Strain gauge pasting and polishing area measuring method based on 5G machine vision

Abstract

The invention belongs to the field of airplane structure strength test, and particularly relates to a strain gauge pasting and polishing area measuring method based on 5G machine vision, which comprises a test structure surface, a handheld terminal, a 5G network system and a cloud platform. To-be-scanned information is provided on the surface of the test structure, the handheld terminal can scan the to-be-scanned information and transmit the to-be-scanned information to the cloud platform through the 5G network system, and a machine vision algorithm is arranged in the cloud platform; according to the first machine vision algorithm, the polishing area can be recognized, recorded and judged; 2, identifying and recording serial number feature image information; thirdly, the grinding area and the feature information are automatically synchronized; before the strain gauge is pasted, the surface of a test structure is scanned through the handheld terminal, then scanned information is transmitted to the cloud platform through the 5G network system, then the pasting position is recognized through the cloud platform by means of a machine vision algorithm, and the recognition result is transmitted to the handheld terminal, so that accurate and efficient pasting of the strain gauge is achieved.

Description

Technical Field

[0001] The present application belongs to the field of aircraft structure strength testing, and in particular, relates to a strain gauge bonding and polishing area measurement method based on 5G machine vision. Background Art

[0002] In the existing full-scale aircraft structure static / fatigue test process, in order to achieve accurate measurement of aircraft structure response, the measurement method of sticking strain gauges on the surface of aircraft structure is generally adopted. Since the surface of the strain gauge substrate is relatively smooth, when sticking it to metal or composite materials, it is necessary to use sandpaper to polish the surface of the structure. On the one hand, the anti-corrosion paint on the metal surface or the protective layer on the surface of the composite material is polished off, and the strain gauge is pasted to the real structure force transmission path; on the other hand, the contact area is increased, the adhesion is increased, the quality and survival rate of strain gauge pasting are improved, and loosening and falling off are prevented. After the strain gauge is pasted, in order to prevent the strain gauge from being exposed to the air for a long time, the temperature and humidity changes will affect the pasting, and at the same time, in order to prevent the installation of loading points, damage inspection, damage prefabrication and other work during the static / fatigue test from touching or damaging the strain gauge, it is necessary to apply special colloids after the strain gauge is pasted for protection. For special environments such as oil tanks and water tanks, even secondary protection is required. To ensure adhesion, the protective pasting colloid also needs to be pasted on the polished structure surface. In summary, under the dual needs of strain gauge pasting and protective colloid pasting, the surface of the structure needs to be polished.

[0003] Assume the strain gauge substrate size is L*H(mm 2 ), in order to meet the requirements of "firm adhesion and reliable protection", the surface grinding area of ​​the structure is usually M to N times the size of the substrate, and the strain gauge needs to be pasted after confirming that the grinding area meets the requirements. The existing grinding area measurement method is visual and standard card comparison, which relies on the experience of technicians and lacks quantitative description. Therefore, a more accurate strain gauge pasting grinding area measurement method is needed. Summary of the invention

[0004] The purpose of this application is to provide a method for measuring the grinding area of ​​a strain gauge pasting based on 5G machine vision, so as to solve the problem that the existing strain gauge pasting is based on experience and is not accurate enough.

[0005] The technical solution of the present application is: a strain gauge pasting and polishing area measurement method based on 5G machine vision, including a test structure surface, a handheld terminal, a 5G network system, and a cloud platform; the test structure surface provides information to be scanned, and the handheld terminal can scan the information to be scanned and transmit it to the cloud platform through the 5G network system. The cloud platform is equipped with a machine vision algorithm; the machine vision algorithm can firstly identify, record and judge the polishing area; secondly, it can identify and record the numbered feature image information to obtain the feature information; thirdly, the polishing area and the feature information are automatically synchronized; after the machine vision algorithm processes the image, it obtains the pasting position or polishing area data, and transmits it back to the handheld terminal.

[0006] Preferably, among the information to be scanned, one type includes the polishing status of the strain gauge pasting part, which forms a polished image of the structure surface after scanning; and the other type includes the strain gauge number feature information, which forms a QR code technology or label after scanning.

[0007] Preferably, the handheld terminal includes a high-precision camera, an ordinary camera, a 5G module and a network cable interface; the high-precision camera is used to collect structural surface polishing images and send them to the 5G module; the ordinary camera is used to collect numbered feature images and send them to the 5G module; the 5G module is used to upload structural surface polishing images and numbered feature images at high speed, and the network cable interface provides wired communication capabilities.

