A visual inspection method for frame quality of heavy-duty special-purpose vehicles

Through visual inspection methods, the automatic detection of heavy-duty special vehicle frames has been solved, and the detection complexity and safety problems in the existing technology have been achieved, efficient and accurate frame quality assessment and risk identification have been achieved, ensuring the safety and transportation of the vehicle efficiency.

CN118608528BActive Publication Date: 2024-11-05LIANGSHAN ZHONGXING MASCH MFG CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411081887.1
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-08-08
Publication Date
2024-11-05
Estimated Expiration
2044-08-08

AI Technical Summary

Technical Problem

The prior art cannot efficiently and accurately detect the quality of heavy-duty special vehicle frames, cannot guarantee the safety of vehicle driving, and the inspection process is complicated, so the body parts need to be removed, which increases labor and inspection time.

Method used

The visual detection method of frame quality of heavy-duty special vehicle is adopted. The detection images are obtained by scanning key points, feature extraction, defect analysis and marking are carried out, risk analysis and judgment are carried out in combination with historical data, and final detection results are output to realize automated detection and maintenance decisions. .

Benefits of technology

It improves the accuracy and efficiency of frame detection, reduces human operation errors, identify potential safety risks, reduces the possibility of accidents, ensures the safety and transportation efficiency of the vehicle, and reduces maintenance costs.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118608528B_ABST
    Figure CN118608528B_ABST
Patent Text Reader

Abstract

The present invention relates to the technical field of vehicle quality inspection, and in particular to a method for visual inspection of the quality of a frame of a heavy-duty special-purpose vehicle, the method comprising: scanning each key point of a frame of a heavy-duty special-purpose vehicle returned for maintenance to obtain each actual inspection image; extracting features from the actual inspection image to obtain actual inspection information, and sequentially performing defect analysis and defect marking on the actual inspection information; making a primary judgment on each key point according to the defect marking result; numbering and integrating each actual inspection information in a historical numbering order to obtain an actual inspection database; comparing the actual inspection database with a historical inspection database to perform risk analysis and risk marking; making a secondary judgment on each key point according to the risk marking result; associating and marking each key point in combination with the primary judgment result and the secondary judgment result, and outputting the final inspection result of each key point according to the associative marking result.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the technical field of vehicle quality detection, in particular to a method for visually detecting the quality of a frame of a heavy-duty special vehicle. Background Art

[0002] After transporting goods, heavy-duty special-purpose vehicles need to be returned to the site for inspection, especially the frame quality of the heavy-duty special-purpose vehicle. However, when conducting quality inspection on the frame of a heavy-duty special-purpose vehicle, the body and other parts of the heavy-duty special-purpose vehicle need to be disassembled first to realize the quality inspection of the frame, which complicates the inspection process, and the vehicle needs to be reinstalled after the quality inspection, which increases the labor force. In addition, risk assessment cannot be performed in the existing frame quality inspection process, and the safe driving of the heavy-duty special-purpose vehicle cannot be guaranteed.

[0003] Chinese patent application publication number CN106767550A discloses an online inspection system for a heavy truck frame assembly, which is used to perform quality inspection on the frame assembly, including an accompanying bracket group for supporting and transporting the frame assembly; and a wireless optical pen for measuring the flatness, straightness and coaxiality characteristic values ​​of the frame assembly; and a CCD digital camera for collecting image information of the frame assembly; and a data analysis platform for capturing the state of the contact when the wireless optical pen contacts the frame assembly and forming three-dimensional coordinate data of the contact; and a display screen for displaying the three-dimensional coordinates of the contact; and a detection frame arranged at a riveting line finishing station for installing the display screen and the CCD digital camera.

[0004] It can be seen that the current vehicle quality inspection technology cannot achieve efficient and accurate inspection and cannot guarantee the safety of vehicle driving. Summary of the invention

[0005] To this end, the purpose of the present invention is to provide a method for visually inspecting the frame quality of a heavy-duty special-purpose vehicle, so as to overcome the problem that the current vehicle quality inspection technology cannot achieve efficient and accurate inspection and cannot ensure the safety of vehicle driving.

[0006] To achieve the above object, the present invention provides a method for visually inspecting the quality of a frame of a heavy-duty special vehicle, comprising:

[0007] Scan the key points of the frame of the heavy-duty special vehicle that returns for maintenance to obtain actual inspection images;

[0008] Based on any actual detection image, extract features from the actual detection image to obtain actual detection information, and perform defect analysis and defect marking on the actual detection information in sequence;

[0009] According to the defect marking results, the actual single-point quality of each key point and the actual overall quality of the frame of the heavy-duty special vehicle are determined once;

[0010] Numbering and integrating each actual detection information in the order of historical numbers to obtain an actual detection database;

[0011] Comparing the actual detection database with the historical detection database to perform risk analysis and risk marking;

[0012] A secondary determination is made on the actual single-point quality of each key point and the actual overall quality of the frame of the heavy-duty special vehicle according to the risk marking result;

[0013] Combine the first judgment result and the second judgment result to associate and mark each key point, and output the final detection result of each key point according to the association marking result.

[0014] Furthermore, the process of extracting features from any actual detection image to obtain actual detection information includes:

[0015] Identify the actual detection image to determine an actual overexposed area and an actual viewing angle limited area;

[0016] Determining whether the actual overexposed area and the actual viewing angle limited area both meet the feature extraction requirements;

[0017] Based on the fact that both the actual overexposed area and the actual viewing angle limited area do not meet the feature extraction requirement, it is determined that the actual detection image meets the single opening condition;

[0018] Determine the number of items in each actual detection image that meet the single start condition, and determine whether to start a single-point re-scanning mode or an overall re-scanning mode according to the number of items;

[0019] The feature extraction requirement is used to measure the scanning quality of the actual detection image.

