Scaffolding quality assessment system and scaffolding quality assessment program

Abstract

To reliably reassure safety of a scaffold in an easy configuration.SOLUTION: A scaffold quality determination system comprises a portable terminal 2 shooting a scaffold 101 to be decided and generating an image to be decided and a scaffold monitoring server 3 storing a sample image which is the image showing a correct condition of the scaffold 101, comparing the sample image to the image to be decided and deciding whether the scaffold 101 to be decided is correctly installed or not.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a scaffolding quality determination system and a scaffolding quality determination program for determining and checking whether a scaffolding is set correctly. [Background technology]

[0002] For example, at power plants, work and construction are frequently carried out for the renovation, repair, and expansion of equipment, and scaffolding (temporary scaffolding) is often used for this work. When scaffolding is set up and used, it must be set up correctly in accordance with occupational safety and health regulations to ensure safety. For example, it is necessary to set up sashes at specified intervals, handrails at specified heights, and baseboards at specified heights. However, there are cases where the scaffolding is not set up correctly due to mistakes made by the scaffolding installer. For this reason, regular safety patrols are conducted, and if any deficiencies are discovered, improvements are made each time, but there is a risk that safety will not be ensured until improvements are made.

[0003] For this reason, a scaffold safety management device is known that can reliably manage the usable / unusable status of scaffolding (see, for example, Patent Document 1). This device displays a scaffold inspection sheet on the terminal of the manager, and when the manager inputs the inspection results for each inspection item on the inspection sheet, it determines whether the scaffold is usable or unusable, and sends the determination result to an alarm device attached to the scaffold entrance. When the alarm device receives the determination result that the scaffold is unusable, it locks the door attached to the scaffold entrance. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Publication No. 2020-133171 Summary of the Invention [Problem to be solved by the invention]

[0005] However, with the scaffolding safety management device described in Patent Document 1, a manager must inspect all inspection items and input the results, which not only takes time and effort, but also carries the risk of inputting incorrect inspection results because the inspections are done manually. In other words, there is a risk that an error will occur in determining whether the scaffolding is usable or not, and safety will not be ensured. In addition, alarm devices and opening / closing doors must be installed at all entrances to the scaffolding, which requires a great deal of cost and effort.

[0006] Therefore, an object of the present invention is to provide a scaffolding quality determination system and a scaffolding quality determination program that can reliably ensure the safety of scaffolding with a simple configuration. [Means for solving the problem]

[0007] In order to solve the above problem, the invention of claim 1 comprises a sample image storage means for storing a sample image which is an image showing the correct state of scaffolding, a photographing means for photographing a scaffold to be judged and generating a judgment target image, and a judging means for comparing the sample image with the judgment target image to judge whether the scaffold to be judged is correctly installed or not, wherein the photographing means is a mobile terminal, and the mobile terminal has ,before This scaffolding quality assessment system is characterized by having installed application software that instructs the position and direction in which the image should be taken so that it is taken from the same position and direction as the sample image was taken.

[0008] The invention of claim 2 is characterized in that, in the scaffolding quality judgment system described in claim 1, the sample image storage means stores the sample image for each type of scaffolding, and the judgment means compares the sample image of scaffolding of the same type as the scaffolding to be judged with the image to be judged.

[0009] The invention of claim 3 is characterized in that, in the scaffolding quality judgment system described in claim 1 or 2, the sample image includes images of each process for correctly installing scaffolding, the image to be judged includes an image of the scaffolding to be judged in a state of being installed, and the judgment means compares the sample image with the image to be judged to judge whether the scaffolding to be judged has been installed in the correct process.

[0010] The invention of claim 4 is characterized in that, in the scaffolding quality determination system described in claims 1 to 3, the determination means uses a learning model for quality determination that has been machine-learned based on past performance data so that when the sample image and the image to be determined are input, it outputs whether the scaffolding to be determined is installed correctly or not.