[0008] Preferably, the handheld terminal further comprises a battery and a charging interface and a touch screen; the battery and the charging interface provide independent power supply and charging functions.

[0009] Preferably, the 5G network system includes antenna modules dispersedly installed on multiple test piece frames, a centrally installed convergence unit Rhub group, a baseband unit BBU and a mobile 5G customer terminal equipment CPE, the antenna module is used for receiving and sending 5G data, the convergence unit Rhub is used for data transmission, and the baseband unit BBU is used for data processing; the mobile 5G customer terminal equipment CPE is used to provide RJ45 network cable to 5G wireless communication capabilities.

[0010] Preferably, the polished area is identified through a neural network, specifically: a certain amount of structural surface polished images are collected, enlarged, rotated, and cropped to obtain an image recognition data set that meets the training quantity requirements, the corresponding neural network is selected, the initial weights and loss function are set, and the weights are corrected through the loss function until the neural network converges to obtain a neural network model that can identify structural surface polished images.

[0011] As a specific implementation, a strain gauge pasting and polishing area measurement system based on 5G machine vision includes the following steps:

[0012] Step 1: Before attaching the strain gauge, arrange the 5G network antenna, convergence unit Rhub and baseband unit BBU, complete network debugging, arrange additional mobile 5G CPE, that is, mobile 5G network terminal, and test the RJ45 network cable access capability;

[0013] Step 2: Prepare one thousand photos of existing strain gauges being pasted and polished, and manually group them; train the machine vision algorithm, and use a handheld terminal to perform recognition tests;

[0014] Step 3: Prepare the strain gauge characteristic QR code and use a handheld terminal to perform an identification test;

[0015] Step 4: Before pasting the strain gauge, select a flat surface near the strain gauge pasting position and paste the characteristic QR code;

[0016] Step 5: Start sanding according to the pasting position;

[0017] Step 6: After polishing to a certain extent, use a handheld terminal to collect images and upload them;

[0018] Step 7: If the 5G network signal can be covered, use the 5G network; if the 5G network signal cannot be covered, use an RJ45 network cable to connect the CPE to complete the image upload;

[0019] Step 8: The machine vision algorithm on the cloud platform performs result recognition and transmits the result back to the handheld terminal;

[0020] Step 9: If the polishing area does not meet the requirements, continue polishing and repeat steps 6 to 8;

[0021] Step 10: If the grinding area meets the requirements, stop grinding, use the handheld terminal to identify the characteristic QR code, synchronize the strain gauge number, type and grinding area, and automatically record them;

[0022] Step 11: Continue pasting the next strain gauge by repeating Step 4 to Step 10.

[0023] The strain gauge pasting and polishing area measurement method based on 5G machine vision of the present application first scans the surface of the test structure through a handheld terminal before pasting the strain gauge, and then transmits the scanned information to the cloud platform through the 5G network system. The cloud platform then uses a machine vision algorithm to identify the pasting position, and transmits the identification result to the handheld terminal, thereby achieving accurate and efficient pasting of the strain gauge. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] In order to more clearly illustrate the technical solution provided by the present application, the following is a brief introduction to the accompanying drawings. Obviously, the accompanying drawings described below are only some embodiments of the present application.

[0025] Figure 1 This is a schematic diagram of the overall structure of this application;

[0026] Figure 2 This is a schematic diagram of the internal components of the handheld terminal of this application;

[0027] Figure 3 This is a schematic diagram of the 5G network composition for this application. DETAILED DESCRIPTION

[0028] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0029] A strain gauge pasting and polishing area measurement system based on 5G machine vision, such as Figure 1 As shown, it includes a test structure surface, a handheld terminal, a 5G network system, and a cloud platform. The test structure surface provides information to be scanned, and the handheld terminal can scan the information to be scanned and transmit it to the cloud platform through the 5G network system. The cloud platform is equipped with a machine vision algorithm, which can process the information to be scanned, obtain the pasting position or polishing area data, and transmit it back to the handheld terminal.

[0030] Among the information to be scanned, one type is the polishing status of the strain gauge pasting area, which forms a polished image of the structure surface after scanning; the other type is the strain gauge number feature information, including the strain gauge number, type, etc., which forms a QR code technology or label after scanning.