[0020] Furthermore, the process of sequentially performing defect analysis and defect marking on the actual detection information includes:

[0021] Conduct defect analysis on any actual detection information to determine whether there is an abnormality, and determine the defect location and defect classification based on the presence of an abnormality;

[0022] assessing actual defect severity based on the defect location and the defect classification;

[0023] Based on the actual defect severity, a corresponding defect level is determined, the defect level includes: primary defect and secondary defect, and the defect is marked according to the defect level.

[0024] Furthermore, the process of performing defect analysis on any actual detection information to determine whether there is an abnormality includes:

[0025] Performing edge analysis, deformation analysis and texture analysis on the actual detection information;

[0026] If any analysis result of the edge analysis, the deformation analysis and the texture analysis is unqualified, it is determined that an abnormality exists;

[0027] Among them, the edge analysis is used to measure the integrity of the key points; the deformation analysis is used to measure the load-bearing stability of the key points; the texture analysis is used to measure the connection stability of the key points;

[0028] Wherein, the defect analysis includes: the edge analysis, the deformation analysis and the texture analysis.

[0029] Further, the process of evaluating the actual defect severity according to the defect location and the defect classification includes:

[0030] Determine the actual number of defects and the actual defect ranges according to the edge analysis results, the deformation analysis results and the texture analysis results, and determine the actual defect type corresponding to each actual defect range;

[0031] Calculating an actual defect score based on each actual defect range, each actual defect type and the actual number of defects, and evaluating the actual defect severity according to the actual defect score;

[0032] Wherein, the defect location includes: the actual defect ranges and the actual number of defects; the defect classification includes: the actual defect types.

[0033] Furthermore, the process of determining the actual single-point quality of each key point and the actual overall quality of the frame of the heavy-duty special vehicle according to the defect marking result includes:

[0034] For any key point, determining the actual single point quality according to the defect level of the key point;

[0035] Based on the defect level being the first-level defect, determining that the actual single-point quality is first-time qualified;

[0036] Based on the defect level being the secondary defect, the actual single point quality is determined to be unqualified once.

[0037] Furthermore, the process of determining the actual single-point quality of each key point and the actual overall quality of the heavy-duty special vehicle frame according to the defect marking result also includes:

[0038] Record the number of key points whose actual single-point quality is the first-time qualified number to obtain the first-time qualified number;

[0039] Record the number of key points where the actual single-point quality is unqualified once, so as to obtain the number of unqualified once;

[0040] Determining whether the actual overall quality is qualified according to the first qualified quantity and the preset first standard quantity;

[0041] Based on the fact that the first qualified quantity is greater than or equal to the first standard quantity, determining that the actual overall quality is unqualified;

[0042] Based on the fact that the number of unqualified items at one time is greater than or equal to 1, it is determined that the actual overall quality is unqualified.

[0043] Furthermore, the process of comparing the actual detection database with the historical detection database to perform risk analysis and risk marking includes:

[0044] For any key point, obtaining the historical defect score of the key point according to the historical detection database;

[0045] Perform risk analysis based on the historical defect score and the actual defect score to determine an actual risk level, where the actual risk level includes: a high risk level and a low risk level;

[0046] The risk marking is performed accordingly based on the actual risk level.

[0047] Furthermore, the process of performing secondary determination on the actual single-point quality of each key point and the actual overall quality of the heavy-duty special vehicle frame according to the risk marking result includes:

[0048] For any key point, determining the actual single point quality according to the actual risk level of the key point;

[0049] Based on the actual risk level being the low risk level, the actual single point quality is determined to be twice qualified, and based on the actual risk level being the high risk level, the actual single point quality is determined to be twice unqualified;

[0050] Record the number of key points whose actual single point quality is twice qualified to obtain the second qualified number, and record the number of key points whose actual single point quality is twice unqualified to obtain the second unqualified number;

[0051] Based on the secondary qualified quantity being greater than or equal to a preset second standard quantity, determining that the actual overall quality is unqualified;

[0052] Based on the fact that the number of secondary failures is greater than or equal to 1, it is determined that the actual overall quality is unqualified.

[0053] Furthermore, the process of combining the primary determination result and the secondary determination result to associate and mark each key point, and outputting the final detection result of each key point according to the association marking result includes:

[0054] For any key point, if the first determination result and the second determination result based on the actual single-point quality of the key point are both qualified, the key point is marked as finally qualified;

[0055] If the number of unqualified items in the first determination result and the second determination result of the actual single-point quality of the key point is equal to 1, then the key point is associated and marked as a type of final unqualified;

[0056] If the number of unqualified items in the first determination result and the second determination result of the actual single-point quality of the key point is equal to 2, then the key point is associated and marked as a second-class final unqualified;

[0057] The final detection result is outputted according to the associated marking result as no maintenance required, immediate maintenance, or immediate replacement.

[0058] Compared with the prior art, the beneficial effects of the present invention are that, through scanning and feature extraction, each key point of the frame can be inspected in detail to ensure that the overall quality of the frame is comprehensively evaluated; through automated feature extraction and defect analysis, errors caused by human operation can be reduced, and the accuracy and efficiency of inspection can be improved; through the integration of historical inspection data and current inspection data, risk analysis and prediction can be performed to provide data support for maintenance decisions; through risk analysis and risk marking, potential safety risks can be identified, and corresponding preventive measures can be taken to reduce the possibility of accidents; by comparing historical data with current data, the changing trend of frame quality can be identified to provide a basis for quality improvement; combining the results of the first and second determinations, each key point of the frame can be associated and marked, and targeted execution strategies can be output to achieve precise maintenance; through automated and data-driven processes, Reduce the cost of inspection and maintenance, and improve overall operational efficiency; through strict quality inspection and risk assessment, it can ensure the safety of heavy-duty special vehicles during transportation and reduce the risk of accidents and damage; help ensure the frame quality of heavy-duty special vehicles, improve transportation safety, and reduce maintenance costs, covering a series of steps from image acquisition, feature extraction to defect analysis, risk assessment, and final judgment to ensure the comprehensiveness and accuracy of inspection; by establishing an actual inspection database and a historical inspection database, and conducting comparative analysis, the inspection results are more objective and can effectively identify potential risks; focus on key points of the frame to avoid missed inspections and misjudgments; not only identify defects, but also conduct risk assessments, which can more effectively guide repair and maintenance work; by establishing a database, you can accumulate inspection data to provide a reference for subsequent maintenance work; it can effectively improve the inspection efficiency and quality of the frames of heavy-duty special vehicles that return to the field for maintenance.

