Inspection method for pressure-welding joint structure, information processing apparatus, and program

The inspection method addresses inaccuracies in existing systems by using measurement values, imaging, and machine learning to accurately assess pressure-welded joint structures, adapting to varying rebar types and standards.

JP2025187624AActive Publication Date: 2025-12-25NTT COMWARE CORP
View PDF 3 Cites 0 Cited by

Patent Information

Application Number
JP2024096592
Authority / Receiving Office
JP · JP
Patent Type
Applications
Current Assignee / Owner
Filing Date
2024-06-14
Publication Date
2025-12-25
Estimated Expiration
2044-06-14

AI Technical Summary

Technical Problem

Existing building inspection systems fail to accurately inspect pressure-welded joint structures due to the use of fixed rebar diameters and image distortion, leading to insufficient accuracy and inability to accommodate varying rebar types and client-specific inspection standards.

Method used

An inspection method that includes inputting measurement values, imaging the structure, recognizing components, calculating actual dimensions per pixel, and selecting between first and second inspection standards, using machine learning to enhance accuracy and adapt to different standards.

Benefits of technology

The method enables precise inspection of pressure-welded joint structures, reducing effort and improving accuracy by accounting for rebar variations and image distortions.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 2025187624000001_ABST
    Figure 2025187624000001_ABST
Patent Text Reader

Abstract

To more accurately inspect a pressure-welding joint structure.SOLUTION: An inspection method for a pressure-welding joint structure includes: an input step of inputting a measured value obtained by measuring, using a measuring tool, a width of a reinforcing bar in the pressure-welding joint structure in which reinforcing bars are connected to each other by pressure-welding; an imaging step of imaging the pressure-welding joint structure; a recognition step of determining whether or not each of a plurality of components of the pressure-welding joint structure is included on the basis of a pressure-welding joint structure image captured in the imaging step; a calculation step of calculating an actual size per pixel in the pressure-welding joint structure image on the basis of the pressure-welding joint structure image and the measured value input in the input step; and an inspection step of inspecting the pressure-welding joint structure on the basis of an image indicating the plurality of components and the actual size per pixel in the pressure-welding joint structure image when it is recognized that the pressure-welding joint structure includes the plurality of components.SELECTED DRAWING: Figure 5
Need to check novelty before this filing date? Find Prior Art

Description

[Technical Field]

[0001] The present invention relates to a method for inspecting a pressure-welded joint structure, an information processing device, and a program. [Background technology]

[0002] A known building inspection system is described in Patent Document 1. This building inspection system determines whether the entire building is included in a captured image of the building based on a first certainty factor indicating the certainty of the entire building included in the captured image, determines whether each of the building's constituent parts is included based on second certainty factors indicating the certainty of each of the building's constituent parts included in the captured image, and, if it is determined that the entire building and the multiple constituent parts are included, performs inspection based on the image showing the constituent parts. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Patent Publication No. 2021-125052 Summary of the Invention [Problem to be solved by the invention]

[0004] However, when measuring the size in each inspection of pressure welded joint structures, a fixed rebar diameter is used for each rebar name, and while inspections that correspond to rebar diameters that vary depending on the manufacturer, type, and lot of the rebar are desired, the technology in Patent Document 1 mentioned above does not accommodate such inspections, and the inspection accuracy is insufficient. In other words, the building inspection system described in Patent Document 1 does not inspect according to the name of each type of rebar.

[0005] The building inspection system described in Patent Document 1 had fixed inspection standards and could not accommodate differences in inspection standards between clients. Furthermore, with the building inspection system described in Patent Document 1, when photographs were taken of the pressure-welded joint structure from above, below, or at an angle, the image was strongly affected by image distortion, which could result in reduced accuracy.

[0006] The present invention has been made in consideration of the above-mentioned problems, and aims to provide an inspection method for a pressure-welded joint structure, an information processing device, and a program that can inspect a pressure-welded joint structure more accurately. [Means for solving the problem]

[0007] (1) One aspect of the present invention is a method for inspecting a pressure-welded joint structure, the method including: an input step in which an information processing device inputs measurement values ​​obtained by measuring the width of a reinforcing bar in a pressure-welded joint structure in which reinforcing bars are connected to each other by pressure welding, the measurement values ​​being measured using a measuring tool; an imaging step in which an inspection terminal device images the pressure-welded joint structure; a recognition step in which the information processing device determines whether each of a plurality of components of the pressure-welded joint structure is included based on the pressure-welded joint structure image captured by the imaging step; a calculation step in which the information processing device calculates an actual dimension per pixel in the pressure-welded joint structure image based on the pressure-welded joint structure image and the measurement values ​​input by the input step; and an inspection step in which the information processing device inspects the pressure-welded joint structure based on the image showing the plurality of components and the actual dimension per pixel in the pressure-welded joint structure image if the information processing device recognizes that the pressure-welded joint structure includes the plurality of components.

[0008] (2) One aspect of the present invention includes a selection step in which the information processing device accepts an operation to select either a first inspection standard or a second inspection standard before the inspection step, and a reporting step in which the information processing device creates inspection result information including a pass / fail judgment result calculated based on the first inspection standard if the first inspection standard is selected by the selection step, and creates inspection result information including a pass / fail judgment result calculated based on the second inspection standard if the second inspection standard is selected by the selection step, and the inspection step may create a pass / fail judgment result according to the first inspection standard if the first inspection standard is selected by the selection step, and create a pass / fail judgment result according to the second inspection standard if the second inspection standard is selected by the selection step.

[0009] (3) In one aspect of the present invention, the inspection step may include setting an upper mask area covering an upper end region of the pressure-welded portion of the pressure-welded joint structure, and a lower mask area covering a lower end region of the pressure-welded portion of the pressure-welded joint structure, setting an upper end mask area covering an upper end region of the upper reinforcing bar of the pressure-welded joint structure, and a lower end mask area covering a lower end region of the lower reinforcing bar of the pressure-welded joint structure, and making the vertical lengths of the upper mask area and the lower mask area shorter than the vertical lengths of the upper end mask area and the lower end mask area.

[0010] (4) In one aspect of the present invention, the recognition step includes the steps of inputting the pressure-welded joint structure image into a machine learning model, calculating a first certainty factor indicating the likelihood of the entire pressure-welded joint structure included in the pressure-welded joint structure image based on the output of the machine learning model, and recognizing whether the entire pressure-welded joint structure is included based on the first certainty factor, and calculating second certainty factors indicating the likelihood of each of the pressure-welded portion, upper reinforcing bar, and lower reinforcing bar of the pressure-welded joint structure based on the output of the machine learning model, and recognizing the configuration of the pressure-welded joint structure based on each of the second certainty factors. and excluding the area recognized as the pressure-welded portion, the upper reinforcing bar, or the lower reinforcing bar from inspection of the pressure-welded joint structure based on the positional relationship between the area recognized as the pressure-welded portion based on the second certainty degree, the area recognized as the upper reinforcing bar based on the second certainty degree, and the area recognized as the lower reinforcing bar based on the second certainty degree, wherein the inspection step may inspect the pressure-welded joint structure when it is determined that the entire pressure-welded joint structure, the pressure-welded portion, the upper reinforcing bar, and the lower reinforcing bar are included.