[0011] The invention of claim 5 is characterized in that, in the scaffolding quality determination system described in claims 1 to 4, the determination means outputs an alarm including what is incorrect if it determines that the scaffolding is not installed correctly. [Effects of the Invention]

[0016] Claim 1 to According to the described invention, a sample image of a correct scaffolding is compared with a target image of the scaffolding to be evaluated, automatically determining whether the scaffolding to be evaluated is installed correctly. This makes it possible to accurately determine the quality of the scaffolding to be evaluated and reliably ensure the safety of the scaffolding. Furthermore, since the system only requires a sample image storage means for storing the sample image and nothing else needs to be installed on the scaffolding itself, the system is simple in configuration and can be easily applied to all scaffolding. Furthermore, since the supervisor or other person in charge only needs to take a photograph of the scaffolding to be evaluated, the burden on the supervisor or other person in charge can be reduced, and anyone, even without knowledge of occupational safety and health regulations, can easily and quickly determine the quality of the scaffolding.

[0017] Claim 2 toAccording to the described invention, a sample image of scaffolding of the same type as the scaffolding to be evaluated is compared with a target image of the scaffolding to be evaluated to determine whether the scaffolding to be evaluated is installed correctly. For example, if the type of scaffolding to be evaluated is a frame scaffolding, a sample image of the frame scaffolding is compared with a target image of the scaffolding to be evaluated to determine whether the scaffolding to be evaluated is installed correctly. This makes it possible to more appropriately determine the quality of the scaffolding to be evaluated depending on the type of scaffolding to be evaluated.

[0018] Claim 3 According to the described invention, sample images of each step for correctly setting up scaffolding are compared with a target image of the scaffold being set up to be judged, and it is judged whether the scaffolding to be judged has been set up in the correct steps. In other words, it is judged whether the scaffolding set up process and procedure are correct, and therefore it is possible to ensure safety during the scaffolding set up.

[0019] Claim 4 According to the described invention, a machine-learned learning model for determining whether the scaffolding being evaluated is installed correctly is used to determine whether it is installed correctly, making it possible to accurately and appropriately determine whether it is installed correctly.

[0020] According to the invention of claim 5, if it is determined that the scaffolding being evaluated is not installed correctly, an alarm is output, which prevents the use of incorrect scaffolding or installing scaffolding in an incorrect process or procedure, making it possible to more reliably ensure safety. Moreover, because the alarm includes information about what is incorrect, it becomes possible to properly improve the incorrect parts and reliably ensure safety. [Brief explanation of the drawings]

[0021] [Figure 1] 1 is a schematic diagram showing a scaffolding quality determination system according to an embodiment of the present invention; [Figure 2] 2 is a schematic block diagram showing the configuration of a scaffolding supervisor server of the scaffolding quality determination system of FIG. 1. FIG. [Figure 3] FIG. 3 is a data structure diagram of a sample image database of the scaffolding supervision server of FIG. 2. [Figure 4] FIG. 4 shows an example of a sample image of a framework scaffold stored in the sample image database of FIG. 3. [Figure 5] FIG. 4 is a diagram showing an example of a sample image of a single-pipe scaffold stored in the sample image database of FIG. 3. [Figure 6] FIG. 4 is a diagram showing an example of a sample image of one step in a regular scaffolding assembly procedure stored in the sample image database of FIG. 3. [Figure 7] 2 is a diagram showing an example of an image of a completed scaffolding to be judged in the scaffolding quality judgment system of FIG. 1. FIG. [Figure 8] 2 is a diagram showing an example of an image of a scaffold to be judged in the scaffold quality judgment system of FIG. 1 in the middle of being installed; FIG. [Figure 9] FIG. 3 is a functional block diagram showing a schematic configuration of a learning model for determining pass / fail of the scaffolding supervision server of FIG. 2. DETAILED DESCRIPTION OF THE INVENTION

[0022] The present invention will be described below based on the illustrated embodiments.