[0059] In particular, by identifying the overexposed areas and view-restricted areas in the actual test images, the impact of these areas on feature extraction can be effectively avoided, thereby improving the accuracy of detection; by setting feature extraction requirements, it can be ensured that the actual test images used meet certain quality standards, thereby ensuring that the extracted feature information is reliable; when the quality of the test image does not meet the requirements, the re-scanning mode is automatically turned on to re-scan the key points to ensure the accuracy and reliability of the test results; by setting a single start condition, the test images that need to be re-scanned can be quickly identified to avoid rescanning all images and improve detection efficiency; according to different situations, different re-scanning modes can be selected to solve problems in a targeted manner and improve detection efficiency; by setting automatic judgment and execution processes, manual intervention can be reduced , reduce the impact of human factors, and improve the objectivity and consistency of test results; by setting feature extraction requirements, it can ensure that the final extracted feature information is more accurate and reliable, providing a better data basis for subsequent analysis and evaluation; by setting strict judgment standards and automated processes, it can effectively improve the quality, efficiency and data quality of detection, and reduce human intervention, and ultimately achieve more accurate and reliable detection results; by identifying the actual overexposed area and the actual viewing angle restriction area, it can ensure the accuracy of feature extraction and reduce misjudgments caused by image quality problems; by judging whether re-scanning is needed, it can optimize the allocation of scanning resources, avoid unnecessary repeated scanning, and improve detection efficiency; by identifying possible problem areas, preventive maintenance measures can be taken to avoid the expansion and deterioration of potential problems.

[0060] In particular, by performing defect analysis and defect marking on actual detection information, the safety status of key points can be effectively evaluated, and corresponding grading can be carried out, so as to achieve timely warning and handling of potential safety hazards; through edge analysis, deformation analysis and texture analysis, the integrity, load-bearing stability and connection stability of key points can be comprehensively evaluated, and potential defects and safety hazards can be effectively identified; based on the defect analysis results, the defects can be classified and evaluated for severity, and graded, so that corresponding measures can be taken in time to prevent accidents and improve safety; by grading defects, different maintenance strategies can be adopted for defects of different levels to avoid excessive or insufficient maintenance and effectively reduce maintenance costs; potential defects can be discovered and repaired in a timely manner to avoid accidents caused by the development of defects It can automatically analyze, evaluate and mark defects, reduce manual operations and improve detection efficiency. It can classify and grade defects to facilitate information management and maintenance, thus improving work efficiency. It can accurately identify the integrity, load-bearing stability and connection stability of key points through edge analysis, deformation analysis and texture analysis, so as to achieve accurate detection and classification of defects. It comprehensively considers defect information from multiple aspects, and can comprehensively evaluate the severity of actual defects, which is helpful to formulate more reasonable maintenance and repair plans. It can quickly determine the location and type of defects, which is helpful to improve the work efficiency of maintenance personnel and reduce maintenance time. It can effectively prevent potential safety accidents through early detection and severity assessment of defects.

[0061] In particular, by directly linking the defect level to the single-point quality judgment, a clear standard is provided for the quality assessment of each key point, ensuring the consistency and reliability of the assessment; by counting the number of qualified and unqualified key points at a time, the overall quality status of the heavy-duty special vehicle frame can be grasped from a macro perspective; by the preset first standard number, a clear basis for judging whether the overall quality of the frame is qualified is provided, reducing the error of subjective judgment; the quality judgment process is standardized, the judgment process is simplified, and the work efficiency of quality control is improved; by judging the single-point quality of key points, potential problem points can be discovered in time, preventive measures or timely intervention can be taken to avoid the expansion of problems; by judging the overall quality of the frame, the safety performance of the heavy-duty special vehicle during use is ensured, and safety accidents caused by quality problems can be reduced; by discovering and solving quality problems early, the maintenance and repair costs caused by quality problems in the later stage can be reduced; by recording the number of qualified and unqualified key points, the safety status of the frame can be easily tracked, and corresponding maintenance measures can be taken in time; by judging the single-point quality, the defective parts can be repaired in a targeted manner to improve maintenance efficiency; the safety and reliability of the heavy-duty special vehicle frame can be effectively improved.

[0062] In particular, by comparing the historical detection database with the actual detection database, it is possible to predict the potential risks of key points and conduct a quantitative assessment of the risks; using historical data to provide a reference for current detection, enhancing the accuracy and comprehensiveness of risk assessment; dividing risks into high-risk levels and low-risk levels helps to take corresponding preventive or emergency measures according to the risk level; by calculating the difference between the actual defect score and the historical defect score, the changing trend of risks can be identified early and timely measures can be taken; according to the risk level, maintenance and inspection plans can be formulated more specifically to optimize resource allocation; through risk analysis and marking, the monitoring and maintenance of high-risk areas can be strengthened in a targeted manner to reduce the possibility of accidents.

[0063] In particular, the secondary determination provides an additional layer of verification for the actual single-point quality of key points, increasing the accuracy and reliability of quality assessment; combining risk levels for quality determination makes quality management pay more attention to the identification and control of potential risks; the secondary determination can dynamically adjust the risk assessment of key points to better adapt to changes in actual conditions; by identifying key points with high risk levels, targeted monitoring and maintenance can be strengthened to improve the overall safety of the heavy-duty special vehicle frame; secondary determination based on risk level helps to identify key points that may require maintenance in advance, thereby achieving preventive maintenance; by counting the number of secondary qualified and unqualified results, maintenance resources can be allocated more effectively, giving priority to high-risk points; the secondary determination can reduce possible misjudgments and missed judgments in the initial determination, ensuring that each key point is correctly evaluated.