[0031] like Figure 2 As shown, preferably, the handheld terminal includes a high-precision camera, a common camera, a 5G module and a network cable interface. The high-precision camera is used to collect the polished image of the structure surface and send it to the 5G module; the common camera is used to collect the number feature image and send it to the 5G module; the 5G module is used to upload the structure surface polished image and the number feature image at high speed, and the network cable interface provides wired communication capabilities to ensure information transmission when the aircraft wing box section, fuel tank and other wireless signals cannot be covered.

[0032] Preferably, the handheld terminal also includes a battery and a charging interface and a touch screen; the battery and the charging interface provide independent power supply and charging functions to ensure that the terminal can be used wirelessly, and the touch screen provides operational capabilities such as information input.

[0033] like Figure 3As shown, preferably, the 5G network system includes antenna modules dispersedly installed on multiple test piece frames, a centrally installed convergence unit Rhub group and a baseband unit BBU. The antenna module is used for receiving and sending 5G data, the convergence unit Rhub is used for data transmission, and the baseband unit BBU is used for data processing. At the same time, in order to ensure the communication needs when the 5G network cannot cover the special structure, an additional mobile 5G customer terminal equipment CPE is configured to provide RJ45 network cable to 5G wireless communication capabilities; the 5G module is used for uploading images and transmitting recognition results back.

[0034] Machine vision services realize three types of functions: one is the identification, recording and judgment of the polishing area; the second is the identification and recording of numbered feature image information to obtain feature information; the third is the automatic synchronization of the polishing area and feature information; among them, the polishing area recognition algorithm requires image training.

[0035] Preferably, the specific method for image training is: collect a certain amount of structural surface polishing images, enlarge, rotate, and crop them to obtain an image recognition data set that meets the training quantity requirements, select the corresponding neural network, set the initial weights and loss function, and modify the weights through the loss function until the neural network converges to obtain a neural network model that can recognize structural surface polishing images.

[0036] Technicians use handheld terminals to capture images, judge the results based on machine vision recognition algorithms, and provide guidance for continued polishing or strain gauge pasting.

[0037] Before pasting the strain gauge, the surface of the test structure is first scanned by a handheld terminal, and then the scanned information is transmitted to the cloud platform through the 5G network system. The cloud platform then uses a machine vision algorithm to identify the pasting position, and the identification result is transmitted to the handheld terminal, thereby achieving accurate and efficient pasting of the strain gauge.

[0038] As a specific implementation, a method for measuring the grinding area of ​​a strain gauge based on 5G machine vision is also included, which specifically includes the following steps:

[0039] Step 1: Before attaching the strain gauge, arrange the 5G network antenna, convergence unit Rhub and baseband unit BBU, complete network debugging, arrange additional mobile 5G CPE, that is, mobile 5G network terminal, and test the RJ45 network cable access capability;

[0040] Step 2: Prepare one thousand photos of existing strain gauges being pasted and polished, and manually group them; train the machine vision algorithm, and use a handheld terminal to perform recognition tests;

[0041] Step 3: Prepare the strain gauge characteristic QR code and use a handheld terminal to perform an identification test;

[0042] Step 4: Before pasting the strain gauge, select a flat surface near the strain gauge pasting position and paste the characteristic QR code;

[0043] Step 5: Start sanding according to the pasting position;

[0044] Step 6: After polishing to a certain extent, use a handheld terminal to collect images and upload them;

[0045] Step 7: If the 5G network signal can be covered, use the 5G network; if the 5G network signal cannot be covered, use an RJ45 network cable to connect the CPE to complete the image upload;

[0046] Step 8: The machine vision algorithm on the cloud platform performs result recognition and transmits the result back to the handheld terminal;

[0047] Step 9: If the polishing area does not meet the requirements, continue polishing and repeat steps 6 to 8;

[0048] Step 10: If the grinding area meets the requirements, stop grinding, use the handheld terminal to identify the characteristic QR code, synchronize the strain gauge number, type and grinding area, and automatically record them;

[0049] Step 11: Continue pasting the next strain gauge by repeating Step 4 to Step 10.