[0011] (5) In one aspect of the present invention, the imaging step may display, on the image captured by the imaging unit, a guide image based on the shape of the pressure-welded joint structure and angle information between the pressure-welded joint structure and a terminal device equipped with the imaging unit.

[0012] (6) One aspect of the present invention is an information processing device comprising: an input unit that inputs measurement values ​​obtained by measuring the width of a reinforcing bar in a pressure-welded joint structure in which reinforcing bars are connected to each other by pressure welding, the measurement values ​​being obtained by using a measuring tool; a recognition unit that determines whether each of a plurality of components of the pressure-welded joint structure is included based on an image of the pressure-welded joint structure taken by a terminal device; a calculation unit that calculates the actual dimensions per pixel in the image of the pressure-welded joint structure based on the image of the pressure-welded joint structure taken by the terminal device and the measurement values ​​input by the input unit; and an inspection unit that, when it is recognized that the pressure-welded joint structure includes the plurality of components, inspects the pressure-welded joint structure based on the image showing the plurality of components and the actual dimensions per pixel in the image of the pressure-welded joint structure.

[0013] (7) One aspect of the present invention is a program that causes a computer of an information processing device to execute the following steps: an input step of inputting measurement values ​​measured using a measuring tool of the width of a reinforcing bar in a pressure-welded joint structure in which reinforcing bars are connected to each other by pressure welding; a recognition step of determining whether each of a plurality of components of the pressure-welded joint structure is included based on an image of the pressure-welded joint structure taken by a terminal device; a calculation step of calculating the actual dimensions per pixel in the image of the pressure-welded joint structure based on the image of the pressure-welded joint structure taken by the terminal device and the measurement values ​​input by the input step; and an inspection step of inspecting the pressure-welded joint structure based on the image showing the plurality of components and the actual dimensions per pixel in the image of the pressure-welded joint structure if it is recognized that the pressure-welded joint structure includes the plurality of components. [Effects of the Invention]

[0014] According to one aspect of the present invention, it is possible to inspect a pressure-welded joint structure more accurately. [Brief explanation of the drawings]

[0015] [Figure 1] FIG. 1 is a diagram showing an example of an outline of an inspection job in an inspection system according to an embodiment. [Figure 2] FIG. 10 is a diagram showing an example of sequential processing in the inspection processing method according to the embodiment. [Figure 3] FIG. 10 is a diagram illustrating an example of parallel processing in the inspection processing method according to the embodiment. [Figure 4] FIG. 10 is a diagram showing an example of batch processing among the inspection processing methods according to the embodiment. [Figure 5] 1 is a block diagram showing an example of the configuration of an inspection system 1 according to an embodiment. [Figure 6] FIG. 2 is a block diagram showing an example of a recognition unit 230 in the embodiment. [Figure 7] FIG. 2 is a diagram showing an example of a processing procedure of the inspection system 1 in the embodiment. [Figure 8]10A and 10B are diagrams for explaining that a pressure-contact portion is measured using a caliper or the like in the embodiment. [Figure 9] 5A to 5C are diagrams showing an example of a captured image, a guide image, and angle information displayed on the inspection terminal device 100 in the embodiment. [Figure 10] 10A and 10B are diagrams showing an example of recognition processing in an embodiment, in which (a) is a diagram showing an image of a pressure-welded joint structure before recognition, and (b) is a diagram showing an image in which the recognized area is superimposed on the pressure-welded joint structure image. [Figure 11] FIG. 10 is a diagram showing a process for correcting an area recognized by an image recognition engine 232 in an embodiment. [Figure 12] 10A and 10B are diagrams showing an example of a process for rotating an image in a process for calculating inspection parameters according to an embodiment, in which FIG. 10A shows the area before rotation, and FIG. 10B shows the area after rotation. [Figure 13] 10 is a diagram showing an example of a process for measuring a width D of a pressure-welded portion in a process for calculating an inspection parameter according to an embodiment. FIG. [Figure 14] 10 is a diagram showing an example of a process for measuring a reinforcing bar diameter d in a process for estimating a reinforcing bar diameter according to an embodiment. FIG. [Figure 15] FIG. 10 is a diagram showing an example of an image obtained when photographing the press-fit joint structure in a vertically oblique direction. [Figure 16] 10A and 10B are diagrams for explaining an example of a process for calculating the length L of the press-contact portion and a pass / fail determination process in the process for calculating the inspection parameters according to the embodiment. [Figure 17] 10A and 10B are diagrams for explaining an example of a process for calculating a bending angle θ and a pass / fail determination process in a process for calculating an inspection parameter according to an embodiment. [Figure 18] 10A and 10B are diagrams for explaining an example of a process for calculating an eccentricity amount e and a pass / fail determination process in the process for calculating inspection parameters according to the embodiment. [Figure 19] 10A and 10B are diagrams for explaining an example of a process for calculating a one-sided bulge Δh and a pass / fail determination process in the process for calculating inspection parameters according to an embodiment. [Figure 20]10A and 10B are diagrams for explaining an example of a process for calculating a significant flange shape F and a pass / fail determination process in the process for calculating inspection parameters according to the embodiment. [Figure 21] 1 is a diagram showing an inspection standard (viewpoint), measurement points, a first inspection standard, a second inspection standard, and processing of the embodiment. FIG. DETAILED DESCRIPTION OF THE INVENTION

[0016] Hereinafter, a method for inspecting a pressure-welded joint structure, an information processing device, and a program to which the present invention is applied will be described with reference to the drawings.

[0017] The inspection system of the embodiment inspects a pressure-welded joint structure using an image (hereinafter referred to as a pressure-welded joint structure image) of the pressure-welded joint structure captured by, for example, a building site worker. The inspection system of the embodiment inspects the pressure-welded joint structure with high accuracy by determining whether the pressure-welded joint structure to be inspected is present in the pressure-welded joint structure image and whether the pressure-welded joint structure conforms to the inspection standard. Furthermore, the inspection system of the embodiment can reduce the inspection effort and inspection reporting effort of the site worker by simply capturing an image of the pressure-welded joint structure and presenting highly accurate inspection results to the site worker. Note that, in the embodiment, a pressure-welded joint structure in which two reinforcing bars are gas-pressure welded is described as the inspection target, but a pressure-welded joint structure in which multiple reinforcing bars are gas-pressure welded may also be the inspection target. Furthermore, the inspection target of the embodiment is not limited to a pressure-welded joint structure that is gas-pressure welded, and pressure-welded joint structures that are pressure-welded by other methods may also be applicable.

