Automobile steering wheel appearance defect detection method based on machine vision

Abstract

The invention provides an automobile steering wheel appearance defect detection method based on machine vision, and relates to the technical field of defect detection, and the method comprises the steps: activating an ultrasonic probe to carry out high-frequency sound wave detection on a target steering wheel; curvilinear processing is carried out on the first echo signal time sequence of the first detection point location; judging whether the first curve deviation meets a curve deviation threshold or not; if not, activating the laser transmitter to perform laser scanning detection on the first detection point location; carrying out local three-dimensional modeling on the first detection point location; comparing the first local point cloud model with a target preset point cloud model; and analyzing the first point cloud deviation set to obtain a target appearance defect. According to the method and the device, the technical problem of inaccurate detection of non-dominant defects due to the fact that some defects appear only under specific conditions in the prior art can be solved, the appearance defects of the steering wheel are detected from multiple levels by combining ultrasonic detection, laser scanning and three-dimensional modeling, and the defect detection accuracy is improved.

Description

Technical Field

[0001] This application relates to the technical field of defect detection, and particularly to a method for detecting appearance defects of automobile steering wheels based on machine vision. Background Art

[0002] The automobile steering wheel is an important part of vehicle driving control. It not only requires good maneuverability and comfort but also needs to meet strict quality standards in appearance. Common appearance defects include but are not limited to scratches, cracks, bubbles, color differences, indentations, irregular shapes, etc. The material of the steering wheel (such as leather, plastic, metal) and the lighting conditions during the production process will affect the recognition effect of machine vision. Visual inspection usually relies on appearance images to capture visible defects on the surface of an object through a camera. Generally, it can only detect surface defects, and it is very difficult to identify those inconspicuous defects located inside the material or difficult to appear under light (such as microcracks, bubbles, or deformations caused by internal stress).

[0003] In summary, there is a technical problem in the prior art that due to some defects only appearing under specific conditions, the detection of non - obvious defects is inaccurate. Summary of the Invention

[0004] The purpose of this application is to provide a method for detecting appearance defects of automobile steering wheels based on machine vision, so as to solve the technical problem in the prior art that due to some defects only appearing under specific conditions, the detection of non - obvious defects is inaccurate.

[0005] To achieve the above purpose, this application provides a method for detecting appearance defects of automobile steering wheels based on machine vision. Among them, the method for detecting appearance defects of automobile steering wheels based on machine vision includes: activating an ultrasonic probe to perform high - frequency acoustic wave detection on the target steering wheel and obtaining an echo signal record; performing curve fitting on the first echo signal time series of the first detection point in the echo signal record to obtain a first echo spline curve; determining whether the first curve deviation obtained by comparing the first echo spline curve with a preset spline curve meets the curve deviation threshold; if the first curve deviation does not meet the curve deviation threshold, activating a laser emitter to perform laser scanning detection on the first detection point to obtain a first laser record; performing local three - dimensional modeling on the first detection point of the target steering wheel based on the first laser record to obtain a first local point cloud model; comparing the first local point cloud model with the target preset point cloud model of the target steering wheel to obtain a first point cloud deviation set; analyzing the first point cloud deviation set to obtain the target appearance defects of the target steering wheel.

[0006] Optionally, obtain a preset influence structure diagram, where the preset influence structure diagram includes a preset ultrasonic influence structure diagram; form an ultrasonic influence factor group according to the preset ultrasonic influence structure diagram; perform dynamic monitoring of high-frequency acoustic wave detection based on the ultrasonic influence factor group to obtain an ultrasonic influence factor parameter set; render the ultrasonic influence factor parameter set to the preset ultrasonic influence structure diagram to obtain a dynamic ultrasonic influence structure diagram; calibrate and adjust the first echo signal timing according to the dynamic ultrasonic influence structure diagram.

[0007] Optionally, obtain a historical ultrasonic detection database; extract a first historical detection data group from the historical ultrasonic detection database, where the first historical detection data group includes first historical detection condition information and first historical detection accuracy; perform a correlation analysis on the first historical detection condition information and the first historical detection accuracy to obtain a first correlation result; construct the preset ultrasonic influence structure diagram according to the first correlation result.

[0008] Optionally, the first historical detection condition information includes first historical steering wheel condition information, first historical probe condition information, and first historical environment condition information.

