording device such as a hard disk and an SSD (Solid State Drive), or a recording medium such as an IC card, an SD card, and a DVD.

1: Skill level determination system 2: Viewpoint detection unit 3: External imaging unit 4: Object recognition processing unit 5: Matching processing unit 6: Determination unit 7: Index value registration unit 8: Display unit 9: Storage unit 10: Viewpoint detection Device 20: Field terminal

Claims (5)

A skill level determination system for determining the skill level of an operator,
An external image pickup unit for picking up a visual field image of an operator,
A viewpoint detection unit that detects the position of the viewpoint of the worker,
From the video imaged by the external imaging unit, the frame images forming the video image are respectively cut out, and the work target object that is projected is recognized for each cut out image, and the drawing area of the work target object is set to the cutout image. An object recognition processing unit that identifies each
In the drawing area specified by the object recognition processing unit, whether or not the viewpoint detected by the viewpoint detection unit is located, a matching processing unit that determines for each cutout image,
Based on the determination result of the matching processing unit, the gaze time during which the worker has gazed within the drawing area for more than a specified time, and the number of times the viewpoint has passed through the drawing area without gazing for the specified time. Deriving the, based on the gaze time and the number of times passed, a determination unit that determines whether or not the work by the worker is properly performed,
A display unit that displays a result determined by the determination unit,
Skill determination system having. In the skill level determination system according to claim 1,
A numerical value relating to the gaze time and the number of times when an experienced person performs the same work as the work performed by the worker, and has an index value registration unit that stores the index value,
The proficiency level determination system, wherein the determination unit determines whether or not the work by the worker is properly performed based on the index value stored in the index value registration unit. In the skill level determination system according to claim 1,
Further, the determination unit by the determination unit, a gaze time, a storage unit for accumulating data associated with the number of times passed,
The determination unit, in the past data accumulated in the accumulation unit, acquires the gaze time and the number of times of passing corresponding to the determination result that the work was properly performed, using these, the worker A proficiency level determination system characterized by determining whether or not the work by is properly performed. In the skill level determination system according to claim 1,
The determination unit further calculates a score value for the work by using the gaze time and the number of times of passing,
The skill level determination system, wherein the display unit displays the score value. In the skill level determination system according to claim 1,
The determination unit, for each work item performed by the worker, derives the gaze time and the number of times passed, for each work item, determines whether or not the work is properly performed,
The skill level determination system, wherein the display unit displays a determination result by the determination unit for each of the work items.

Images