[0018] <Inspection system overview> FIG. 1 is a diagram showing an example of an outline of an inspection job in an inspection system according to an embodiment. The inspection system of the embodiment is an information processing system for receiving inspection requests from a large number of user companies and handling inspection work. Inspection work is managed in the hierarchical structure shown in FIG. 1. An administrator at the user company uses the back office function of the back office terminal device 110 to link group information, construction information, and inspection location information in a tree structure. An inspector uses the inspection function of the inspection terminal device 100 to obtain group information, construction information, and inspection location information. Using the inspection function, the inspection terminal device 100 links and registers inspection result information for gas pressure welded joints with inspection location information. In the inspection system, for example, group information or construction information may include information indicating inspection standards. The information indicating inspection standards is information for selecting or specifying one of multiple inspection standards, such as a first inspection standard and a second inspection standard. The first inspection standard and the second inspection standard differ, for example, in the inspection method, inspection standard values, report format, etc. The first inspection standard is, for example, the "Japan Reinforced Concrete Joints Association Standard," and the second inspection standard is, for example, the "JIS Standard." Whether the first inspection standard or the second inspection standard is used depends on the user company, group, and construction site. This allows the inspection system to output inspection results collectively for each inspection location, which is the unit for performing inspection work.The inspection system can also output report information for each inspection location.The inspection system may output inspection results in addition to the report as report information.

[0019] FIG. 2 is a diagram showing an example of sequential processing in the inspection processing method according to the embodiment. The sequential processing involves a series of processes including photographing the welded joint structure using the inspection terminal device 100, recognition processing and pass / fail judgment processing using a machine learning model (AI), and display processing of the pass / fail judgment result. The sequential processing involves photographing the welded joint structure, recognition processing, pass / fail judgment processing, and display processing of the pass / fail judgment result for one welded joint structure to be inspected, and then processing for the next welded joint structure. This sequential processing is suitable for reliably confirming the pass / fail judgment result for each welded joint structure or for sharing the pass / fail judgment result for each welded joint structure with an on-site inspector during an on-site inspection.

[0020] FIG. 3 is a diagram showing an example of parallel processing in the inspection processing method according to the embodiment. In parallel processing, the recognition process and pass / fail judgment process begin once the photographing of the pressure-welded joint structure is completed. However, the inspection terminal device 100 performs photographing of the next pressure-welded joint structure in parallel with the recognition process and pass / fail judgment process. This allows the inspector to start the next photographing work without waiting for the pass / fail judgment result, improving work efficiency. The inspection terminal device 100 may notify the result of the pass / fail judgment process using the machine learning model as the text "pass" or "fail" in a notification field on the photographing screen. The inspection terminal device 100 allows the worker to check the status of the pass / fail judgment process and details of the pass / fail judgment result from an inspection status list in response to operation by the worker. Furthermore, even if the inspection terminal device 100 is temporarily unable to communicate at the work site, the photographing of the pressure-welded joint structure can proceed.

[0021] FIG. 4 is a diagram showing an example of batch processing among the inspection processing methods according to the embodiment. In batch processing, a plurality of pressure-welded joint structures are imaged by the inspection terminal device 100, and then recognition processing and pass / fail judgment processing are performed collectively on the plurality of pressure-welded joint structure images imaged by the inspection terminal device 100. This allows, for example, pressure-welded joint structures to be imaged and inspected by the inspection terminal device 100 even if the communication environment at the work site is insufficient, thereby expanding the range of applications for the inspection system.

[0022] <Inspection system configuration> FIG. 5 is a block diagram showing an example of the configuration of the inspection system 1 according to the embodiment. The inspection system 1 includes, for example, one or more inspection terminal devices 100 and an inspection server device 200. The inspection terminal devices 100 and the inspection server device 200 are connected to, for example, a communication network NW. Each device connected to the communication network NW includes a communication interface such as a network interface card (NIC) or a wireless communication module (not shown in FIG. 1). The communication network NW includes, for example, the Internet, a wide area network (WAN), a local area network (LAN), a cellular network, etc.

[0023] The inspection terminal device 100 is a portable terminal device equipped with a camera device (image capturing unit 102), such as a smartphone or tablet terminal. The inspection terminal device 100 includes, for example, the image capturing unit 102, a display unit 104, and a control unit 106. The inspection terminal device 100 launches a UA (User Agent) such as a browser or an application program. The UA is, for example, an application (control unit 106) for communicating with the inspection server device 200. The inspection terminal device 100 uses the inspection application as the UA to perform display processing on the display unit 104 using content received from the inspection server device 200, and to accept operations on a touch panel integrated with the display unit 104. The inspection terminal device 100 also uses the inspection application to provide operation information based on user operations, images of the pressure-welded joint structure, and the like to the inspection server device 200.

[0024] The inspection server device 200 is an information processing device that receives inspection requests for pressure-welded joint structures and provides inspection results for the pressure-welded joint structures. The inspection server device 200 includes, for example, an inspection processing unit 210 and a recognition unit 230. Functional units such as the inspection processing unit 210 and the recognition unit 230 are realized by a processor such as a CPU (Central Processing Unit) executing a program stored in a program memory. Furthermore, some or all of these functional units may be realized by hardware such as an LSI (Large Scale Integration), an ASIC (Application Specific Integrated Circuit), or an FPGA (Field-Programmable Gate Array), or may be realized by a combination of software and hardware.

[0025] The inspection processing unit 210 includes, for example, an input unit 212 , a calculation unit 214 , an inspection unit 216 , a reporting unit 218 , and a control unit 220 . The input unit 212 inputs a measurement value obtained by measuring the width of a pressure-welded joint structure in which reinforcing bars are connected together by pressure welding using a measuring tool. For example, a person operates the inspection terminal device 100 to input the width of the reinforcing bars in the pressure-welded joint structure measured using a measuring tool such as a vernier caliper to the inspection terminal device 100. The inspection terminal device 100 transmits measurement value information indicating the width of the pressure-welded joint structure to the inspection server device 200 via the communication network NW. This allows the input unit 212 to acquire the measurement value based on the received measurement value information. The calculation unit 214 calculates the actual size per pixel in the press-fit joint structure image based on the press-fit joint structure image captured by the inspection terminal device 100 and the measurement values ​​input by the input unit 212 . When it is recognized that the pressure-welded joint structure includes a plurality of components, the inspection unit 216 inspects the pressure-welded joint structure based on an image showing the plurality of components and the actual dimensions per pixel in the pressure-welded joint structure image. The inspection unit 216 outputs a pass / fail judgment result based on the inspection standard as the inspection result. The reporting unit 218 creates inspection result information including the pass / fail judgment result output from the inspection unit 216 . The control unit 220 comprehensively controls the operation of the inspection processing unit 210. The control unit 220 receives requests from the inspection terminal device 100 and transmits various content data to the inspection terminal device 100. The content data is, for example, data for displaying a registration screen before the start of the inspection, data for displaying a screen during the inspection, and data for displaying the pass / fail judgment result. Furthermore, the control unit 220 controls the activation of the image recognition engine 232 in the recognition unit 230 during the inspection.

[0026] The recognition unit 230 determines whether or not each of the multiple components of the insulation displacement joint structure is included based on the insulation displacement joint structure image captured by the inspection terminal device 100.

[0027] FIG. 6 is a block diagram showing an example of the recognition unit 230 according to the embodiment. The recognition unit 230 includes, for example, a training image database 240, a training processing unit 242, and a recognition processing unit 244.