[0064] In particular, by combining the results of the first and second determinations, the quality status of key points can be evaluated more accurately, reducing misjudgments and missed judgments; through associated marking, key points are divided into final qualified, first-class final unqualified and second-class final unqualified, providing multi-level quality grading for maintenance decisions; according to the associated marking results, maintenance decisions (no maintenance required, immediate maintenance, immediate replacement) are directly output, providing clear guidance for maintenance work; according to different unqualified levels, maintenance resources can be allocated in a targeted manner to ensure that resources are used most effectively; through accurate test results and reasonable maintenance decisions, unnecessary maintenance and replacement can be avoided, thereby reducing maintenance costs; ensuring that the quality of key points is effectively monitored, and timely maintenance measures are taken for unqualified points, thereby improving the safety of the overall structure; associated marking and final test results provide detailed records for quality management and maintenance history, enhancing traceability. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] Figure 1 It is a flow chart of a method for visually inspecting the quality of a frame of a heavy-duty special vehicle according to an embodiment of the present invention;

[0066] Figure 2It is a flow chart of risk analysis and risk marking in the visual inspection method for frame quality of heavy-duty special-purpose vehicles according to an embodiment of the present invention;

[0067] Figure 3 It is a flow chart of secondary determination in the visual inspection method for the frame quality of a heavy-duty special vehicle according to an embodiment of the present invention;

[0068] Figure 4 The present invention is a flowchart of performing association marking in a method for visually inspecting the quality of a frame of a heavy-duty special-purpose vehicle according to an embodiment of the present invention. DETAILED DESCRIPTION

[0069] In order to make the objects and advantages of the present invention more clearly understood, the present invention is further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0070] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only used to explain the technical principles of the present invention and are not intended to limit the protection scope of the present invention.

[0071] It should be noted that, in the description of the present invention, terms such as "up", "down", "left", "right", "inside" and "outside" indicating directions or positional relationships are based on the directions or positional relationships shown in the drawings. This is merely for the convenience of description and does not indicate or imply that the device or element must have a specific orientation, be constructed and operated in a specific orientation. Therefore, it cannot be understood as a limitation on the present invention.

[0072] See also Figure 1 As shown, the present invention provides a method for visually inspecting the quality of a frame of a heavy-duty special vehicle, comprising:

[0073] Step S100, scanning each key point of the frame of the heavy-duty special vehicle returned for maintenance to obtain each actual detection image;

[0074] Step S200, based on any actual detection image, extracting features from the actual detection image to obtain actual detection information, and sequentially performing defect analysis and defect marking on the actual detection information;

[0075] Step S300, determining the actual single-point quality of each key point and the actual overall quality of the frame of the heavy-duty special vehicle according to the defect marking result;

[0076] Step S400, numbering and integrating each actual detection information according to the historical numbering sequence to obtain an actual detection database;

[0077] Step S500, comparing the actual detection database with the historical detection database to perform risk analysis and risk marking;

[0078] Step S600, performing a secondary determination on the actual single-point quality of each key point and the actual overall quality of the heavy-duty special vehicle frame according to the risk marking result;

[0079] Step S700 , associate and mark each key point based on the first determination result and the second determination result, and output the final detection result of each key point according to the association marking result.

[0080] Specifically, the embodiments of the present invention can perform detailed inspections on each key point of the frame through scanning and feature extraction to ensure that the overall quality of the frame is comprehensively evaluated; through automated feature extraction and defect analysis, errors caused by human operation can be reduced, and the accuracy and efficiency of inspection can be improved; through integration of historical inspection data and current inspection data, risk analysis and prediction can be performed to provide data support for maintenance decisions; through risk analysis and risk marking, potential safety risks can be identified, and corresponding preventive measures can be taken to reduce the possibility of accidents; by comparing historical data with current data, the changing trend of frame quality can be identified to provide a basis for quality improvement; by combining the primary and secondary determination results, each key point of the frame can be associated and marked, and a targeted execution strategy can be output to achieve precise maintenance; through automated and data-driven processes, inspection time can be reduced. and maintenance costs, improving overall operational efficiency; through strict quality inspection and risk assessment, it can ensure the safety of heavy-duty special-purpose vehicles during transportation and reduce the risk of accidents and damage; it helps to ensure the frame quality of heavy-duty special-purpose vehicles, improve transportation safety, and reduce maintenance costs, covering a series of steps from image acquisition, feature extraction to defect analysis, risk assessment, and final judgment to ensure the comprehensiveness and accuracy of inspection; by establishing an actual inspection database and a historical inspection database, and conducting comparative analysis, the inspection results are more objective and can effectively identify potential risks; focus on key points of the frame to avoid missed inspections and misjudgments; not only identify defects, but also conduct risk assessments, which can more effectively guide repair and maintenance work; by establishing a database, it can accumulate inspection data to provide a reference for subsequent maintenance work; it can effectively improve the inspection efficiency and quality of the frames of heavy-duty special-purpose vehicles that return to the field for maintenance.

[0081] Specifically, the process of extracting features from any actual detection image to obtain actual detection information in this embodiment includes:

[0082] Step S211, identifying the actual detection image to determine an actual overexposed area and an actual viewing angle limited area;

[0083] Step S212, determining whether the actual overexposed area and the actual viewing angle limited area both meet the feature extraction requirements;

[0084] Step S213, based on the fact that both the actual overexposed area and the actual viewing angle limited area do not meet the feature extraction requirement, it is determined that the actual detection image meets the single opening condition;

[0085] Step S214, determining the number of items in each actual detection image that meet the single start condition, and judging whether to start a single-point re-scanning mode or an overall re-scanning mode according to the number of items;

[0086] The feature extraction requirement is used to measure the scanning quality of the actual detection image.