[0009] Optionally, obtain any condition index; match an arbitrary correlation coefficient corresponding to the any condition index in the first correlation result; when the arbitrary correlation coefficient reaches a correlation threshold, retrieve a predetermined rendering strategy to render the any condition index to the preset ultrasonic influence structure diagram.

[0010] Optionally, according to the predetermined rendering strategy, use the any condition index as an influence endpoint and the arbitrary correlation coefficient as an influence length; obtain an arbitrary influence structure of the any condition index according to the influence endpoint and the influence length; add the arbitrary influence structure to an initial influence structure diagram to obtain the preset ultrasonic influence structure diagram.

[0011] Optionally, perform matrix vectorization processing on the dynamic ultrasonic influence structure diagram to obtain a dynamic ultrasonic influence vector; use the dynamic vector eigenvalue of the dynamic ultrasonic influence vector as a weight to calibrate and adjust the first echo signal timing.

[0012] Optionally, obtain a first echo scatter plot of the first echo signal timing; perform polynomial regression fitting processing on the first echo scatter plot to obtain a first echo fitting formula; use the first echo fitting curve of the first echo fitting formula as the first echo spline curve.

[0013] Optionally, extract the preset laser influence structure diagram from the preset influence structure diagram; form a laser influence factor group according to the preset laser influence structure diagram; perform dynamic monitoring of laser scanning detection based on the laser influence factor group to obtain a laser influence factor parameter set; render the laser influence factor parameter set to the preset laser influence structure diagram to obtain a dynamic laser influence structure diagram; calibrate and adjust the first laser record according to the dynamic laser influence structure diagram.

[0014] The technical solutions provided in this application have at least the following technical effects or advantages: Activate the ultrasonic probe to perform high-frequency acoustic wave detection on the target steering wheel and obtain an echo signal record; curve the time series of the first echo signal at the first detection point in the echo signal record to obtain a first echo spline curve; determine whether the first curve deviation obtained by comparing the first echo spline curve with the preset spline curve conforms to the curve deviation threshold; if the first curve deviation does not conform to the curve deviation threshold, activate the laser emitter to perform laser scanning detection on the first detection point to obtain a first laser record; perform local three-dimensional modeling on the first detection point of the target steering wheel based on the first laser record to obtain a first local point cloud model; compare the first local point cloud model with the target preset point cloud model of the target steering wheel to obtain a first point cloud deviation set; analyze the first point cloud deviation set to obtain the target appearance defect of the target steering wheel. That is to say, by using ultrasonic detection to perform high-frequency acoustic wave detection on the steering wheel, curve the echo signal, compare it with the preset spline curve, determine whether there is a deviation, if there is a deviation, activate the laser emitter to perform laser scanning, and perform local three-dimensional modeling, compare it with the target preset point cloud model, analyze the point cloud deviation, and identify the appearance defect, which improves the accuracy of detecting the appearance defect of the automotive steering wheel.

[0015] The above description is only an overview of the technical solutions of this application. In order to be able to understand the technical means of this application more clearly, it can be implemented according to the content of the specification. And in order to make the above and other purposes, features and advantages of this application more obvious and understandable, the following specifically illustrates the embodiments of this application. It should be understood that the content described in this part is not intended to identify the key or important features of the embodiments of this application, nor is it used to limit the scope of this application. Other features of this application will become easily understood through the following description. Description of the Drawings

[0016] To more clearly illustrate the technical solutions in the present application or the prior art, the following will briefly introduce the accompanying drawings required for the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are only exemplary. For those of ordinary skill in the art, without creative efforts, other accompanying drawings can also be obtained based on the provided drawings.

[0017] Figure 1 It is a schematic flowchart of the method for detecting the appearance defects of an automotive steering wheel based on machine vision in the present application; Figure 2 It is a schematic flowchart of calibrating and adjusting the timing of the first echo signal in the method for detecting the appearance defects of an automotive steering wheel based on machine vision in the present application. Detailed implementation manners

[0018] The present application provides a method for detecting the appearance defects of an automotive steering wheel based on machine vision, which solves the technical problem in the prior art that due to some defects only appearing under specific conditions, the detection of non - obvious defects is inaccurate. By using ultrasonic detection to perform high - frequency acoustic wave detection on the steering wheel, curve - processing the echo signal, comparing it with a preset spline curve to judge whether there is a deviation. If there is a deviation, activate the laser emitter for laser scanning and perform local three - dimensional modeling, compare it with the target preset point - cloud model, analyze the point - cloud deviation, and identify the appearance defects, thereby improving the accuracy of detecting the appearance defects of the automotive steering wheel.