[0028] The training image database 240 is a database that stores, for example, images previously acquired from the inspection terminal device 100 or other data server devices as training images. The training image database 240 stores, for example, positive example images as training images. A positive example image is an image of a pressure-welded joint structure in a normal state. A positive example image is, for example, an image of the entire rebar, an image of the upper rebar, an image of the lower rebar, and an image of the pressure-welded portion formed by pressure-welding the upper rebar and the lower rebar, an image of the upper rebar, an image of the lower rebar, and an image of the pressure-welded portion. The positive example image is assigned tag information corresponding to the entire rebar, tag information corresponding to the upper rebar, tag information corresponding to the lower rebar, or tag information corresponding to the pressure-welded portion, and is registered in the training image database 240. In addition to positive example images, images of objects other than the pressure-welded joint structure (objects in an abnormal state or defective products) may be stored as training images in the training image database 240. The upper reinforcing bar, the lower reinforcing bar, and the pressure-welded portion are each an example of a component part of the pressure-welded joint structure.

[0029] The learning processing unit 242 receives training images from the training image database 240. The learning processing unit 242 performs recognition processing using the training images to obtain a determination result. When a positive example image is input, the learning processing unit 242 learns processing parameters of the recognition processing unit 244 (recognition model 244A) so that a determination result is obtained that the entire pressure-welded joint structure or a component of the pressure-welded joint structure is present, and generates learning result data. When an image different from the pressure-welded joint structure is input, the learning processing unit 242 learns processing parameters of the recognition processing unit 244 (recognition model 244A) so that a determination result is obtained that the entire pressure-welded joint structure or a component of the pressure-welded joint structure is not present, and generates learning result data. The processing parameters of the recognition processing unit 244 (recognition model 244A) are, for example, filters (also referred to as weights or biases) included in a neural network. The learning result data is stored in the learning result data storage unit 2441.

[0030] The recognition processing unit 244 is a component corresponding to the image recognition engine 232. The recognition processing unit 244 has, for example, a learning result data storage unit 2441 and a determination unit 2442. The learning result data storage unit 2441 accumulates processing parameters of the recognition processing unit 244 as learning results, and the processing parameters of the recognition processing unit 244 are updated by the learning processing unit 242. When the recognition processing unit 244 acquires a captured image from the inspection terminal device 100, it extracts feature amounts using a recognition model 244A based on the processing parameters of the recognition processing unit 244. The determination unit 2442 determines whether the entire pressure-welded joint structure and each of the component parts of the pressure-welded joint structure are present based on the extracted feature amounts. The recognition processing unit 244 outputs the determination result by the determination unit 2442 as the recognition result. In addition, the recognition unit 230 generates an image (mask image) representing the area within the captured image that is recognized as the entire pressure-welded joint structure, and images (mask images, images showing the component parts) representing the areas within the image that are recognized as each of the component parts of the pressure-welded joint structure, and passes these to the inspection processing unit 210.

[0031] FIG. 7 is a diagram showing an example of a processing procedure of the inspection system 1 in the embodiment. First, the inspection terminal device 100 starts an inspection application based on an operation by an inspector (step S100). The inspection terminal device 100 displays on the display unit 104 a screen that prompts an operation to input basic inspection information such as information specifying the construction work, information specifying the inspection processing method, and information selecting the inspection standard using the inspection application. The inspection terminal device 100 acquires the information specifying the construction work, information specifying the inspection processing method, and information selecting the inspection standard based on the operation of the inspector, and transmits them to the inspection server device 200. As a result, the inspection system 1 accepts an operation to select either the first inspection standard or the second inspection standard before the inspection.

[0032] Next, the inspection terminal device 100 displays on the display unit 104 a screen for inputting the measured value of the rebar width used in the pressure-welded joint structure. The rebar width is an actual size corresponding to "d" described below. The inspection terminal device 100 acquires the measurement value information of the rebar width measured with a measuring tool based on the operation of the inspector. The inspection terminal device 100 transmits the acquired measurement value information of the rebar width to the inspection server device 200, and the input unit 212 of the inspection server device 200 inputs the measurement value information of the rebar width (step S102). 8 is a diagram for explaining that the pressure-welded portion is measured using a vernier caliper etc. An inspector measures the actual dimension d of the portion other than the pressure-welded portion using a measuring tool such as a vernier caliper.

[0033] The inspection terminal device 100 starts up the imaging unit 102 in response to control of the inspection application, begins photographing the pressure-welded joint structure, and displays the photographed pressure-welded joint structure image on the display unit 104. The inspection terminal device 100 also displays a guide image superimposed on the photographed pressure-welded joint structure image. In this way, the inspection terminal device 100 acquires a pressure-welded joint structure image by photographing the entire rebar with the guide image displayed on the display unit 104, and transmits the acquired pressure-welded joint structure image to the inspection server device 200. In this way, the input unit 212 inputs the pressure-welded joint structure (step S104).

[0034] FIG. 9 is a diagram showing an example of a captured image, a guide image, and angle information displayed on the inspection terminal device 100 in the embodiment. The guide image is an image that assists the inspection terminal device 100 in capturing an image so that the pressure-welded joint structure is included within a predetermined range within the captured image. The guide image includes, for example, an upper guide image, a pressure-welded portion guide image, and a lower guide image. Furthermore, the inspection terminal device 100 may display a message image at the top of the screen saying, "Please insert the rebar into the guide and take a photo." This allows the inspection terminal device 100 to capture an image that includes the entire rebar, the upper rebar, the lower rebar, and the pressure-welded portion when the capture button is selected with the upper guide image aligned with the upper rebar, the pressure-welded portion guide image aligned with the pressure-welded portion, and the lower guide image aligned with the lower rebar.

[0035] As shown in the lower left of FIG. 9 , the inspection terminal device 100 may display, on the image captured by the imaging unit 102 during shooting, a guide image based on the shape of the welded joint structure and angle information between the welded joint structure and the inspection terminal device 100 equipped with the imaging unit 102. The angle information is information indicating the inclination of the inspection terminal device 100 with respect to the ground. The angle information indicates the value of the terminal angle detected by an angle sensor built into the inspection terminal device 100. When the rebar is installed vertically from the ground, the inspection terminal device 100 shoots so that the terminal angle is 90 degrees. When the rebar is installed horizontally to the ground, the inspection terminal device 100 shoots so that the terminal angle is 0 degrees. This angle correction assist function of the inspection terminal device 100 can assist the user in correcting the shooting angle.

[0036] The image recognition engine 232 of the inspection server device 200 acquires the pressure-welded joint structure image and performs a recognition process to determine whether the entire rebar, upper rebar, lower rebar, and pressure-welded portion are present in the pressure-welded joint structure image (step S106). At this time, the image recognition engine 232 inputs the pressure-welded joint structure image into the machine learning model, calculates a first certainty factor indicating the certainty of the entire rebar, and recognizes whether the entire pressure-welded joint structure is included based on the first certainty factor. Also, the image recognition engine 232 inputs the pressure-welded joint structure image into the machine learning model, calculates second certainty factors indicating the certainty of each of the upper rebar, the lower rebar, and the pressure-welded part, and recognizes whether each of the components of the pressure-welded joint structure is included based on each of the second certainty factors. For example, if the confidence levels for the entire rebar, the upper rebar, the lower rebar, and the pressure-welded portion are each equal to or greater than a threshold, the image recognition engine 232 determines that the entire rebar, the upper rebar, the lower rebar, and the pressure-welded portion are present. If the inspection server device 200 recognizes that the entire rebar, the upper rebar, the lower rebar, and the pressure-welded portion are all present (step S108: YES), the inspection server device 200 proceeds to step S110. If the inspection server device 200 recognizes that any of the entire rebar, the upper rebar, the lower rebar, and the pressure-welded portion are not present (step S108: NO), the inspection server device 200 notifies the inspection terminal device 100 to re-take a photograph (step S112) and re-acquires the pressure-welded joint structure image (step S104).