[0023] 1 to 9 show an embodiment of the present invention, and Fig. 1 is a schematic configuration diagram showing a scaffolding quality determination system 1 according to this embodiment. This scaffolding quality determination system 1 is a system that determines and checks whether a scaffolding (temporary scaffolding) 101 is set up correctly, and in this embodiment, it is configured by freely communicating with a mobile terminal (photography means) 2 carried and operated by a person such as M who is responsible for managing the assembly work of the scaffolding 101, and a scaffolding supervision server 3. Alternatively, the configuration and functions of the scaffolding supervision server 3, which will be described later, may be incorporated into the mobile terminal 2, and the scaffolding quality determination system 1 may be configured by the mobile terminal 2 alone, or the scaffolding supervision server 3 may be configured by a plurality of computers, servers, etc.

[0024] The mobile terminal 2 is a terminal that takes an image of the scaffolding 101 to be judged and generates a judgment target image TG. As long as it has a photographing function, a communication function, a display, etc., it may be any type of device, such as a multi-function mobile terminal, a personal computer, or a dedicated terminal. In this embodiment, however, it is configured as a smartphone. In addition, application software (hereinafter referred to as the "good / bad" judgment app) for receiving the scaffolding good / bad" judgment service provided by the scaffolding good / bad" judgment system 1 is installed in the mobile terminal 2. This good / bad" judgment app can transmit the captured judgment target image TG to the scaffolding supervision server 3 and receive the judgment results from the scaffolding supervision server 3 and display them on the display. In other words, by simply installing the good / bad" judgment app on a general-purpose smartphone, anyone can receive the scaffolding good / bad" judgment service and learn and confirm whether the scaffolding 101 is set up correctly.

[0025] Here, when photographing the scaffolding 101 to be evaluated using the mobile terminal 2, it is preferable to photograph the scaffolding 101 so that the evaluation target image TG can be compared with the sample image SG described later and the pass / fail judgment can be more appropriately performed. That is, it is preferable to photograph the scaffolding 101 from the same position, direction, angle, etc. (overall, directly in front, directly to the side, diagonally right from the front, diagonally left from the front, etc.) as the scaffolding 101 from which the sample image SG was photographed. For this reason, when photographing the scaffolding 101 to be evaluated using the mobile terminal 2, the pass / fail judgment app provides guidance and instructions on the position, direction, angle, etc. to be photographed (for example, by displaying a template on the display or by providing audio guidance). Furthermore, by providing guidance and instructions on the photographing conditions (stored in the sample image database 32 described later) including autofocus information (distance information) and lens angle of view information when the sample image SG was photographed, it is possible to photograph the scaffolding 101 under the same photographing conditions and more appropriately perform the pass / fail judgment.

[0026] Furthermore, the judgment target image TG includes not only an image of the scaffolding 101 in a completed state where the scaffolding 101 to be judged has been completely installed, but also an image of the scaffolding 101 in a state where it is being installed (in the middle of completion). That is, in order to judge the quality of the completed scaffolding 101, the mobile terminal 2 may take an image of the scaffolding 101 in a completed state and send it to the scaffolding supervision server 3, or in order to judge the quality of the installation process (procedure) of the scaffolding 101, the scaffolding 101 in the middle of completion may be taken and sent to the scaffolding supervision server 3. Here, when the quality judgment app sends the judgment target image TG to the scaffolding supervision server 3, status information indicating the completed state is input for the judgment target image TG in a completed state, and status information indicating the installation state or assembly process is input for the judgment target image TG in the middle of completion, and these status information are sent to the scaffolding supervision server 3.

[0027] The scaffolding supervision server 3 is a server managed and operated by assembly construction companies and companies that provide scaffolding quality assessment services, and as shown in Figure 2, it mainly comprises a communication unit 31, a sample image database (sample image storage means) 32, an assessment task (assessment means) 33, a learning task 34, a learning model for quality assessment 35, a performance database for quality assessment 36, and a central control unit 37 that controls these.

[0028] The communication unit 31 is an interface for communicating with the mobile terminal 2 and the like.

