Data acquisition and transmission method and system of ultrasonic detector using AI

Abstract

The application provides a data acquisition and transmission method and system of an ultrasonic detector using AI, comprising the following steps: S1. transmitting measurement data information generated by the ultrasonic detector to a data collector through a USB interface; S2. uploading the data information file to a decoding device by the data collector; S3. automatically operating professional decoding software to decode the data information file by the decoding device using an AI model, and converting the decoded data information file into an HTML chart file; and S4. uploading the HTML chart file to an AI server through a workshop production network and performing data analysis and processing on the HTML chart file. The method and system connect the ultrasonic device and the factory network, build an Internet of Things platform for the ultrasonic device, and enable the measurement data generated by the ultrasonic detector to be automatically converted and transmitted by AI, thereby avoiding quality accidents caused by manual errors and omissions, improving the intelligence and processing efficiency of ultrasonic data sorting work, and realizing automatic and intelligent decoding and transmission.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of data processing, in particular to a data acquisition and transmission method and system of an ultrasonic detector using AI. BACKGROUND

[0002] The ultrasonic non-destructive testing technology is the current mainstream detection method for resistance spot welding nugget quality. The existing ultrasonic measuring instrument has the following shortcomings: ① the system version is too old and does not have a wireless network module; ② the local detection data needs to be decoded by professional software; ③ the professional decoding software needs to be mastered and used after professional training; ④ the professional decoding software has no API and is not open source. Therefore, under the condition that the operating system of the ultrasonic detector is too old and does not have networking function, the measurement data can only be exported from the ultrasonic detector by the staff using a USB flash disk, and then the data format decoding and conversion are performed by the data conversion software of the equipment manufacturer. The local data storage and export of the ultrasonic detector is heavy, and the process quality engineer cannot obtain the welding spot detection results of the day in time. In order to keep up with the pace of digital transformation of enterprises, a data acquisition and transmission method and system suitable for the ultrasonic detector are urgently needed. SUMMARY

[0003] In view of the defects in the prior art, the purpose of the present application is to provide a data acquisition and transmission method and system of an ultrasonic detector using AI, so that the related detection data generated by the ultrasonic detector is automatically decoded, converted and transmitted by AI, avoiding quality accidents caused by manual errors and omissions, improving the intelligence and processing efficiency of the system, and realizing automatic intelligent decoding and transmission.

[0004] To achieve the above technical effects, the present application adopts the following technical solutions:

[0005] According to a first aspect of the present application, a data acquisition and transmission method of an ultrasonic detector using AI is provided, characterized by comprising the following steps:

[0006] S1. transmitting the welding spot measurement data information file generated by the ultrasonic detector to the data collector through the USB interface;

[0007] S2. uploading the data information file to the decoding device through the SSH protocol via the workshop WiFi network by the data collector;

[0008] S3. automatically operating the professional decoding software to decode the data information file by the decoding device using the AI model, and converting the decoded data information file into an HTML chart file; the decoding step mainly includes recognizing and determining various process interfaces and dialog boxes of the professional decoding software by using the component visual positioning engine in the image recognition console, and simultaneously using the OCR module to recognize the interface text elements and real-time detect the program window state of the professional decoding software;

[0009] S4. uploading the HTML chart file to the AI server through the workshop production network and performing data analysis and processing on the HTML chart file;

[0010] The step S3 of decoding the data information file by using the AI model to automatically operate the professional decoding software specifically comprises the following steps:

[0011] S3.1 starting the professional decoding software and checking the state to ensure successful start;

[0012] S3.2 positioning interface elements including buttons, input boxes and drop-down menus based on AI template matching and SIFT feature recognition and simulating user operation to complete interface operation and parameter input of the professional decoding software, thereby generating an HTML chart file; the step of recognizing and positioning interface elements and simulating user operation comprises: (1) recognizing and positioning buttons by using template matching + HSV color space verification, and completing interactive actions by simulating mouse clicking buttons; (2) recognizing and positioning input boxes by using OCR positioning + cursor position calculation, and simulating manual keyboard input to fill in the input boxes; (3) recognizing and positioning drop-down menus by using area scanning + contour analysis, and simulating manual direction key selection parameters;