[0087] In this embodiment, the actual over-exposed area is an area that cannot be identified in the actual detection image due to the reflection or refraction phenomenon on the vehicle body surface; the actual viewing angle limited area is an area that cannot be identified in the actual detection image due to the structure of the vehicle body that cannot scan the vehicle frame from an ideal viewing angle;

[0088] Calculating a first ratio between the actual overexposed area and the actual overall area corresponding to the actual detection image; the first ratio=actual overexposed area / actual overall area;

[0089] If the first proportion is greater than or equal to 10%, it is determined that the actual overexposed area does not meet the feature extraction requirement;

[0090] If the first proportion is less than 10%, it is determined that the actual overexposed area meets the feature extraction requirement;

[0091] Calculating a second ratio between the actual viewing angle limit area and the actual overall area corresponding to the actual detection image; the second ratio=actual viewing angle limit area / actual overall area;

[0092] If the second proportion is greater than or equal to 10%, it is determined that the actual viewing angle restriction area does not meet the feature extraction requirement;

[0093] If the second proportion is less than 10%, it is determined that the actual viewing angle restriction area meets the feature extraction requirement;

[0094] Based on the fact that both the actual overexposed area and the actual viewing angle limited area do not meet the feature extraction requirement, it is determined that the actual detection image meets the single opening condition;

[0095] Based on the number of items that meet the feature extraction requirement in the actual overexposed area and the actual viewing angle limited area being greater than or equal to 1, it is determined that the actual detection image does not meet the single opening condition;

[0096] Determine the number of items in each actual detection image that meet the single opening condition, and record it as the single matching number;

[0097] Calculate the third ratio by calculating the single matching number and the total number of images of each actually detected image;

[0098] The third proportion = the number of single matches / the total number of images;

[0099] If the third proportion is greater than or equal to 20%, it is determined to start the overall re-scanning mode;

[0100] If the third proportion is less than 20%, it is determined to start the single-point re-scanning mode, and each key point that meets the single start condition is re-scanned according to the single-point re-scanning mode.

[0101] Specifically, the embodiments of the present invention can effectively avoid the influence of overexposed areas and viewing angle restricted areas in the actual detection image, thereby improving the accuracy of detection; by setting feature extraction requirements, it can ensure that the actual detection image used meets certain quality standards, thereby ensuring that the extracted feature information is reliable; when the quality of the detection image does not meet the requirements, the re-scanning mode is automatically turned on to re-scan the key points to ensure the accuracy and reliability of the detection results; by setting a single start condition, the detection image that needs to be re-scanned can be quickly identified to avoid rescanning all images, thereby improving detection efficiency; different re-scanning modes can be selected according to different situations to solve the problem in a targeted manner and improve detection efficiency; by setting automatic judgment and execution processes, the Manual intervention reduces the impact of human factors and improves the objectivity and consistency of test results. By setting feature extraction requirements, it can ensure that the final extracted feature information is more accurate and reliable, providing a better data basis for subsequent analysis and evaluation. By setting strict judgment standards and automated processes, it can effectively improve the quality, efficiency and data quality of detection, reduce manual intervention, and ultimately achieve more accurate and reliable detection results. By identifying the actual overexposed area and the actual viewing angle restriction area, it can ensure the accuracy of feature extraction and reduce misjudgments caused by image quality problems. By judging whether rescanning is needed, the allocation of scanning resources can be optimized, unnecessary repeated scanning can be avoided, and detection efficiency can be improved. By identifying possible problem areas, preventive maintenance measures can be taken to avoid the expansion and deterioration of potential problems.

[0102] Specifically, the process of sequentially performing defect analysis and defect marking on the actual detection information in this embodiment includes:

[0103] Step S221, performing defect analysis on any actual detection information to determine whether there is an abnormality, and determining defect location and defect classification based on the presence of an abnormality;

[0104] Step S222, evaluating the actual defect severity according to the defect location and the defect classification;

[0105] Step S223, determining a corresponding defect level based on the actual defect severity, the defect level including: primary defect and secondary defect, and marking the defect according to the defect level.

[0106] Specifically, the process of performing defect analysis on any actual detection information to determine whether there is an abnormality includes:

[0107] Performing edge analysis, deformation analysis and texture analysis on the actual detection information;

[0108] If any analysis result of the edge analysis, the deformation analysis and the texture analysis is unqualified, it is determined that an abnormality exists;

[0109] Among them, the edge analysis is used to measure the integrity of the key points; the deformation analysis is used to measure the load-bearing stability of the key points; the texture analysis is used to measure the connection stability of the key points;

[0110] Wherein, the defect analysis includes: the edge analysis, the deformation analysis and the texture analysis.

[0111] In this embodiment, when the edge analysis result shows that the actual damage range is greater than or equal to the standard damage range, the edge analysis result is determined to be abnormal; when the deformation analysis result shows that the actual deformation range is greater than or equal to the standard deformation range, the deformation analysis result is determined to be abnormal; when the texture analysis result shows that the actual crack range is greater than or equal to the standard crack range, the texture analysis result is determined to be abnormal.

[0112] Specifically, the process of evaluating the actual defect severity according to the defect location and the defect classification includes:

[0113] Determine the actual number of defects and the actual defect ranges according to the edge analysis results, the deformation analysis results and the texture analysis results, and determine the actual defect type corresponding to each actual defect range;

[0114] Calculating an actual defect score based on each actual defect range, each actual defect type and the actual number of defects, and evaluating the actual defect severity according to the actual defect score;

[0115] Wherein, the defect location includes: the actual defect ranges and the actual number of defects; the defect classification includes: the actual defect types.