[0029] The sample image database 32 is a database that stores sample images SG, which are images showing the correct state of the scaffolding 101, and stores sample images SG for each type of scaffolding. That is, as shown in Fig. 3, a plurality of sample images SG are stored for each scaffolding type (kind of scaffolding) 1 to n, and these sample images SG include a sample image SG of the scaffolding 101 in a completed state after assembly, and sample images SG of each process for correctly setting up the scaffolding 101. Also, corresponding to each sample image SG, shooting conditions including autofocus information (distance information) and lens angle of view information at the time of shooting are stored.

[0030] Here, types of scaffolding include frame scaffolding, single-pipe scaffolding, suspended scaffolding, rolling tower, etc. Sample images SG are stored to enable comparison with the evaluation target image TG (described later) and more appropriately judge pass / fail. That is, in this embodiment, images of a properly assembled and completed scaffolding 101 taken from multiple positions, directions, angles, etc. (overall, directly front, directly side, diagonally right from the front, diagonally left from the front, etc.) are stored as sample images SG, allowing appropriate comparison of any of the sample images SG with the evaluation target image TG. Similarly, images of the states of each process for properly installing the scaffolding 101 (states in the middle of completion) taken from multiple positions, directions, angles, etc. are also stored as sample images SG. Furthermore, for the sample images SG in the completed state, status information indicating the completed state is stored, and for the sample images SG in the middle of completion, status information indicating the installation state or assembly process (e.g., the second floor is in the middle of being assembled) is stored.

[0031] For example, sample images SG of a completed frame scaffolding as shown in FIG. 4, a completed single-pipe scaffolding as shown in FIG. 5, or a sample image SG of one process (one stage) for properly installing frame scaffolding as shown in FIG. 6 are stored in the sample image database 32. Here, in the sample image SG of the completed state, necessary components such as jacks, frames, braces, scaffolding boards, slats, handrails, and baseboards are installed without excess or deficiency in accordance with occupational safety and health regulations, and these components are installed in the correct positions and with the correct dimensions. Also, in the sample image SG of one stage in the middle of completion, necessary components for the process are installed without excess or deficiency in accordance with occupational safety and health regulations, and these components are installed in the correct positions and with the correct dimensions.

[0032] The judgment task 33 is a task program that compares the judgment target image TG received from the mobile terminal 2 with the sample image SG to judge whether the scaffolding 101 to be judged has been installed correctly. At this time, the judgment is made by comparing the judgment target image TG with the sample image SG of scaffolding of the same type as the scaffolding 101 to be judged. Here, whether it has been installed correctly includes both whether it has been installed correctly when assembly is complete and whether it has been installed in the correct process and procedure during completion.

[0033] Specifically, first, the type of scaffolding 101 in the judgment target image TG received from the mobile terminal 2 is determined. In this embodiment, the judgment target image TG is subjected to image analysis, and the type of scaffolding 101 is automatically determined based on characteristic components corresponding to the type of scaffolding 101. For example, in the case of a frame scaffolding, a frame 111 as shown in FIG. 4 is installed, so if the frame 111 is extracted in the judgment target image TG, it is determined to be a frame scaffolding. Similarly, in the case of a single-pipe scaffolding, a clamp 121 as shown in FIG. 5 is installed, so if the clamp 121 is extracted in the judgment target image TG, it is determined to be a single-pipe scaffolding. Alternatively, the type of scaffolding may be selected and input using a pass / fail judgment app on the mobile terminal 2, so that the type of scaffolding is transmitted to the scaffolding supervision server 3 along with the judgment target image TG.

[0034] Next, a sample image SG of scaffolding of the same type as the type of scaffolding 101 of the judged judgment target image TG is acquired and selected from the sample image database 32. At this time, if the scaffolding 101 of the judgment target image TG is in a completed state, the sample image SG of the completed scaffolding 101 is selected, and if the scaffolding 101 of the judgment target image TG is in a process of being completed, the sample image SG of the scaffolding 101 in a process of being completed is selected. Specifically, the sample image SG of the same state information as the above-mentioned state information (information indicating whether the scaffolding 101 is in a completed state or in what state it is installed) received together with the judgment target image TG is selected. Alternatively, whether the scaffolding 101 of the judgment target image TG is in a completed state or in a process of being completed may be automatically determined based on whether or not the scaffolding 101 of the judgment target image TG has a characteristic configuration that is included in the scaffolding 101 in a completed state (for example, in the case of a frame scaffolding, there is a handrail on the top layer but no frame).