[0037] Note that the image recognition engine 232 determined whether the certainty of each of the entire rebar, the upper rebar, the lower rebar, and the pressure-welded portion exceeded a threshold, but is not limited to this. If the total value of the certainty of the entire rebar, the upper rebar, the lower rebar, and the pressure-welded portion exceeds a threshold, the image recognition engine 232 may generate a recognition result indicating that all of the entire rebar, the upper rebar, the lower rebar, and the pressure-welded portion are present. Furthermore, the image recognition engine 232 may change the certainty threshold for each of the entire rebar, the upper rebar, the lower rebar, and the pressure-welded portion. This allows the image recognition engine 232 to recognize each of the entire rebar, the upper rebar, the lower rebar, and the pressure-welded portion with high accuracy.

[0038] FIG. 10 is a diagram showing an example of recognition processing in an embodiment, where (a) is a diagram showing an image of a welded joint structure before recognition, and (b) is a diagram showing an image in which the recognized area is superimposed on the image of the welded joint structure. The image recognition engine 232 inputs the pressure-welded joint structure image shown in Fig. 10(a) into a machine learning model (the image recognition engine 232 itself), and recognizes the upper rebar region, the pressure-welded portion region, the lower rebar region, and the entire rebar region shown in Fig. 10(b). The recognized regions are used to mean the regions that show the respective outer shapes (polygons) of the entire rebar, the upper rebar, the lower rebar, and the pressure-welded portion included in the pressure-welded joint structure image. The image recognition engine 232 assigns tag information to each of the upper rebar area, the pressure-welded area, the lower rebar area, and the entire rebar area. For example, the upper rebar area is assigned tag information with item number "1," the lower rebar area is assigned tag information with item number "2," the pressure-welded area is assigned tag information with item number "3," and the entire rebar area is assigned tag information with item number "4."

[0039] FIG. 11 is a diagram showing a process for correcting an area recognized by the image recognition engine 232 in this embodiment. The inspection server device 200 may exclude areas recognized as a pressure-welded portion, an upper reinforcing bar, or a lower reinforcing bar from inspection of the pressure-welded joint structure based on the positional relationship between the area recognized as a pressure-welded portion based on the second certainty level, the area recognized as an upper reinforcing bar based on the second certainty level, and the area recognized as a lower reinforcing bar based on the second certainty level. The multiple regions recognized by the image recognition engine 232 are classified into, for example, patterns 1 to 5. When there are holes in multiple regions as in pattern 1, the recognition unit 230 ignores the holes. When three regions are adjacent to each other but one region is an outlying region, as in pattern 2, the recognition unit 230 deletes the outlying region. When one small area is isolated from one large area as in pattern 3, the recognition unit 230 deletes the isolated area. When the image is divided into two regions aligned in a predetermined direction as in pattern 4, the recognition unit 230 ignores the division because it does not affect the calculation of the width or center line by the calculation unit 214. When nothing is recognized in a portion surrounded by multiple regions, as in pattern 5, and the multiple regions are separated into islands, the recognition unit 230 determines that a recognition error has occurred.

[0040] The calculation unit 214 calculates the size [mm] per pixel in the pressure-welded joint structure image from the measurement value input in step S102 and the recognition result of the pressure-welded joint structure image input in step S104 (step S106) (step S110). The calculation unit 214 obtains dp as the number of pixels per mm from the ratio of the number d of pixels of the rebar diameter in the pressure-welded joint structure image to the measurement value da [mm], as in Equation 1. dp[pixel / mm]= da[mm]÷d[pixel] (Formula 1) For example, if da is 23.8 millimeters and d is 100 pixels, the calculation unit 214 calculates da as 0.238 [pixel / mm] from 23.8 divided by 100. This allows the calculation unit 214 to estimate the actual size per pixel. By using the rebar size [mm] as a reference, the inspection server device 200 can reliably measure the measurement values ​​[mm] of each part, which will be described later.

[0041] The inspection unit 216 calculates inspection parameters (step S114). At this time, the inspection unit 216 calculates the parameters using an image showing the component parts acquired from the image recognition engine 232. The image showing the component parts is an inspection image, and is an image showing an area of ​​the captured image that has been recognized as a component part by the image recognition engine 232 (recognition model).

[0042] Next, the inspection unit 216 performs an inspection using the calculated calculation parameters (step S116). The inspection process is a pass / fail judgment process that indicates whether the selected inspection criterion is met. If a first inspection criterion is selected, the inspection unit 216 creates a pass / fail judgment result according to the first inspection criterion, and if a second inspection criterion is selected, the inspection unit 216 creates a pass / fail judgment result according to the second inspection criterion. The details of the calculation of the inspection parameters and the inspection details will be described later. The inspection server device 200 may transmit pass / fail judgment results corresponding to the pressure-welded joint structure images to the inspection terminal device 100, causing the inspection terminal device 100 to display the pass / fail judgment results for each pressure-welded joint structure image. The details of the inspection details and the display of the inspection results will be described later.

[0043] Next, the reporting unit 218 creates test result information indicating the test results corresponding to the selected test criteria (step S118). Next, the testing server device 200 transmits the created test result information to the test terminal device 100 (step S120). If a first test criteria is selected, the reporting unit 218 creates test result information including a pass / fail judgment result calculated based on the first test criteria, and if a second test criteria is selected, the reporting unit 218 creates test result information including a pass / fail judgment result calculated based on the second test criteria. The test result information is, for example, information indicating the test contents and results corresponding to the test criteria and the test items, and is information that causes the test terminal device 100 to display content including the test contents and test results. Next, the testing server device 200 determines whether the testing is complete (step S122). For example, if the testing server device 200 receives an operation to end the testing (step S122: YES), the testing server device 200 determines that the testing is complete and ends the processing of this flowchart. For example, if the testing server device 200 receives an operation to proceed to the next image capture (step S122: NO), the testing server device 200 inputs the press-welded joint structure image for the next test object in step S104 (step S124), and repeats the processing from step S106 onwards.

[0044] 12A and 12B are diagrams showing an example of image rotation processing in the process of calculating inspection parameters according to an embodiment, where (a) is a diagram showing the area before rotation and (b) is a diagram showing the area after rotation. Note that in FIG. 12, the bending of the pressure-welded joint structure is exaggerated for the purpose of explaining the processing. 12(a), the calculation unit 214 calculates the center line of each of the upper reinforcing bar region and the lower reinforcing bar region by the least squares method, and calculates the average of the center line positions as the center line of the entire reinforcing bar. Note that the center line may be set using an area excluding the upper and lower ends of the pressure welded joint structure. Next, the calculation unit 214 determines the center coordinates of the pressure-welded area and sets it as the center of the entire image, and rotates the entire image so that the center line of the entire reinforcing bar is vertical, as shown in FIG. 12(b). By correcting the orientation of the pressure-contact region, the calculation unit 214 can correctly measure the width and length of the pressure-contact region.