[0116] Specifically, the actual number of defects is used to measure the number of anomalies in the integrity, load-bearing stability and connection stability of the key points;

[0117] The actual defect range determined according to the edge analysis result is the actual damage range; the actual defect range determined according to the deformation analysis result is the actual deformation range; the actual defect range determined according to the texture analysis result is the actual crack range;

[0118] The actual defect type corresponding to the actual damage range is: integrity defect;

[0119] The actual defect type corresponding to the actual deformation range is: stability defect;

[0120] The actual defect type corresponding to the actual crack range is: connectivity defect;

[0121] The actual defect score = (actual damage range × integrity defect type compensation parameter + actual deformation range × stability defect type compensation parameter + actual crack range × connectivity defect type compensation parameter) × actual defect number compensation parameter;

[0122] If the actual defect score is greater than or equal to the preset standard defect score, the actual defect severity is determined to be severe and dangerous, and the defect level is determined to be the first-level defect;

[0123] If the actual defect score is less than the standard defect score, the actual defect severity is determined to be risky, and the defect level is determined to be the secondary defect;

[0124] In this embodiment, the integrity defect type compensation parameter is set to 0.6, the stability defect type compensation parameter is set to 0.7; the connectivity defect type compensation parameter is set to 0.7; the actual defect number compensation parameter is proportional to the actual defect number; for example: when the actual defect number is 2, the actual defect number compensation parameter is set to 0.8; the standard defect score is used to measure the degree of defects when the key points can operate safely; the standard defect score is set to 0.3-0.5; and adaptive selection is made according to the quality assessment standards of the heavy-duty special vehicle frame and the service life of the heavy-duty special vehicle.

[0125] Specifically, the embodiments of the present invention can effectively evaluate the safety status of key points and make corresponding grade divisions by performing defect analysis and defect marking on actual detection information, thereby realizing timely early warning and processing of potential safety hazards; through edge analysis, deformation analysis and texture analysis, it can comprehensively evaluate the integrity, load-bearing stability and connection stability of key points, and effectively identify potential defects and safety hazards; based on the defect analysis results, the defects are classified and evaluated for severity, and graded, so that corresponding measures can be taken in time to prevent accidents and improve safety; by grading defects, different maintenance strategies can be adopted for defects of different grades to avoid excessive or insufficient maintenance and effectively reduce maintenance costs; potential defects can be discovered and repaired in a timely manner to avoid accidents caused by defects Development leads to greater failures and reduces maintenance costs; it can automatically complete defect analysis, evaluation and marking, reduce manual operations and improve detection efficiency; by classifying and grading defects, it can more conveniently manage information and maintain maintenance work, thereby improving work efficiency; through edge analysis, deformation analysis and texture analysis, it can accurately identify the integrity, load-bearing stability and connection stability issues of key points, thereby achieving accurate detection and classification of defects; it comprehensively considers defect information from multiple aspects and can comprehensively evaluate the severity of actual defects, which is helpful to formulate more reasonable maintenance and repair plans; it can quickly determine the location and type of defects, which helps to improve the work efficiency of maintenance personnel and reduce maintenance time; through early detection and severity assessment of defects, potential safety accidents can be effectively prevented.

[0126] Specifically, the process of determining the actual single-point quality of each key point and the actual overall quality of the heavy-duty special vehicle frame according to the defect marking results in this embodiment includes:

[0127] Step S311, for any key point, determining the actual single point quality according to the defect level of the key point;

[0128] Step S312, based on the defect level being the first-level defect, determining that the actual single point quality is once qualified;

[0129] Step S313: Based on the defect level being the secondary defect, the actual single point quality is determined to be unqualified once.

[0130] Specifically, the process of determining the actual single-point quality of each key point and the actual overall quality of the heavy-duty special vehicle frame according to the defect marking result also includes:

[0131] Step S321, recording the number of key points whose actual single point quality is the first-time qualified number to obtain the first-time qualified number;

[0132] Step S322, recording the number of key points where the actual single point quality is unqualified once, to obtain the number of unqualified once;

[0133] Step S323, judging whether the actual overall quality is qualified according to the first qualified quantity and the preset first standard quantity;

[0134] Step S324, based on the one-time qualified quantity being greater than or equal to the first standard quantity, determining that the actual overall quality is unqualified;

[0135] Step S325: based on the one-time unqualified number being greater than or equal to 1, determining that the actual overall quality is unqualified.

[0136] Specifically, the embodiment of the present invention directly associates the defect level with the single-point quality judgment, provides a clear standard for the quality assessment of each key point, and ensures the consistency and reliability of the assessment; by counting the number of qualified and unqualified key points at a time, the overall quality status of the heavy-duty special vehicle frame can be grasped from a macro perspective; by the preset first standard number, a clear basis for judging whether the overall quality of the frame is qualified is provided, reducing the error of subjective judgment; the quality judgment process is standardized, the judgment process is simplified, and the work efficiency of quality control is improved; by judging the single-point quality of key points, potential problem points can be discovered in time, preventive measures or timely intervention can be taken to avoid the expansion of problems; by judging the overall quality of the frame, the safety performance of the heavy-duty special vehicle during use is ensured, and safety accidents caused by quality problems are reduced; by discovering and solving quality problems early, the maintenance and repair costs caused by quality problems in the later stage can be reduced; by recording the number of qualified and unqualified key points, the safety status of the frame can be easily tracked, and corresponding maintenance measures can be taken in time; by judging the single-point quality, the defective parts can be repaired in a targeted manner to improve the maintenance efficiency; the safety and reliability of the heavy-duty special vehicle frame can be effectively improved.

[0137] Specifically, see Figure 2 As shown, it is a flow chart of risk analysis and risk marking, wherein the process of comparing the actual detection database with the historical detection database to perform risk analysis and risk marking includes:

[0138] Step S510, for any key point, obtaining a historical defect score of the key point according to the historical detection database;

[0139] Step S520, performing risk analysis according to the historical defect score and the actual defect score to determine an actual risk level, wherein the actual risk level includes: a high risk level and a low risk level;

[0140] Step S530: performing corresponding risk marking based on the actual risk level.

[0141] Specifically, a first difference is calculated according to the historical defect score and the actual defect score;

[0142] The first difference = actual defect score - historical defect score;

[0143] If the first difference is greater than or equal to a preset first evaluation value, determining that the actual risk level is the high risk level;

[0144] If the first difference is less than the first evaluation value, determining that the actual risk level is the low risk level;

[0145] In this embodiment, the first evaluation value is used to measure the defect variation range of the key point; the specific setting is also affected by the quality assessment standard of the heavy-duty special vehicle frame and the service life of the heavy-duty special vehicle.