[0050] Finally, it should be noted that: the drawings of the embodiments disclosed in the present invention only involve structures related to the embodiments disclosed in the present invention, and other structures can refer to the general design. In the absence of conflict, the same embodiment and different embodiments of the present invention can be combined with each other;

[0051] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for measuring the grinding area of ​​a strain gauge based on 5G machine vision, characterized in that: It includes the surface of a test structure, a handheld terminal, a 5G network system, and a cloud platform; the surface of the test structure provides information to be scanned, the handheld terminal can scan the information to be scanned, and transmit it to the cloud platform through the 5G network system. The cloud platform is equipped with a machine vision algorithm; the machine vision algorithm can firstly identify, record and judge the polishing area; secondly, it can identify and record the numbered feature image information to obtain the feature information; thirdly, the polishing area and the feature information are automatically synchronized; after processing the image, the machine vision algorithm obtains the pasting position or polishing area data, and transmits it back to the handheld terminal.

2. The method for measuring the grinding area of ​​a strain gauge based on 5G machine vision according to claim 1, characterized in that: The information to be scanned includes, for example, the polishing status of the strain gauge pasting part, which forms a polishing image of the structure surface after scanning; and for example, the number characteristic information of the strain gauge, which forms a QR code technology or label after scanning.

3. The method for measuring the grinding area of ​​a strain gauge based on 5G machine vision according to claim 1, characterized in that: The handheld terminal includes a high-precision camera, an ordinary camera, a 5G module and a network cable interface; the high-precision camera is used to collect structure surface polishing images and send them to the 5G module; the ordinary camera is used to collect numbered feature images and send them to the 5G module; the 5G module is used to upload structure surface polishing images and numbered feature images at high speed, and the network cable interface provides wired communication capabilities.

4. The method for measuring the grinding area of ​​a strain gauge based on 5G machine vision as claimed in claim 3, characterized in that: The handheld terminal also includes a battery and a charging interface and a touch screen; the battery and the charging interface provide independent power supply and charging functions.

5. The method for measuring the grinding area of ​​a strain gauge based on 5G machine vision according to claim 1, characterized in that: The 5G network system includes antenna modules installed dispersedly on multiple test piece frames, a centrally installed convergence unit Rhub group, a baseband unit BBU and a mobile 5G customer terminal device CPE. The antenna modules are used to receive and send 5G data, the convergence unit Rhub is used for data transmission, and the baseband unit BBU is used for data processing; the mobile 5G customer terminal device CPE is used to provide RJ45 network cable to 5G wireless communication capabilities.

6. The method for measuring the grinding area of ​​a strain gauge based on 5G machine vision according to claim 1, characterized in that: The polished area is identified through neural network, specifically: a certain amount of structural surface polished images are collected, and after enlarging, rotating and cropping, an image recognition data set that meets the training quantity requirements is obtained, the corresponding neural network is selected, the initial weights and loss function are set, and the weights are corrected through the loss function until the neural network converges to obtain a neural network model that can identify structural surface polished images.

7. A strain gauge pasting and polishing area measurement system based on 5G machine vision, using the measurement method as described in any one of claims 1-6, characterized in that: The steps include: Step 1: Before attaching the strain gauge, arrange the 5G network antenna, convergence unit Rhub and baseband unit BBU, complete network debugging, arrange additional mobile 5G CPE, that is, mobile 5G network terminal, and test the RJ45 network cable access capability; Step 2: Prepare one thousand photos of existing strain gauges being pasted and polished, and manually group them; train the machine vision algorithm, and use a handheld terminal to perform recognition tests; Step 3: Prepare the strain gauge characteristic QR code and use a handheld terminal to perform an identification test; Step 4: Before pasting the strain gauge, select a flat surface near the strain gauge pasting position and paste the characteristic QR code; Step 5: Start sanding according to the pasting position; Step 6: After polishing to a certain extent, use a handheld terminal to collect images and upload them; Step 7: If the 5G network signal can be covered, use the 5G network; if the 5G network signal cannot be covered, use an RJ45 network cable to connect the CPE to complete the image upload; Step 8: The machine vision algorithm on the cloud platform performs result recognition and transmits the result back to the handheld terminal; Step 9: If the polishing area does not meet the requirements, continue polishing and repeat steps 6 to 8; Step 10: If the grinding area meets the requirements, stop grinding, use the handheld terminal to identify the characteristic QR code, synchronize the strain gauge number, type and grinding area, and automatically record them; Step 11: Continue pasting the next strain gauge by repeating Step 4 to Step 10.

Images