[0013] The step S4 of uploading the HTML chart file to the AI server through the workshop production network specifically comprises the following steps:

[0014] S4.1 starting a browser and navigating to a login interface of the AI server, recognizing various elements of the login interface including positioning boxes, password boxes and verification code boxes by using an image recognition module component;

[0015] S4.2 for simple text in the verification code box, using perspective transformation + denoising processing + OCR recognition method to fill in the verification code; for distorted text in the verification code box, using a CNN model training method to fill in the verification code;

[0016] S4.3 after successfully logging in to the AI server, positioning a product information box and a file uploading box by using the image recognition module, inputting product information in the product information box and filling in a path of the HTML chart file to be uploaded in the file uploading box, and clicking an upload file button to upload the HTML chart file to the AI server.

[0017] Optionally, the data collector in the step S1 has an independent IP address.

[0018] Optionally, the image recognition module component in the step S4.1 comprises OpenCV template matching, OCR recognition and color space analysis positioning.

[0019] Optionally, in step S4, the AI server receives and analyzes the data through the workshop production network in the SFTP protocol.

[0020] According to a second aspect of the present application, a data acquisition and transmission system of an ultrasonic detector using AI is provided for implementing the above method, comprising a data collector, a decoding device and an AI server; wherein the data collector is used to receive the weld measurement data information file of the ultrasonic detector through the USB interface; the decoding device is used to receive the weld measurement data information file and run the AI model to automatically operate the professional decoding software, and realize file decoding and HTML chart generation and uploading through computer vision technology; and the AI server receives the generated HTML chart file through the workshop production network and analyzes it.

[0021] Compared with the prior art, the present application has the following beneficial effects:

[0022] The method and system provided by the present application connect the ultrasonic equipment and the factory network, build an Internet of Things platform for the ultrasonic equipment, and enable the related to-be-decoded measurement data generated by the ultrasonic detector to be automatically converted and transmitted by AI, thereby avoiding quality accidents caused by manual errors and omissions, improving the intelligence and processing efficiency of ultrasonic data sorting work, and realizing automatic intelligent decoding and transmission. BRIEF DESCRIPTION OF DRAWINGS

[0023] Other features, objects and advantages of the present application will become more apparent from the following detailed description of non-limiting embodiments with reference to the attached drawings:

[0024] Figure 1 A step flowchart of the method described in the first embodiment;

[0025] Figure 2 A technical scheme architecture diagram of the related links in step S3 and step S4 of the method described in the first embodiment. DETAILED DESCRIPTION

[0026] In order to make the purposes, technical schemes and advantages of the embodiments of the present application clearer, the technical schemes of the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all the embodiments. The components of the embodiments of the present application described and shown in the drawings herein can be arranged and designed in various different configurations.

[0027] Therefore, the following detailed description of embodiments of the application provided in the accompanying drawings is not intended to limit the scope of the application as claimed, but merely represents selected embodiments of the application. Based on the embodiments in the application, all other embodiments obtained by those of ordinary skill in the art without creative labor are within the scope of protection of the application.

[0028] It should be noted that similar reference numbers and letters represent similar items in the following drawings, so once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings. In addition, all directional indications (such as up, down, left, right, front, back, bottom, etc.) in the application are only used to explain the relative positional relationship, movement, etc. between components in a certain posture (as shown in the drawings), and if the certain posture changes, the directional indications also change accordingly. Further, the description of "first", "second", etc. in the application is only for description purposes and cannot be understood as indicating or implying the relative importance or implicitly indicating the number of technical features indicated.