[0045] FIG. 13 is a diagram showing an example of a process for measuring a reinforcing bar diameter d in a process for estimating a reinforcing bar diameter according to the embodiment. The calculation unit 214 sets an upper mask region that covers the upper end region of the press-welded portion of the press-welded joint structure, and a lower mask region that covers the lower end region of the press-welded portion of the press-welded joint structure. Furthermore, the calculation unit 214 sets an upper end mask region that covers the upper end region of the upper rebar of the press-welded joint structure, and a lower end mask region that covers the lower end region of the lower rebar of the press-welded joint structure. The calculation unit 214 makes the vertical lengths of the upper mask region and the lower mask region shorter than the vertical lengths of the upper end mask region and the lower end mask region. The calculation unit 214 makes the size of the mask region smaller the closer it is to the center of the image of the press-welded joint structure. For example, the calculation unit 214 sets the vertical lengths of the upper mask region and the lower mask region to 1 / 5 of the width D of the pressure-welded portion. As shown in Fig. 13(a), the calculation unit 214 sets the lengths of the upper mask region and the lower mask region to 1 / 5 of the width D of the pressure-welded portion, the length of the upper end mask region to 25% of the upper rebar region, and the length of the lower end mask region to 25% of the lower rebar region. The calculation unit 214 obtains the maximum number of pixels d1 in the width of the upper rebar area excluding the upper mask area, the lower mask area, the upper end mask area, and the lower end mask area, as shown in Figure 13(b), and obtains the maximum number of pixels d2 in the width of the lower rebar area. The calculation unit 214 adopts the smaller of the maximum number of pixels d1 acquired from the upper rebar region and the maximum number of pixels d2 acquired from the lower rebar region as the rebar diameter d of the pressure welded joint structure.

[0046] FIG. 14 shows an example of an image of a pressure-welded joint structure captured in a vertically oblique direction. For example, when a pressure-welded joint structure is captured from above, the actual rebar diameter d is constant, but the higher the pressure-welded joint structure image, the larger it appears. In response to this, the calculation unit 214 can reduce the vertical length of the upper mask region and the lower mask region by making the vertical length of the upper end mask region and the lower end mask region shorter than the vertical length of the upper end mask region and the lower end mask region, thereby reducing the impact of capturing the pressure-welded joint structure from an oblique direction. Furthermore, even if an end of the region recognized by the recognition unit 230 is missing and deviates from the pressure-welded joint structure in the image, the calculation unit 214 can avoid measuring the pressure-welded joint structure based on the size of the end region.

[0047] FIG. 15 is a diagram showing an example of a process for measuring the width D of the press-contact portion in the process for calculating the inspection parameters according to the embodiment. With the crimped joint structure image rotated as described above, the calculation unit 214 acquires the leftmost and rightmost positions in the crimped portion region, and acquires the number of pixels from the leftmost to the rightmost positions as the crimped portion width D. The calculation unit 214 can acquire the actual size of the crimped portion from the crimped portion width D [number of pixels] and the actual size per pixel. Inspection unit 216 determines that the inspection of width D of the pressure-welded portion passes if width D of the pressure-welded portion is equal to or greater than a reference value. The reference value is changed depending on the inspection standard selected by the witness inspector from the first inspection standard and the second inspection standard. When the first inspection standard is selected, the inspection unit 216 sets the reference value to the length written in the numerical portion of the name of the welded joint structure (for example, 22.0 mm for D22) x the number of pixels per mm dp x the inspection standard coefficient of "1.4". When the second inspection standard is selected, the inspection unit 216 sets the reference value to the nominal diameter (for example, 22.2 mm for D22) specified for each type of rebar x the number of pixels per mm dp x the inspection standard coefficient, which is "1.4." This allows the testing server device 200 to accurately determine whether the width D of the pressure-contact portion is acceptable or not.

[0048] FIG. 16 is a diagram for explaining an example of the process for calculating the length L of the press-contact portion and the pass / fail determination process in the process for calculating the inspection parameters according to the embodiment. The calculation unit 214 calculates the average value of the rebar diameters in the upper rebar area and the lower rebar area, removes a rectangular area of ​​the width of the average value from the pressure-welded area, and determines the number of pixels from the upper end to the lower end of the remaining pressure-welded area as the length L of the pressure-welded area. If the length L of the pressure-contact portion is equal to or greater than a reference value, the inspection unit 216 determines that the inspection of the length L of the pressure-contact portion has passed. The reference value is changed depending on the inspection standard selected by the inspector from the first inspection standard and the second inspection standard. When the first inspection standard is selected, the inspection unit 216 sets the reference value to the length written in the numerical portion of the name of the welded joint structure (for example, 22.0 mm for D22) x the number of pixels per mm dp x the inspection standard coefficient of "1.1". When the second inspection standard is selected, the inspection unit 216 sets the reference value to the nominal diameter (for example, 22.2 mm for D22) specified for each type of rebar x the number of pixels per mm dp x the inspection standard coefficient of "1.1". This allows the testing server device 200 to accurately determine whether the length L of the pressure-contact portion is acceptable or not.

[0049] FIG. 17 is a diagram for explaining an example of the process of calculating the bending angle θ and the pass / fail determination process in the process of calculating the inspection parameters according to the embodiment. The calculation unit 214 calculates the bending angle θ between the upper reinforcing bar region and the lower reinforcing bar region based on the difference in the inclination between the center line of the upper reinforcing bar region and the center line of the lower reinforcing bar region. The inspection unit 216 determines that the inspection of the bending angle θ is passed if the bending angle θ calculated by the calculation unit 214 is equal to or smaller than a reference value according to the inspection standard. The reference value is changed according to the inspection standard selected by the inspector from the first inspection standard and the second inspection standard. When the first inspection standard is selected, the inspection section 216 sets the standard value to "2.0." When the second inspection standard is selected, the inspection section 216 sets the standard value to "3.5". This allows the testing server device 200 to accurately determine whether the bending angle θ is acceptable or not.

[0050] FIG. 18 is a diagram for explaining an example of the process for calculating the amount of eccentricity e and the pass / fail determination process in the process for calculating the inspection parameters according to the embodiment. The calculation unit 214 obtains the intersection between the center line of the upper rebar area and the center line connecting the left end of the pressure-welded portion and the right end of the pressure-welded portion, and the intersection between the center line of the lower rebar area and the center line of the pressure-welded portion, and calculates the distance [number of pixels] between the two intersections as the eccentricity e. The inspection unit 216 determines that the inspection of the eccentricity e has passed if the eccentricity e calculated by the calculation unit 214 is equal to or less than the reference value. The reference value is changed depending on the inspection standard selected by the inspector from the first inspection standard and the second inspection standard. When the first inspection standard is selected, the inspection unit 216 sets the reference value to the length written in the numerical portion of the name of the welded joint structure (for example, 22.0 mm for D22) x the number of pixels per mm dp x the inspection standard coefficient of "1 / 5". When the second inspection standard is selected, the inspection unit 216 sets the reference value to the nominal diameter (for example, 22.2 mm for D22) specified for each type of rebar x the number of pixels per mm dp x the inspection standard coefficient of "1 / 5". This allows the testing server device 200 to accurately determine whether the eccentricity amount e is acceptable or not.