[0035] Next, the selected sample image SG is compared with the target image TG to determine the scaffolding of the target. It is determined whether the scaffolding 101 in the judgment target image TG has been installed correctly or in the correct process. That is, when the scaffolding 101 in the judgment target image TG is in a completed state, it is determined by performing image analysis of the sample image SG and the judgment target image TG whether the necessary components such as jacks, frames, braces, scaffolding boards / cloth boards, rails, handrails, and baseboards are installed in the correct positions and with the correct dimensions in the same way as the scaffolding 101 in the sample image SG.

[0036] For example, when comparing the sample image SG shown in FIG. 5 with the judgment target image TG shown in FIG. 7, the following defects are extracted and determined. First, the scaffolding 101 in the sample image SG has a brace 122 attached, but the scaffolding 101 in the judgment target image TG does not have a brace 122 attached. Second, the height of the baseboard 123 of the scaffolding 101 in the judgment target image TG is lower than the height of the baseboard 123 of the scaffolding 101 in the sample image SG. Third, the height position of the handrail 124 of the scaffolding 101 in the judgment target image TG is lower than the height position of the handrail 124 of the scaffolding 101 in the sample image SG. Here, when determining whether the dimensions and positions of components are acceptable or not, as in the second and third cases, if the dimensions and positions of the components are within a predetermined tolerance range compared to the dimensions and positions of the components in the sample image SG, they are determined to be "acceptable." If they are outside the predetermined tolerance range, they are determined to be "defective." Furthermore, if extra tools or components are extracted in the judgment target image TG compared to the sample image SG, they are deemed to have been left behind on the scaffolding 101 and are determined to be "defective."

[0037] On the other hand, if the scaffolding 101 of the judgment target image TG is in the middle of completion, a sample image SG that is similar to the installation state of the scaffolding 101 of the judgment target image TG is selected. That is, as described above, a sample image SG with the same state information as the state information of the judgment target image TG is selected. Then, similar to the scaffolding 101 of the sample image SG, image analysis of the sample image SG and the judgment target image TG is performed to determine whether the necessary components are installed in the scaffolding 101 of the judgment target image TG without excess or deficiency, and whether these components are installed in the correct positions and with the correct dimensions.

[0038] For example, when the sample image SG shown in Fig. 6, which shows the second layer in the middle of assembly, is compared with the judgment target image TG shown in Fig. 8, the following defects are extracted and judged: In the scaffolding 101 of the sample image SG, all of the diagonal braces 112 of the first layer are installed before the second layer is assembled, but in the scaffolding 101 of the judgment target image TG, some of the diagonal braces 112 are not installed.

[0039] The result of this determination is then transmitted to the mobile device 2. That is, if the determination is "good", a determination result of "good installation" is transmitted to the mobile device 2. In response to this, "good installation" is displayed on the display of the mobile device 2. On the other hand, if the determination is "bad", that is, if the installation is not correct, a determination result of "bad installation" is transmitted to the mobile device 2. This determination result includes warning information indicating what is wrong (the defective part), and indicates, for example, which diagonal braces 122 are not attached and which baseboards 123 are low on the determination target image TG. In response to this, the warning information is displayed on the display of the mobile device 2, and an alarm sound is emitted.

[0040] The judgment task 33 uses a pass / fail judgment learning model 35 that is machine-learned based on past performance data so that when a sample image SG and a judgment target image TG are input, it outputs whether the scaffolding 101 to be judged is installed correctly or not. This pass / fail judgment learning model 35 is created by a learning task 34.