[0019] Next, the technical solutions in the present application will be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments of the present application. It should be understood that the present application is not limited by the example embodiments described here. Based on the embodiments of the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present application. Additionally, it should be noted that for the sake of convenience of description, only the parts related to the present application are shown in the accompanying drawings rather than all of them.

[0020] Embodiment, please refer to the attached Figure 1 , the present application provides a method for detecting the appearance defects of an automotive steering wheel based on machine vision. Among them, the method for detecting the appearance defects of an automotive steering wheel based on machine vision specifically includes the following steps: S100: Activate the ultrasonic probe to perform high - frequency acoustic wave detection on the target steering wheel and obtain an echo signal record.

[0021] Specifically, a suitable ultrasonic probe is selected and activated to emit high-frequency acoustic signals into the target steering wheel. Commonly used ultrasonic probe tools such as inspection ultrasonic flaw detectors, like Panametrics 5800, send and receive ultrasonic signals inside the object being detected. Appropriate frequencies and probe types are set according to the material and size of the measurement object to ensure the accuracy and effectiveness of the test. An ultrasonic probe is an instrument that uses ultrasonic technology for detection. It emits high-frequency acoustic waves (usually with frequencies ranging from dozens of kilohertz to several megahertz) into an object and receives the echo signals reflected from the inside or surface of the object. It is usually used to detect the internal structure of materials and can identify defects such as cracks, bubbles, and holes.

[0022] The high-frequency acoustic signals emitted by the ultrasonic probe will penetrate the surface of the steering wheel and generate different reflected echoes according to different internal structures or defects (such as cracks, bubbles, deformations, etc.). For example, assume the target steering wheel is made of plastic. If there is a tiny crack in it, the presence of the crack will cause the ultrasonic waves to reflect at the crack and return to the probe. If the crack is small enough, the intensity of the echo signal will be weaker than that of the echo signal in the normal area, and the reflection time will also be different. By analyzing the changes in the echo signals, the position and size of the crack can be determined.

[0023] After the emission and reflection of the ultrasonic signals are completed, the echo signals are automatically recorded. Through ultrasonic detection, defects inside or on the surface of the steering wheel are detected, especially latent defects such as tiny cracks, bubbles, and material inhomogeneity. By analyzing the timing sequence and waveform changes recorded in the echo signals, internal defects can be effectively distinguished from normal areas, providing important data support for subsequent detection work (such as laser scanning and 3D modeling), and improving the accuracy and efficiency of overall defect detection.

[0024] S200: Curve the timing sequence of the first echo signal at the first detection point in the recorded echo signals to obtain the first echo spline curve.

[0025] Specifically, the recorded echo signals are the acoustic signals reflected from the surface of the target object (such as the steering wheel) back to the probe after high-frequency acoustic wave detection of the target steering wheel. They contain the acoustic waves emitted by the probe and the echo data reflected from the target object, usually existing in the form of a time series, recording the change of echo intensity over time. The first detection point refers to a specific position point that the ultrasonic probe focuses on during the detection process. Different detection points can represent different positions on the surface of the target object, which may involve different surface characteristics and defects. The timing sequence of the first echo signal related to the first detection point is obtained from the ultrasonic detection.

[0026] Perform curve fitting on the timing sequence of the first echo signal, process the data in the echo signal timing sequence to make it a smooth curve. Before curve fitting, it is first necessary to calibrate and adjust the timing sequence of the first echo signal. Construct a preset ultrasonic influence structure diagram from the historical detection conditions and their corresponding detection accuracies in the historical ultrasonic detection database, and then form an ultrasonic influence factor group according to the preset ultrasonic influence structure diagram, which includes all factors affecting the ultrasonic detection results (such as temperature, humidity, probe frequency, steering wheel material, etc.). Monitor the process of high-frequency acoustic wave detection according to the ultrasonic influence factor group and collect the parameter value sets of all influence factors. Map these real-time monitoring parameters onto the preset ultrasonic influence structure diagram to reflect the influence of the real-time changes of each influence factor on the ultrasonic detection result during the actual detection process. Calibrate and adjust the timing sequence of the first echo signal according to the preset ultrasonic influence structure diagram mapped with real-time monitoring parameters to ensure the accuracy of the signal, reduce interference or errors, and thus improve the detection accuracy.