[0146] Specifically, the embodiments of the present invention can predict the potential risks of key points and conduct quantitative risk assessment by comparing the historical detection database with the actual detection database; use historical data to provide a reference for current detection, thereby enhancing the accuracy and comprehensiveness of risk assessment; divide risks into high-risk levels and low-risk levels, which helps to take corresponding preventive or emergency measures according to the risk level; by calculating the difference between the actual defect score and the historical defect score, the changing trend of risks can be identified early and measures can be taken in time; according to the risk level, maintenance and inspection plans can be formulated more specifically to optimize resource allocation; through risk analysis and marking, the monitoring and maintenance of high-risk areas can be strengthened in a targeted manner to reduce the possibility of accidents.

[0147] Specifically, see Figure 3 As shown, it is a flowchart of secondary determination, wherein the process of secondary determination of the actual single point quality of each key point and the actual overall quality of the heavy-duty special vehicle frame according to the risk marking result includes:

[0148] Step S610, for any key point, determining the actual single point quality according to the actual risk level of the key point;

[0149] Step S620: if the actual risk level is the low risk level, the actual single point quality is determined to be twice qualified; if the actual risk level is the high risk level, the actual single point quality is determined to be twice unqualified;

[0150] Step S630, recording the number of each key point where the actual single point quality is twice qualified to obtain the second qualified number, and recording the number of each key point where the actual single point quality is twice unqualified to obtain the second unqualified number;

[0151] Step S640, based on the secondary qualified number being greater than or equal to a preset second standard number, determining that the actual overall quality is unqualified;

[0152] Step S650: based on the fact that the number of secondary unqualified products is greater than or equal to 1, it is determined that the actual overall quality is unqualified.

[0153] Specifically, the embodiments of the present invention provide an additional verification layer for the actual single-point quality of key points through secondary determination, thereby increasing the accuracy and reliability of quality assessment; quality determination is performed in combination with risk levels, so that quality management pays more attention to the identification and control of potential risks; through secondary determination, the risk assessment of key points can be dynamically adjusted to better adapt to changes in actual conditions; by identifying key points with high risk levels, monitoring and maintenance can be strengthened in a targeted manner, thereby improving the overall safety of the heavy-duty special vehicle frame; secondary determination based on risk level helps to identify key points that may need maintenance in advance, thereby achieving preventive maintenance; by counting the number of secondary qualified and unqualified results, maintenance resources can be allocated more effectively, and high-risk points can be processed first; through secondary determination, the possible misjudgment and missed determination in the initial determination can be reduced, ensuring that each key point is correctly evaluated.

[0154] Specifically, see Figure 4 As shown in FIG. 1 , it is a flowchart for performing association marking, wherein the process of combining the primary determination result and the secondary determination result to perform association marking on each key point, and outputting the final detection result of each key point according to the association marking result includes:

[0155] Step S710, for any key point, if the first determination result and the second determination result based on the actual single point quality of the key point are both qualified, then the key point is associated and marked as finally qualified;

[0156] Step S720, if the number of unqualified items in the first determination result and the second determination result of the actual single-point quality of the key point is equal to 1, then the key point is associated and marked as a type of final unqualified;

[0157] Step S730, based on the number of unqualified items in the first determination result and the second determination result of the actual single-point quality of the key point being equal to 2, the key point is associated and marked as a second-class final unqualified;

[0158] Step S740: outputting the final detection result as no maintenance required, immediate maintenance required, or immediate replacement required according to the associated marking result.

[0159] Among them, if the associated marking result is the final qualified, the final detection result is determined to be no maintenance required; if the associated marking result is the final unqualified of the first category, the final detection result is determined to be immediate maintenance; if the associated marking result is the final unqualified of the second category, the final detection result is determined to be immediate replacement.

[0160] Specifically, the embodiment of the present invention can more accurately evaluate the quality status of key points and reduce misjudgments and missed judgments by combining the results of the first judgment and the second judgment; through associated marking, the key points are divided into final qualified, one type of final unqualified and two types of final unqualified, providing multi-level quality grading for maintenance decisions; according to the associated marking results, the maintenance decision (no maintenance required, immediate maintenance, immediate replacement) is directly output, providing clear guidance for maintenance work; according to different unqualified levels, maintenance resources can be allocated in a targeted manner to ensure that resources are used most effectively; through accurate detection results and reasonable maintenance decisions, unnecessary maintenance and replacement can be avoided, thereby reducing maintenance costs; it ensures that the quality of key points is effectively monitored, and maintenance measures are taken in time for unqualified points, thereby improving the safety of the overall structure; the associated marking and final detection results provide detailed records for quality management and maintenance history, and enhance traceability.

[0161] The calculation compensation parameters and the calculation adjustment parameters described in the present invention have two functions: one is to balance the left and right dimensions of the formula, and the other is to adjust the numerical results. No specific assignment is performed in this embodiment. In addition, the calculation formulas in this embodiment are used to intuitively reflect the adjustment relationship between the numerical values, such as positive correlation and negative correlation. Unless otherwise specified, the parameter values ​​that are not specifically limited to numerical values ​​are all positive.

[0162] So far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.