[0041] That is, the learning task 34 uses past performance data recorded and accumulated in a performance database 36 for quality determination to create a quality determination learning model 35 using a known machine learning algorithm such as a neural network. This performance database 36 for quality determination stores sample images SG and judgment target images TG that have been used and compared in the past, and the sample images SG and judgment target images TG. This is a database in which performance data including the pass / fail judgment results (actual pass / fail judgment as to whether the scaffolding 101 to be judged is installed correctly or not) actually judged by a manager or the like as to whether or not the scaffolding 101 of the judgment target image TG is installed correctly based on SG. Note that the past performance data includes data created based on the actual judgment results of the actual images SG, TG and the scaffolding 101 to be judged, as well as data created for pre-training, etc.

[0042] 9, this learning task 34 uses machine learning and deep learning using a neural network to create a neural network based on performance data recorded in a performance database 36 for quality assurance, with the sample image SG and the image to be judged TG as an input layer, the quality assurance results as an output layer, and image analysis processing from the images SG and TG to the quality assurance results as an intermediate layer. The learning task 34 then uses the performance data of the quality assurance learning model 35 as learning data to learn various parameters in the intermediate layer. That is, the learning task 34 learns various parameters in the intermediate layer so that quality assurance results (whether the scaffolding 101 to be judged for the image to be judged TG is installed correctly and what is incorrect) can be obtained without omission from the sample image SG and the image to be judged TG with high accuracy.

[0043] As described above, with this scaffolding quality assessment system 1, the manager M or an equivalent simply photographs the scaffolding 101 to be assessed using the mobile device 2 and transmits the assessment target image TG to the scaffolding supervision server 3, and the assessment result is transmitted to the mobile device 2. That is, as described above, the sample image SG of the correct scaffolding 101 is compared with the assessment target image TG of the scaffolding 101 to be assessed, and it is automatically determined whether the scaffolding 101 to be assessed is correctly installed. This makes it possible to properly assess the quality of the scaffolding 101 to be assessed and reliably ensure the safety of the scaffolding 101. Furthermore, since the scaffolding supervision server 3 only needs to be equipped with a sample image database 32 that stores the sample image SG, and nothing else needs to be equipped on the scaffolding 101 itself, the configuration is simple and can be easily applied to all scaffoldings 101. Moreover, since the manager M or an equivalent only needs to photograph the scaffolding 101 to be assessed, the burden on the manager M or an equivalent can be reduced, and anyone, even those without knowledge of occupational safety and health regulations, can easily and quickly assess the quality of the scaffolding 101.

[0044] Moreover, a sample image SG of scaffolding 101 of the same type as the scaffolding 101 to be judged is compared with a judgment target image TG of the scaffolding 101 to be judged to judge whether the scaffolding 101 to be judged has been installed correctly. For example, if the type of the scaffolding 101 to be judged is a frame scaffolding, a judgment is made by comparing the sample image SG of the frame scaffolding with the judgment target image TG of the scaffolding 101 to be judged. Therefore, it becomes possible to more appropriately judge the quality of the scaffolding 101 to be judged depending on the type of the scaffolding 101 to be judged.

[0045] Furthermore, the sample images SG of each step for correctly setting up the scaffolding 101 are compared with the judgment target image TG of the scaffolding 101 being set up to be judged, and it is judged whether the scaffolding 101 to be judged has been set up in the correct steps. In other words, it is judged whether the setting up process and procedure of the scaffolding 101 are correct, and therefore it is possible to ensure safety during the setting up of the scaffolding 101.

[0046] In addition, since the machine-learned learning model 35 for determining whether the scaffolding 101 to be determined is installed correctly or not is used to determine whether it is installed correctly, it is possible to accurately and appropriately determine whether it is installed correctly or not.

[0047] If it is determined that the scaffolding 101 to be judged is not installed correctly, an alarm is output to the mobile terminal 2, so that it is possible to prevent the use of an incorrect scaffolding 101 or installing the scaffolding 101 in an incorrect process or procedure. This prevents the installation of 01, making it possible to ensure safety more reliably. Moreover, because the warning includes information about what is incorrect, it is possible to properly improve the incorrect parts and ensure safety.