[0027] By selecting a suitable mathematical model (such as polynomial regression, spline curve) to fit the discrete data points, connect the discrete echo signal data points into a smooth curve. Usually, the fitting method is used to smooth the data. Draw a scatter plot based on the timing sequence of the first echo signal, and each data point appears as a scatter point on the chart. Perform polynomial regression fitting on the timing sequence data of the first echo signal, use a polynomial model to fit the data points in the scatter plot, and select a suitable polynomial to fit the signal according to the order of regression. Select the regression order (such as quadratic polynomial, cubic polynomial, etc.). If the order is too low, it may not accurately capture the changes in the signal; if the order is too high, it may lead to overfitting. Calculate the echo signal values at different time points through the regression formula and draw the echo fitting curve to obtain a smooth echo signal trend line.

[0028] Further perform spline curve processing on the echo fitting curve, select an appropriate spline type (such as cubic spline), and fit the data through piecewise polynomials to ensure continuity and smoothness at each node. Spline curves can effectively avoid the overfitting problem of polynomial regression, especially when there are large fluctuations between data points. Through polynomial regression and spline curve fitting, the echo signal is smoothed, removing noise and irregular fluctuations, making the trend of the signal more clear. Through the regression and spline curve fitting methods, the interference of noise on signal analysis can be effectively reduced.

[0029] S300: Judge whether the first curve deviation obtained by comparing the first echo spline curve with the preset spline curve meets the curve deviation threshold.

[0030] Specifically, the preset spline curve is a curve pre-constructed based on known normal conditions or reference data, representing the trend or characteristics that the echo signal should exhibit under normal circumstances, and is generated based on theoretical models, historical data, or detection results under ideal conditions. Calculate the deviation between the first echo spline curve and the preset spline curve. The deviation can be calculated by comparing the echo signal intensity values at each time point to calculate the difference between the two curves, or by calculating the average deviation, maximum deviation, or root mean square error (RMSE) over a period of time to evaluate the overall deviation. Compare the calculated first curve deviation with the curve deviation threshold. The curve deviation threshold is a preset tolerance threshold used to determine whether the deviation between the two curves is within an acceptable range. If the deviation meets the threshold, it indicates that the performance of the echo signal meets the expectations, and subsequent detection and analysis can continue. If the deviation does not meet the threshold, it means that the echo signal is abnormal and may have defects (such as surface cracks, pores, etc.), and further detection or remedial measures need to be taken, such as using more precise methods like laser scanning and 3D modeling to locate the defects. By comparing the curve deviations, potential defects can be identified at the moment when the signal changes significantly, reducing the subjectivity of manual judgment, making the detection process more automated, and being less susceptible to the limitations of the operator's experience.

[0031] S400: If the first curve deviation does not meet the curve deviation threshold, activate the laser emitter to perform laser scanning detection on the first detection point to obtain a first laser record.

[0032] Specifically, if the first curve deviation does not meet the curve deviation threshold, it indicates that the echo signal is abnormal. Automatically activate the laser emitter to perform laser scanning detection on the first detection point to accurately measure the minute changes and defects on the surface of the target steering wheel (or other object surfaces). A laser emitter is a device used to emit laser beams and is commonly used for non-contact surface measurement and scanning. In defect detection, lasers can be used to accurately measure the shape, displacement, or defects of the target surface. The laser emitter emits a laser beam towards the surface of the target steering wheel. The laser beam irradiates the target surface and reflects back. The laser receiver captures the reflected optical signal and calculates the three-dimensional coordinates of each point on the surface. The laser scanning results are recorded to generate a first laser record, including the three-dimensional coordinates of the surface points, surface morphology, depth information, etc. Laser scanning can provide very precise three-dimensional surface data and can detect minute defects (such as surface micro-cracks, depressions, or irregular surface deformations) that cannot be detected by ultrasonic echo signals. Ultrasonic waves are suitable for detecting deep or larger defects, while laser scanners are used to accurately detect surface defects. By combining the two, the comprehensiveness and accuracy of defect detection can be effectively improved.

[0033] S500: Based on the first laser record, perform local three-dimensional modeling on the first detection point of the target steering wheel to obtain a first local point cloud model.