[0051] FIG. 19 is a diagram for explaining an example of the process for calculating the one-sided bulge Δh and the pass / fail determination process in the process for calculating the inspection parameters according to the embodiment. The calculation unit 214 extends an upper rectangle whose width is the diameter of the upper reinforcing bar to the pressure-welded portion along the inclination of the center line of the upper reinforcing bar, and extends a lower rectangle whose width is the diameter of the lower reinforcing bar to the pressure-welded portion along the inclination of the center line of the lower reinforcing bar. The calculation unit 214 calculates the distance h between the left end of the pressure-welded portion area and the upper rectangle. 上1 and the distance h between the right edge of the pressure weld area and the upper rectangle 上2 , the distance h between the left edge of the pressure weld area and the lower rectangle 下1 and the distance h between the right edge of the pressure weld area and the lower rectangle 下2 The calculation unit 214 measures h 上1 and h 上2 Absolute value of the difference Δh 上 , h 下1 and h 下2 Absolute value of the difference Δh 下 The calculation unit 214 calculates the absolute value Δh 上 and absolute value Δh 下 is the one-sided bulge, and the larger one is the one-sided bulge Δh [pixels] of the pressure-welded joint structure. The inspection unit 216 determines that the inspection for the one-sided bulge Δh is passed if the one-sided bulge Δh [pixels] calculated by the calculation unit 214 is equal to or less than the reference value. The reference value is changed depending on the inspection standard selected by the inspector from the first inspection standard and the second inspection standard. When the first inspection standard is selected, the inspection unit 216 sets the reference value to the length written in the numerical portion of the name of the welded joint structure (for example, 22.0 mm for D22) x the number of pixels per mm dp x the inspection standard coefficient of "1 / 5". When the second inspection standard is selected, the inspection unit 216 sets the reference value to the nominal diameter (for example, 22.2 mm for D22) specified for each type of rebar x the number of pixels per mm dp x the inspection standard coefficient of "1 / 5". This allows the testing server device 200 to accurately determine whether the one-sided bulge Δh is acceptable or not.

[0052] FIG. 20 is a diagram for explaining an example of the process of calculating the significant flange shape F and the pass / fail determination process in the process of calculating the inspection parameters according to the embodiment. The calculation unit 214 calculates the significant flange shape F by calculating F=D÷L using the width D of the crimped portion and the length L of the crimped portion calculated as described above. The significant flange shape F is a value that indicates the ratio of the width D of the crimped portion to the length L of the crimped portion. If the significant flange shape F calculated by the calculation unit 214 is equal to or less than the inspection reference value of "1.6", the inspection unit 216 determines that the inspection for the significant flange shape F has passed. This allows the testing server device 200 to accurately determine whether the flange shape F is significant or not.

[0053] 21 is a diagram showing the inspection criteria (viewpoints), measurement locations, first inspection criteria, second inspection criteria, and processing of the embodiment. The inspection system 1 of the embodiment switches the rebar width d depending on whether the first inspection criteria or the second inspection criteria is selected, and calculates the width D of the pressure-welded portion, the length L of the pressure-welded portion, the eccentricity e, the bending angle θ, the one-sided bulge Δh, and the significant flange shape, and can make a pass / fail judgment.

[0054] <Effects of the embodiment> According to the inspection system 1 of the embodiment described above, the system includes an input step of inputting measurement values ​​measured using a measuring tool for the width of reinforcing bars in a pressure-welded joint structure in which reinforcing bars are connected to each other by pressure welding; an imaging step of imaging the pressure-welded joint structure; a recognition step of determining whether each of the multiple components of the pressure-welded joint structure is included based on the pressure-welded joint structure image captured in the imaging step; a calculation step of calculating the actual dimensions per pixel in the pressure-welded joint structure image based on the pressure-welded joint structure image and the measurement values ​​input in the input step; and, if it is recognized that the pressure-welded joint structure includes multiple components, an inspection step of inspecting the pressure-welded joint structure based on the image showing the multiple components and the actual dimensions per pixel in the pressure-welded joint structure image.Therefore, the pressure-welded joint structure can be inspected more accurately by dividing it into its component parts.

[0055] According to the embodiment, the inspection system 1 includes, before the inspection step, a selection step of accepting an operation to select either a first inspection standard or a second inspection standard, and a reporting step of creating inspection result information including a pass / fail judgment result calculated based on the first inspection standard if the first inspection standard is selected in the selection step, and creating inspection result information including a pass / fail judgment result calculated based on the second inspection standard if the second inspection standard is selected in the selection step. The inspection step can create a pass / fail judgment result according to the first inspection standard if the first inspection standard is selected in the selection step, and create a pass / fail judgment result according to the second inspection standard if the second inspection standard is selected in the selection step. Thus, according to the inspection system 1, even if there are multiple different inspection standards, the pass / fail judgment result can be created according to the selected inspection standard. This allows, for example, the selection of either the "Japan Reinforced Concrete Joints Association Standard" or the "JIS Standard" to be input in advance, and when determining pass / fail, the calculation unit 214 can perform calculations depending on which of the specified inspection standards is selected, and further, when outputting the pass / fail determination result, inspection result information in accordance with the specified inspection standard can be output.

[0056] According to the inspection system 1 of the embodiment, based on the positional relationship of the area recognized as the pressure-welded portion, the area recognized as the upper reinforcing bar, and the area recognized as the lower reinforcing bar, the area recognized as the pressure-welded portion, the upper reinforcing bar, or the lower reinforcing bar is excluded from the inspection of the pressure-welded joint structure, and the pressure-welded joint structure is inspected only when it is determined that the entire pressure-welded joint structure, the pressure-welded portion, the upper reinforcing bar, and the lower reinforcing bar are included. As a result, the inspection system 1 can accurately inspect the target pressure-welded joint structure even when the image contains reinforcing bars other than the target or objects of similar shape, as when imaging and inspecting a specific pressure-welded joint structure in an environment where many pressure-welded joint structures are present.

[0057] According to the embodiment of the inspection system 1, a guide image based on the shape of the pressure-welded joint structure and angle information between the pressure-welded joint structure and the inspection terminal device 100 equipped with the imaging unit 102 are displayed on the image captured by the imaging unit 102, thereby assisting in capturing images of the inspection terminal device 100 and the pressure-welded joint structure in a parallel position, thereby improving measurement accuracy.

[0058] In addition, the programs for executing the various processes of the inspection terminal device 100 and the inspection server device 200 in this embodiment may be recorded on a computer-readable recording medium, and the programs recorded on the recording medium may be read into and executed by a computer system, thereby performing the various processes described above related to the inspection terminal device 100 and the inspection server device 200.