[0029] First embodiment

[0030] The present embodiment provides a data acquisition and transmission method for an ultrasonic detector using AI, as shown in the following steps: Figure 1

[0031] S1. The ultrasonic detector scans the weld and transmits the image of the weld, the classification, and the original encrypted measurement data information file to be decoded to the data acquisition device through the USB interface; in this step, the data acquisition device uses an embedded device to connect the ultrasonic device, and is recognized by the ultrasonic device as a USB flash disk;

[0032] S2. The data acquisition device uploads the data information to the decoding device through the SSH protocol via the workshop WiFi network; in this step, the data acquisition device has an independent IP address and can export the device end data to the decoding device in a wireless manner;

[0033] S3. The decoding device uses the AI model locally loaded and trained to automatically operate the professional decoding software to decode the input measurement file (original encrypted data information to be decoded), and converts it into an html chart file after decoding; the decoding process includes identifying and determining various process interfaces and dialog boxes of the professional decoding software using the component visual positioning engine in the image recognition console, and simultaneously using the OCR module to identify interface text elements and real-time detect the program window state of the professional decoding software;

[0034] In this step, UI automation is combined with computer vision in depth to solve the technical solution of black box software operation automation and login verification code problems.

[0035] ​S4. Upload the html chart file to the AI server through the workshop production network and perform data analysis and processing.

[0036] Figure 2 is a technical solution architecture diagram for the two links of "AI model decodes data and converts it into an Html chart file" in step S3 and "uploading the Html chart file to the AI server through the workshop production network" in step S4. It describes the complete automated processing flow from inputting the measurement file to decoding and generating the HTML chart file, and then automatically logging into the server to upload the file using the AI model. Specifically:

[0037] The step S3 of using the AI model to automatically operate the professional decoding software to decode the file specifically includes the following steps:

[0038] S3.1 Start the professional decoding software and check the status to ensure successful startup;

[0039] S3.2 Simulate user operation menu navigation, file dialog input, and function button triggering through software running, including interface element recognition and positioning based on AI template matching and SIFT features, as well as simulating user operation to complete interface operation and parameter input for decoding, thereby generating an HTML chart file; the positioned interface elements include buttons, input boxes, and drop-down menus; the identified and positioned interface elements and simulated user operation include: (1) using template matching + HSV color space verification to identify and position buttons, and completing interactive actions by simulating mouse clicks; (2) using OCR positioning + cursor position calculation to identify and position input boxes, and simulating manual keyboard input to fill in; (3) identifying and positioning drop-down menus through area scanning + contour analysis, and simulating manual directional key selection parameters; after completing interface operation and parameter input for decoding, start the decoding process and output the HTML chart in compressed package format.

[0040] The step S4 of transmitting the html chart file to the AI server through the workshop production network, i.e., automatically uploading the file to the AI server through the browser login WEB interface, specifically includes the following steps:

[0041] S4.1 Start the browser and navigate to the login interface of the AI server, identify various elements of the WEB page including the positioning box, password box, and verification code box through image recognition module components such as OpenCV template matching, OCR recognition, and color space analysis positioning.

[0042] S4.2 For simple text in the verification code box, use perspective transformation + denoising + OCR recognition method to fill in the verification code; for distorted text in the verification code box, use CNN model training method to fill in the verification code.

[0043] S4.3 After successfully logging in the AI server, the image recognition module is used to locate the product information box and the file upload box. The product information is input in the product information box, and the path of the HTML chart file to be uploaded is filled in the file upload box. Clicking the upload file button submits the form to upload the HTML chart file to the AI server for archiving.

[0044] The method provided in the embodiment connects the ultrasonic device and the factory network, builds an Internet of Things platform for the ultrasonic device, and enables the AI to automatically convert and transmit the related measurement data generated by the ultrasonic detector, thereby avoiding quality accidents caused by manual errors and omissions, improving the system intelligence and processing efficiency, and realizing automatic intelligent decoding and transmission.