[0034] Specifically, after laser scanning, calibrate and adjust the first laser record to eliminate measurement errors caused by factors such as environmental impact, surface reflectivity differences, and laser emission angles. After calibration, a more accurate first laser record is obtained, which contains the corrected data of the target surface obtained by laser scanning. Based on the first laser record, perform local three-dimensional modeling on the first detection point of the target steering wheel. The first detection point is a specific position during the laser scanning process, corresponding to the data of a certain area on the steering wheel surface, which can be the center point, edge point, or an area with a special shape.

[0035] Extract the point cloud data around the first detection point from the first laser record, and use three-dimensional modeling technology to process these points to generate a three-dimensional model. By connecting adjacent points in the point cloud, generate a triangular mesh, thereby constructing the surface model of the target area. Use algorithms (such as the least squares method) to fit the point cloud data to generate a smooth surface or geometric shape. Convert the point cloud data into a uniform voxel grid model for representing objects in three-dimensional space. Local three-dimensional modeling refers to constructing a three-dimensional geometric model of a local area of the target object based on the point cloud data obtained by laser scanning. Usually, the three-dimensional geometric model is a point cloud composed of several points, and these points are connected or fitted through certain algorithms to form a three-dimensional surface.

[0036] After local three-dimensional modeling, a first local point cloud model is obtained, which is composed of a large amount of point cloud data. Each point has a three-dimensional coordinate indicating its position in space. The point cloud model is composed of a series of discrete three-dimensional points and usually presents the shape of the target object's surface. Further analyze the structure, surface flatness, and possible defects of the target object based on the three-dimensional points. Local three-dimensional modeling enables the accurate reconstruction of the surface shape of the target steering wheel, can truly reflect the geometric characteristics and surface structure of the target object, quickly identify abnormal areas on the steering wheel surface, and perform accurate positioning.

[0037] S600: Compare the first local point cloud model with the target preset point cloud model of the target steering wheel to obtain a first point cloud deviation set.

[0038] Specifically, the first partial point cloud model is compared with the target preset point cloud model of the target steering wheel to identify and quantify the differences between the two. The target preset point cloud model of the target steering wheel is a three-dimensional point cloud model of the target steering wheel preset in advance according to design or quality standards, representing the surface geometry, expected dimensions, shape and other characteristics of the steering wheel in an ideal state, usually a standard model generated by computer-aided design (CAD) tools. The preset point cloud model can be generated through measurement and calculation, and is usually not affected by errors in the actual manufacturing process, representing a perfect steering wheel shape, representing the ideal geometry and surface characteristics of the target steering wheel.

[0039] The corresponding points in the two models are compared through a point cloud processing algorithm, and the distance between each point in the first partial point cloud model and the nearest point in the target preset point cloud model is calculated. The Iterative Closest Point (ICP) algorithm is used to align the two point clouds, repeatedly finding the nearest point of each point in one point cloud in the other point cloud, and then continuously optimizing the transformation parameters between the point clouds until the minimum error is reached. After the point cloud comparison, a first point cloud deviation set is generated, recording the spatial positions and deviation values of all the points with differences. The first point cloud deviation set contains the deviation values between all points, indicating the deviations caused by manufacturing errors, material properties, surface unevenness, etc. in the local area. Each deviation point usually represents the spatial position difference between the actual surface and the ideal model. By obtaining the point cloud deviation set, the defects on the surface of the steering wheel can be quantitatively analyzed to help determine the size, shape and position of the defects.

[0040] S700: Analyze the first point cloud deviation set to obtain the target appearance defects of the target steering wheel.

[0041] Specifically, analyze the first point cloud deviation set to identify the appearance defects existing on the surface of the steering wheel, including defects such as poor roundness, warping, surface unevenness, cracks or breakages. Poor roundness means that there are deviations in the circular part of the steering wheel, which may be manifested as part of it protruding or recessing compared to other parts. By comparing the deviations of the points in the deviation set with the ideal circular model, the roundness error can be calculated. By comparing the point cloud data with the ideal circular model, the point cloud is circularly fitted using a circle fitting algorithm (such as the least squares method), and the fitting error is calculated. If the deviation exceeds a certain threshold, it is considered that there is a defect of poor roundness.

[0042] Warping refers to the unevenness of the surface of the steering wheel, with bending or bulging occurring. Through the analysis of the data in the point cloud deviation set, it can be found that the deviations in certain areas on the surface are relatively large and show a consistent shape change (such as bending or bulging). Using a surface fitting algorithm, such as least squares plane fitting or B-spline surface fitting, the deviation between the actual point cloud and the fitted plane is calculated. If the deviation values show some regular changes (such as gradually increasing from the center outwards), it indicates that there may be a warping phenomenon.