[0048] Although the embodiments of the present invention have been described in detail above, the specific configuration is not limited to these embodiments, and the present invention also includes design changes and the like that do not deviate from the gist of the present invention. For example, in the above embodiment, a case has been described in which a manager M or the like transmits a judgment target image TG to the scaffold supervision server 3 using a mobile terminal 2, and transmits and returns a judgment result to the mobile terminal 2, but it is also possible for one or more people to transmit a judgment target image TG to the scaffold supervision server 3 using their own terminals, and for the judgment result to be transmitted to a terminal of a specific person (for example, manager M).

[0049] On the other hand, by installing the following scaffolding quality judgment program in a general-purpose computer, it is possible to provide the computer with judgment functions similar to those of the above-described scaffolding quality judgment system 1. That is, the scaffolding quality judgment program causes the computer to function as: sample image storage means (sample image database 32) that stores sample images SG, which are images showing the correct state of scaffolding; and judgment means (judgment task 33) that compares a judgment target image TG generated by photographing the scaffolding to be judged with the sample image SG to judge whether the scaffolding to be judged has been installed correctly or not, where the sample image storage means stores sample images SG for each type of scaffolding, the judgment means compares the judgment target image TG with the sample image SG of scaffolding of the same type as the scaffolding to be judged, and further, the sample image SG includes images of each step for correctly installing the scaffolding, and the judgment target image TG includes an image of the scaffolding to be judged in its installed state, and the judgment means compares the sample image SG with the judgment target image TG to judge whether the scaffolding to be judged has been installed in the correct step or not. In addition, the judgment means uses a learning model 35 for determining whether or not the scaffolding to be judged is installed correctly, which has been machine-learned based on past performance data, so that when a sample image SG and an image to be judged TG are input, a result is output as to whether or not the scaffolding to be judged is installed correctly. [Explanation of symbols]

[0050] 1. Scaffolding quality assessment system 2. Mobile device (photography means) 3. Scaffolding Supervisor Server 32 Sample image database (sample image storage means) 33 Judgment task (judgment means) 34 Learning Tasks 35 Learning model for pass / fail judgment 36 Performance database for quality judgement 101 Scaffolding SG sample image TG Judgment target image

Claims

1. a sample image storage means for storing a sample image that shows a correct scaffolding state; an imaging means for capturing an image of the scaffolding to be judged and generating an image to be judged; a determination means for comparing the sample image with the determination target image to determine whether the scaffolding to be determined is correctly installed; Equipped with the photographing means is a mobile terminal, application software is installed on the mobile terminal for instructing the position and direction to photograph so that the image is photographed from the same position and direction as the sample image was photographed; A scaffolding quality determination system characterized by:

2. The sample image storage means stores the sample image for each type of scaffolding, The determination means compares the sample image of a scaffold of the same type as the scaffold to be determined with the image to be determined.

2. The scaffolding quality determination system according to claim 1.

3. The sample images include images of each step for correctly installing the scaffolding, The image to be determined includes an image in which the scaffolding to be determined is installed, The determination means compares the sample image with the image to be determined and determines whether the scaffolding to be determined has been installed in a correct process.

3. The scaffolding quality determination system according to claim 1 or 2.

4. the judgment means uses a learning model for judgment of whether or not the scaffolding to be judged is installed correctly, the learning model being machine-learned based on past performance data, so that when the sample image and the judgment target image are input, a judgment of whether or not the scaffolding to be judged is installed correctly is output.

4. The scaffolding quality determination system according to claim 1, wherein the scaffolding quality determination system is a scaffolding quality determination system.

5. When the determination means determines that the device is not installed correctly, it outputs a warning including information about what is incorrect.

5. The scaffolding quality determination system according to claim 1, wherein the scaffolding quality determination system is a scaffolding quality determination system.

Images