[0045] Embodiment 2

[0046] The embodiment provides a data acquisition and transmission system of an ultrasonic detector using AI, which comprises a data collector, a decoding device and an AI server. The data collector is used to receive measurement data of the ultrasonic detector through a USB interface. The decoding device is used to receive the measurement data and automatically operate professional decoding software by running an AI model, and realizes file decoding and HTML chart generation through computer vision technology. The AI server receives and analyzes the decoded data through a workshop production network in an SFTP protocol.

[0047] The specific embodiments of the application are described above, and through the above description, relevant personnel can make various changes and modifications without deviating from the technical idea of the application.

Claims

1. A data acquisition and transmission method for an ultrasonic testing instrument utilizing AI, characterized in that, Includes the following steps: S1. Transfer the weld joint measurement data file generated by the ultrasonic testing instrument to the data acquisition unit via USB interface; S2. The data acquisition device uploads the data information file to the decoding device via the workshop WiFi network and the SSH protocol; S3. The decoding device uses an AI model to automatically operate professional decoding software to decode data information files. After decoding, it converts the data information files into HTML chart files. The decoding steps mainly include using the visual positioning engine component in the image recognition console to identify and determine various process interfaces and dialog boxes of the professional decoding software, while using the OCR module to identify interface text elements and detect the program window status of the professional decoding software in real time. S4. Upload the HTML chart file to the AI ​​server via the workshop production network and perform data analysis and processing on it; Step S3, which involves using an AI model to automatically operate professional decoding software to decode the data information file, specifically includes the following steps: S3.1 Launch the professional decoding software and check its status to ensure successful startup; S3.2 Based on AI, template matching and SIFT feature recognition are used to locate interface elements and simulate user operations to complete various interface operations and parameter inputs of professional decoding software, thereby generating HTML chart files; the located interface elements include buttons, input boxes and drop-down menus; the steps of identifying and locating interface elements and simulating user operations include: (1) using template matching + HSV color space verification to identify and locate buttons, and completing interactive actions by simulating mouse clicks on buttons; (2) using OCR positioning + cursor position calculation to identify and locate input boxes, and simulating manual keyboard input to fill input boxes; (3) using area scanning + contour analysis to identify and locate drop-down menus, and simulating manual directional key selection of parameters; Step S4, uploading the HTML chart file to the AI ​​server via the workshop production network, specifically includes the following steps: S4.1 launches the browser and navigates to the AI ​​server's login interface. It uses the image recognition module component to identify various elements on the login interface, including the positioning box, password box, and verification code box. S4.2 For simple text in the CAPTCHA box, perspective transformation + noise reduction + OCR recognition method is used to fill in the CAPTCHA; for distorted text in the CAPTCHA box, CNN model training method is used to fill in the CAPTCHA. After successfully logging into the AI ​​server using S4.3, the image recognition module is used to locate the product information box and the file upload box. Product information is entered in the product information box, and the path to the HTML chart file to be uploaded is entered in the file upload box. The upload file button is then clicked to upload the HTML chart file to the AI ​​server.

2. The data acquisition and transmission method of the ultrasonic testing instrument using AI according to claim 1, characterized in that, The data collector in step S1 has an independent IP address.

3. The data acquisition and transmission method of the ultrasonic testing instrument using AI according to claim 1, characterized in that, The image recognition module components described in step S4.1 include OpenCV template matching, OCR recognition, and color space analysis and localization.

4. The data acquisition and transmission method of the ultrasonic testing instrument using AI according to claim 1, characterized in that, In step S4, the AI ​​server receives and analyzes data via the workshop production network using the SFTP protocol.

5. A data acquisition and transmission system for an ultrasonic testing instrument utilizing AI, used to implement the method described in any one of claims 1 to 4, characterized in that, The system includes a data acquisition unit, a decoding device, and an AI server. The data acquisition unit receives weld joint measurement data files from an ultrasonic testing instrument via a USB interface. The decoding device receives the weld joint measurement data files and runs an AI model to automatically operate professional decoding software, using computer vision technology to decode the files and generate and upload HTML charts. The AI ​​server receives and analyzes the generated HTML chart files through the workshop production network.

Images