[0043] Surface unevenness refers to local depressions or bulges on the surface of the steering wheel. In some areas where the point cloud deviations are concentrated, the deviation values may be significantly higher than those in other areas, indicating surface irregularities. By analyzing the local curvature or dispersion of the point cloud, check for obvious depressions or bulges in local areas. Local curvature calculation methods can be used to determine whether the surface is flat. Cracks or damages usually manifest as obvious discontinuities on the surface of the steering wheel. Some points in the point cloud data may show obvious missing or morphological mutations. Analyze the sparse or discontinuous areas in the point cloud data to determine whether there are cracks or surface defects. By calculating the sharp changes in the deviation values, identify possible fracture or damage areas.

[0044] According to the analysis results of the first point cloud deviation set, a defect report is generated, recording the defect type, defect location, defect severity, repair suggestions, etc. Through a comprehensive analysis of the point cloud deviation set, defects in multiple dimensions, such as shape, surface quality, dimensions, etc., can be detected, greatly reducing the workload of manual inspection, reducing human errors, and improving the detection efficiency. By using the analysis method of point cloud data, tiny defects that cannot be detected by traditional detection means can be identified, providing higher detection accuracy, which helps to detect subtle appearance problems such as tiny warping, poor roundness, or surface unevenness.

[0045] Further, as shown in the appendix Figure 2 This application S200 includes: S210: Obtain a preset influence structure diagram, where the preset influence structure diagram includes a preset ultrasonic influence structure diagram; S220: Form an ultrasonic influence factor group according to the preset ultrasonic influence structure diagram; S230: Perform dynamic monitoring of high-frequency acoustic wave detection based on the ultrasonic influence factor group to obtain an ultrasonic influence factor parameter set; S240: Render the ultrasonic influence factor parameter set to the preset ultrasonic influence structure diagram to obtain a dynamic ultrasonic influence structure diagram; S250: Calibrate and adjust the first echo signal timing according to the dynamic ultrasonic influence structure diagram.

[0046] Specifically, obtaining a preset influence structure diagram, a pre-set graphical structure, contains various conditions and factors that affect the accuracy of ultrasonic detection. The preset influence structure diagram includes the relationships between various factors, including the preset ultrasonic influence structure diagram, which is a specific type of the preset influence structure diagram, specifically used to describe the influence of various factors (such as the steering wheel, probe, environment, etc.) during the ultrasonic detection process on the detection accuracy.

[0047] Based on the historical ultrasonic detection database, multiple detection conditions and their corresponding multiple detection precisions are determined. Randomly select a data group from them, and extract the corresponding first historical detection condition information and the first historical detection precision. Conduct a correlation analysis on the first historical detection condition information and the first historical detection precision to determine the influence degree of each condition index in the first historical detection condition information on the detection precision, that is, the correlation coefficient. For the first historical detection condition information and the first historical detection precision, construct a corresponding initial influence structure diagram. By analogy, according to multiple detection conditions and their corresponding detection precisions, construct corresponding influence structure diagrams respectively, and finally complete the construction of the preset ultrasonic influence structure diagram, which shows various factors affecting ultrasonic detection and the relationships between them.

[0048] According to the preset ultrasonic influence structure diagram, form an ultrasonic influence factor group, which consists of all conditions affecting the ultrasonic signal detection results, such as physical conditions (such as temperature, humidity), equipment settings (such as probe frequency, probe angle), and target object characteristics (such as steering wheel material, thickness), etc. Conduct dynamic monitoring of high-frequency acoustic wave detection based on the ultrasonic influence factor group, and conduct real-time monitoring and recording of various condition factors (such as probe frequency, steering wheel material, etc.) during the ultrasonic detection process. During the process of high-frequency acoustic wave detection, monitor the changes of various influence factors such as environmental conditions and equipment settings in real time, and record their parameter sets to obtain an ultrasonic influence factor parameter set, including temperature, humidity, probe frequency, acoustic wave reflectivity, etc.

[0049] Render the recorded ultrasonic influence factor parameter set onto the preset ultrasonic influence structure diagram to create a dynamic ultrasonic influence structure diagram. According to the influence factor parameters collected in real time, map the rendered ultrasonic influence factor parameter set to the corresponding positions in the preset influence structure diagram. For example, if the probe frequency changes, then the corresponding part of the frequency in the influence structure diagram will be updated dynamically to show the new influence length or direction. Through rendering, the dynamic ultrasonic influence structure diagram can reflect in real time the current detection conditions and their impacts on ultrasonic signals.