[0059] Note that the term "computer system" here may include hardware such as the OS and peripheral devices. Furthermore, if a WWW system is used, the term "computer system" also includes the homepage provision environment (or display environment). Furthermore, "computer-readable recording media" refers to storage devices such as flexible disks, magneto-optical disks, ROMs, and writable non-volatile memory such as flash memory, portable media such as CD-ROMs, and hard disks built into computer systems.

[0060] Furthermore, "computer-readable recording medium" refers to the volatile memory (e.g., DRAM (Dynamic Random Access Memory)) within a computer system that acts as a server or client when a program is transmitted via a network such as the Internet or a communication line such as a telephone line. This also includes devices that hold a program for a certain period of time, such as a random access memory (Random Access Memory). The program may also be transmitted from a computer system that stores the program in a storage device to another computer system via a transmission medium or by transmission waves in the transmission medium.

[0061] Here, the "transmission medium" for transmitting the program refers to a medium that has the function of transmitting information, such as a network (communication network) such as the Internet or a communication line (communication line) such as a telephone line. The program may also be one that realizes part of the above-mentioned functions. Furthermore, it may be a so-called differential file (differential program) that can realize the above-mentioned functions in combination with a program already recorded in a computer system.

[0062] Although the embodiment of the present invention has been described above in detail with reference to the drawings, the specific configuration is not limited to this embodiment, and includes designs within the scope of the gist of the present invention. For example, an embodiment of the present invention may be configured with only the inspection terminal device 100. [Explanation of symbols]

[0063] 100 Inspection terminal device 102 Imaging unit 104 Display section 106 Control Unit 200 Inspection server device 210 Inspection processing section 212 Input section 214 Calculation Department 216 Inspection Department 218 Reporting Department 220 Control Unit 230 Recognition part 232 Image Recognition Engine 240 training image database 242 Learning processing unit 244 Recognition processing section 244A Recognition Model 2441 Learning result data storage unit 2442 Judgment section

Claims

1. an input step in which an information processing device inputs a measurement value obtained by measuring the width of the reinforcing bars in a pressure-welded joint structure in which reinforcing bars are connected to each other by pressure welding, the measurement value being measured by a measuring tool; an imaging step in which an inspection terminal device images the insulation displacement joint structure; a recognition step in which the information processing device determines whether each of a plurality of components of the pressure-welded joint structure is included based on the pressure-welded joint structure image captured in the imaging step; a calculation step in which the information processing device calculates an actual size per pixel in the press-fit joint structure image based on the press-fit joint structure image and the measurement values ​​input in the input step; an inspection step in which, when the information processing device recognizes that the pressure-welded joint structure includes the plurality of components, the information processing device inspects the pressure-welded joint structure based on an image showing the plurality of components and an actual size per pixel in the pressure-welded joint structure image; A method for inspecting a pressure-welded joint structure, comprising:

2. a selection step in which the information processing device receives an operation to select either a first inspection standard or a second inspection standard before the inspection step; a reporting step in which the information processing device creates inspection result information including a pass / fail judgment result calculated based on the first inspection standard when the first inspection standard is selected in the selection step, and creates inspection result information including a pass / fail judgment result calculated based on the second inspection standard when the second inspection standard is selected in the selection step, the inspection step creates a pass / fail judgment result in accordance with the first inspection standard when the first inspection standard is selected in the selection step, and creates a pass / fail judgment result in accordance with the second inspection standard when the second inspection standard is selected in the selection step; A method for inspecting a pressure-welded joint structure according to claim 1.

3. The inspection step includes: an upper mask region covering an upper end region of the pressure-welded portion of the pressure-welded joint structure, and a lower mask region covering a lower end region of the pressure-welded portion of the pressure-welded joint structure; An upper end mask area covering an upper end area of ​​the upper reinforcing bar in the pressure welding joint structure and a lower end mask area covering a lower end area of ​​the lower reinforcing bar in the pressure welding joint structure are set, The vertical lengths of the upper mask region and the lower mask region are made shorter than the vertical lengths of the upper end mask region and the lower end mask region. A method for inspecting a pressure-welded joint structure according to claim 1.

4. The recognition step includes: inputting the pressure-welded joint structure image into a machine learning model; calculating a first confidence level indicating the likelihood of the entire pressure-welded joint structure being included in the pressure-welded joint structure image based on the output of the machine learning model, and recognizing whether the entire pressure-welded joint structure is included based on the first confidence level; A step of calculating second certainty factors indicating the likelihood of each of the pressure-welded portion, upper reinforcing bar, and lower reinforcing bar of the pressure-welded joint structure based on the output of the machine learning model, and recognizing whether each of the components of the pressure-welded joint structure is included based on each of the second certainty factors; and excluding the area recognized as the pressure-welded portion, the upper reinforcing bar, or the lower reinforcing bar from the inspection of the pressure-welded joint structure based on a positional relationship between the area recognized as the pressure-welded portion based on the second certainty level, the area recognized as the upper reinforcing bar based on the second certainty level, and the area recognized as the lower reinforcing bar based on the second certainty level, In the inspection step, when it is determined that the entire pressure-welded joint structure, the pressure-welded portion, the upper reinforcing bar, and the lower reinforcing bar are included, the pressure-welded joint structure is inspected. A method for inspecting a pressure-welded joint structure according to claim 1.

5. 2. The method for inspecting a pressure-welded joint structure according to claim 1, wherein the imaging step displays a guide image based on the shape of the pressure-welded joint structure and angle information between the pressure-welded joint structure and a terminal device equipped with the imaging unit on the image captured by the imaging unit.

6. An input unit for inputting a measurement value obtained by measuring the width of the reinforcing bar in a pressure-welded joint structure in which reinforcing bars are connected to each other by pressure welding, using a measuring tool; a recognition unit that determines whether each of a plurality of components of the pressure-welded joint structure is included based on an image of the pressure-welded joint structure captured by a terminal device; a calculation unit that calculates an actual size per pixel in the press-welded joint structure image based on the press-welded joint structure image captured by the terminal device and the measurement values ​​input by the input unit; an inspection unit that, when it is recognized that the plurality of components are included in the pressure-welded joint structure, inspects the pressure-welded joint structure based on an image showing the plurality of components and an actual size per pixel in the pressure-welded joint structure image; An information processing device comprising:

7. The computer of the information processing device An input step of inputting a measurement value obtained by measuring the width of the reinforcing bar in a pressure-welded joint structure in which reinforcing bars are connected to each other by pressure welding using a measuring tool; a recognition step of determining whether each of a plurality of components of the pressure-welded joint structure is included based on an image of the pressure-welded joint structure captured by a terminal device; a calculation step of calculating an actual size per pixel in the press-welded joint structure image based on the press-welded joint structure image captured by a terminal device and the measurement values ​​input in the input step; an inspection step of inspecting the pressure-welded joint structure based on an image showing the plurality of components and an actual size per pixel in the pressure-welded joint structure image when it is recognized that the pressure-welded joint structure includes the plurality of components; A program that executes the following.

Citation Information

Patent Citations

  • Flow velocity distribution measurement method

    JP2017161329A

  • Building inspection method, information processing device, terminal device, program and building inspection system

    JP2021125052A

  • Reinforcing bar determination device and reinforcing bar determination method

    WO2021024499A1