[0163] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A visual inspection method for the quality of a heavy-duty special vehicle frame, characterized in that: include: Scan the key points of the frame of the heavy-duty special vehicle that returns for maintenance to obtain actual inspection images; Based on any actual detection image, extract features from the actual detection image to obtain actual detection information, and perform defect analysis and defect marking on the actual detection information in sequence; According to the defect marking results, the actual single-point quality of each key point and the actual overall quality of the frame of the heavy-duty special vehicle are determined once; Numbering and integrating each actual detection information in the order of historical numbers to obtain an actual detection database; Comparing the actual detection database with the historical detection database to perform risk analysis and risk marking; A secondary determination is made on the actual single-point quality of each key point and the actual overall quality of the frame of the heavy-duty special vehicle according to the risk marking result; Combine the first judgment result and the second judgment result to associate and mark each key point, and output the final detection result of each key point according to the association marking result; The process of combining the primary determination result and the secondary determination result to associate and mark each key point, and outputting the final detection result of each key point according to the association marking result includes: For any key point, if the first determination result and the second determination result based on the actual single-point quality of the key point are both qualified, the key point is marked as finally qualified; If the number of unqualified items in the first determination result and the second determination result of the actual single-point quality of the key point is equal to 1, then the key point is associated and marked as a type of final unqualified; If the number of unqualified items in the first determination result and the second determination result of the actual single-point quality of the key point is equal to 2, then the key point is associated and marked as a second-class final unqualified; The final detection result is outputted according to the associated marking result as no maintenance required, immediate maintenance, or immediate replacement.

2. The visual inspection method for the frame quality of a heavy-duty special vehicle according to claim 1 is characterized in that: The process of extracting features from any actual detection image to obtain actual detection information includes: Identify the actual detection image to determine an actual overexposed area and an actual viewing angle limited area; Determining whether the actual overexposed area and the actual viewing angle limited area both meet the feature extraction requirements; Based on the fact that both the actual overexposed area and the actual viewing angle limited area do not meet the feature extraction requirement, it is determined that the actual detection image meets the single opening condition; Determine the number of items in each actual detection image that meet the single start condition, and determine whether to start a single-point re-scanning mode or an overall re-scanning mode according to the number of items; The feature extraction requirement is used to measure the scanning quality of the actual detection image.

3. The visual inspection method for the frame quality of a heavy-duty special vehicle according to claim 2 is characterized in that: The process of sequentially performing defect analysis and defect marking on the actual detection information includes: Conduct defect analysis on any actual detection information to determine whether there is an abnormality, and determine the defect location and defect classification based on the presence of an abnormality; assessing actual defect severity based on the defect location and the defect classification; Based on the actual defect severity, a corresponding defect level is determined, the defect level includes: primary defect and secondary defect, and the defect is marked according to the defect level.

4. The visual inspection method for the frame quality of a heavy-duty special vehicle according to claim 3 is characterized in that: The process of performing defect analysis on any actual detection information to determine whether there is an abnormality includes: Performing edge analysis, deformation analysis and texture analysis on the actual detection information; If any analysis result of the edge analysis, the deformation analysis and the texture analysis is unqualified, it is determined that an abnormality exists; Among them, the edge analysis is used to measure the integrity of the key points; the deformation analysis is used to measure the load-bearing stability of the key points; the texture analysis is used to measure the connection stability of the key points; Wherein, the defect analysis includes: the edge analysis, the deformation analysis and the texture analysis.

5. The method for visually inspecting the frame quality of a heavy-duty special vehicle according to claim 4, characterized in that: The process of evaluating the actual defect severity according to the defect location and the defect classification comprises: Determine the actual number of defects and the actual defect ranges according to the edge analysis results, the deformation analysis results and the texture analysis results, and determine the actual defect type corresponding to each actual defect range; Calculating an actual defect score based on each actual defect range, each actual defect type and the actual number of defects, and evaluating the actual defect severity according to the actual defect score; Wherein, the defect location includes: the actual defect ranges and the actual number of defects; the defect classification includes: the actual defect types.

6. The method for visually inspecting the frame quality of a heavy-duty special vehicle according to claim 5, characterized in that: The process of determining the actual single-point quality of each key point and the actual overall quality of the heavy-duty special vehicle frame according to the defect marking result includes: For any key point, determining the actual single point quality according to the defect level of the key point; Based on the defect level being the first-level defect, determining that the actual single-point quality is first-time qualified; Based on the defect level being the secondary defect, the actual single point quality is determined to be unqualified once.

7. The visual inspection method for the frame quality of a heavy-duty special vehicle according to claim 6 is characterized in that: The process of determining the actual single-point quality of each key point and the actual overall quality of the heavy-duty special vehicle frame according to the defect marking result also includes: Record the number of key points whose actual single-point quality is the first-time qualified number to obtain the first-time qualified number; Record the number of key points where the actual single-point quality is unqualified once, so as to obtain the number of unqualified once; Determining whether the actual overall quality is qualified according to the first qualified quantity and the preset first standard quantity; Based on the fact that the first qualified quantity is greater than or equal to the first standard quantity, determining that the actual overall quality is unqualified; Based on the fact that the number of unqualified items at one time is greater than or equal to 1, it is determined that the actual overall quality is unqualified.

8. The method for visually inspecting the quality of a heavy-duty special vehicle frame according to claim 7, characterized in that: The process of comparing the actual detection database with the historical detection database to perform risk analysis and risk marking includes: For any key point, obtaining the historical defect score of the key point according to the historical detection database; Perform risk analysis based on the historical defect score and the actual defect score to determine an actual risk level, where the actual risk level includes: a high risk level and a low risk level; The risk marking is performed accordingly based on the actual risk level.

9. The method for visually inspecting the quality of a heavy-duty special vehicle frame according to claim 8, characterized in that: The process of performing secondary determination on the actual single-point quality of each key point and the actual overall quality of the heavy-duty special vehicle frame according to the risk marking result includes: For any key point, determining the actual single point quality according to the actual risk level of the key point; Based on the actual risk level being the low risk level, the actual single point quality is determined to be twice qualified, and based on the actual risk level being the high risk level, the actual single point quality is determined to be twice unqualified; Record the number of key points whose actual single point quality is twice qualified to obtain the second qualified number, and record the number of key points whose actual single point quality is twice unqualified to obtain the second unqualified number; Based on the secondary qualified quantity being greater than or equal to a preset second standard quantity, determining that the actual overall quality is unqualified; Based on the fact that the number of secondary failures is greater than or equal to 1, it is determined that the actual overall quality is unqualified.