[0050] Calibrate the timing of the first echo signal according to the dynamic ultrasonic influence structure diagram. The timing of the echo signal is affected by various factors, such as ambient temperature, humidity changes, or probe settings. If these conditions change, it may cause the timing of the echo signal to shift, thus affecting the detection accuracy. First, convert the dynamic ultrasonic influence structure diagram into a matrix form, and convert the data in it into a vector form through vectorization processing. According to each dynamic ultrasonic influence vector, extract the dynamic vector eigenvalue, which reflects the weighted effect and importance of the influencing factors, and use it as the weight. Calibrate and adjust the timing of the first echo signal according to the weight, and finally obtain a more accurate echo signal timing. Through the rendering of the dynamic ultrasonic influence structure diagram, determine the real-time influence of each influencing factor on the detection process, and the real-time calibration and adjustment of the timing of the first echo signal reduce the signal offset caused by environmental and equipment changes, ensuring the accuracy and consistency of the echo signal, thereby improving the detection accuracy.

[0051] Furthermore, the present application further includes the following steps: S211: Obtain the historical ultrasonic detection database; S212: Extract the first historical detection data group from the historical ultrasonic detection database, where the first historical detection data group includes the first historical detection condition information and the first historical detection accuracy; S213: Perform a correlation analysis on the first historical detection condition information and the first historical detection accuracy to obtain a first correlation result; S214: Construct the preset ultrasonic influence structure diagram according to the first correlation result.

[0052] S2131: The first historical detection condition information includes the first historical steering wheel condition information, the first historical probe condition information, and the first historical environmental condition information.

[0053] Specifically, obtain the historical ultrasonic detection database through the ultrasonic detection device, the detection records of different batches of steering wheels in the production process, or the records provided by the historical test and detection instruments, which usually include the conditions of each detection (such as probe type, test environment, specific state of the steering wheel) and the corresponding detection accuracy. Randomly select a piece of data from the historical acoustic wave detection database as the first historical detection data group, including the first historical detection condition information and the first historical detection accuracy.

[0054] The first historical detection condition information includes various conditions that affect the accuracy and results of ultrasonic detection, such as the first historical steering wheel condition information, the first historical probe condition information, and the first historical environmental condition information. The first historical steering wheel condition information refers to the information related to the physical characteristics of the steering wheel, such as material, shape, surface condition, etc.; the first historical probe condition information refers to the information related to the ultrasonic probe itself, such as the frequency, model, sensitivity, etc. of the probe; the first historical environmental condition information refers to the conditions of the external environment during the detection process, such as factors like temperature, humidity, noise, and light.

[0055] Perform a correlation analysis on the first historical detection condition information and the first historical detection accuracy, such as Pearson correlation coefficient, regression analysis, etc., to determine the influence of different detection conditions (such as environment, probe, steering wheel characteristics) on the detection accuracy. Taking the Pearson correlation coefficient analysis as an example, calculate the correlation coefficient between each pair of variables, that is, the correlation between each historical detection condition information and the detection accuracy. The result is between -1 and +1. +1 indicates a perfect positive correlation, -1 indicates a perfect negative correlation, and 0 indicates no correlation. According to the specific relationship between each detection condition (steering wheel material, probe frequency, environmental temperature, etc.) and the detection accuracy, as well as the degree of influence of each variable on the detection accuracy, obtain the first correlation result.

[0056] According to each condition index and its corresponding correlation coefficient in the first historical detection condition information, introduce a predetermined rendering strategy. Using the condition index as the influence endpoint and the correlation coefficient as the influence length, obtain the respective influence structures corresponding to each condition index, and add them to the initial influence structure diagram. Finally, complete the construction of the preset ultrasonic influence structure diagram, including the degree of influence of each condition index in the first historical detection condition information on the detection accuracy. By deeply analyzing the influence of different detection conditions on the detection accuracy, clarify which factors are the most critical for ultrasonic detection, so as to make targeted adjustments and optimizations in actual detection, accurately adjust the detection process, and ensure the maximization of the final detection accuracy.

[0057] Furthermore, this application also includes the following steps: S2141: Obtain any condition index; S2142: Match the corresponding arbitrary correlation coefficient of the any condition index in the first correlation result; S2143: When the arbitrary correlation coefficient reaches the correlation threshold, call the predetermined rendering strategy to render the any condition index to the preset ultrasonic influence structure diagram.

[0058] Specifically, determine any conditional indicator to be analyzed, which can be a specific condition or factor that may affect the test result during ultrasonic testing, such as the temperature and humidity of the test environment, the material of the steering wheel, the frequency of the probe, etc. Different conditional indicators have different degrees of influence on the ultrasonic testing accuracy. According to any conditional indicator, match the corresponding correlation coefficient in the first correlation result. When any correlation coefficient reaches the correlation threshold, retrieve the predetermined rendering strategy. The correlation threshold refers to the threshold set in the correlation analysis. When the correlation coefficient of a certain conditional indicator reaches or exceeds this threshold value, it is considered that this condition has a significant impact on the testing accuracy and decides to further analyze or visually display it as an important factor.

[0059] Take any conditional indicator as the influencing endpoint and the corresponding any correlation coefficient as the influencing length. According to the predetermined rendering strategy, form any influencing structure corresponding to the any conditional indicator. The predetermined rendering strategy refers to the process of visualizing the correlation analysis results in the form of graphs, charts, etc. By the correlation coefficients of different conditions, dynamically display which conditions have a greater impact on the testing accuracy and show the relative importance of each factor. The rendering method can adopt forms such as color coding (for example, high correlation is represented by red and low correlation is represented by green), size change, graph or annotation (add annotations in the graph to clearly indicate the importance of certain conditions), etc., and dynamically adjust parameters such as color, size, and position in the graph according to the correlation coefficient of the conditional indicator to highlight the conditions that have a greater impact on the testing accuracy.

[0060] Render each conditional indicator in the first historical test condition information correspondingly to form the corresponding influencing structure diagram, and finally obtain the preset ultrasonic influencing structure diagram, which represents the relationship between various influencing factors in ultrasonic testing and shows the correlation between each conditional indicator and the testing accuracy. Through the preset rendering strategy, automatically identify and display important influencing factors, reduce manual intervention and subjective judgment, and realize a more intelligent testing process.

[0061] Furthermore, this application also includes the following steps: S21431: According to the predetermined rendering strategy, take the any conditional indicator as the influencing endpoint and the any correlation coefficient as the influencing length; S21432: Obtain the any influencing structure of the any conditional indicator according to the influencing endpoint and the influencing length; S21433: Add the any influencing structure to the initial influencing structure diagram to obtain the preset ultrasonic influencing structure diagram.

[0062] Specifically, according to the determined predetermined rendering strategy, any conditional indicator is used as the influencing endpoint, and any correlation coefficient is used as the influencing length. The predetermined rendering strategy refers to the rules and methods for converting detection conditions and their correlation coefficients into a visual form, specifying how to use the influencing endpoint (conditional indicator) and the influencing length (correlation coefficient) to create a graphical representation. The influencing endpoint represents the starting point of a specific condition, and the influencing length is determined according to the correlation coefficient obtained from the first correlation result.

[0063] Based on the obtained influencing endpoint and influencing length, any influencing structure corresponding to any conditional indicator is constructed, whose length is the absolute value of the correlation coefficient, and the direction represents the direction of influence. The generated any influencing structure is added to the initial influence result graph to form a preset ultrasonic influence structure graph. That is to say, for all conditional indicators in the first historical detection condition information, the above calculations are performed, using them as influencing endpoints and their corresponding correlation coefficients as influencing lengths to form an influencing structure.

[0064] The preset ultrasonic influence structure graph is the final graph formed after continuous updating, rendering, and adding, including the influences of all conditions and correlation coefficients, intuitively understanding the influences of various factors on the ultrasonic detection accuracy. The preset ultrasonic influence structure graph includes each condition and its corresponding influencing length, and the positive or negative influence of each condition is displayed, using different colors or different-shaped marks to distinguish the influence intensities of different conditions. Through the influence structure graph, the conditions that have a greater impact on the detection accuracy can be directly judged. The influence structure graph can be continuously updated to maintain the latest detection optimization strategy and can be automatically adjusted according to different detection environments.

[0065] Furthermore, the present application further includes the following steps: S251: Perform matrix vectorization processing on the dynamic ultrasonic influence structure graph to obtain a dynamic ultrasonic influence vector; S252: Use the dynamic vector eigenvalue of the dynamic ultrasonic influence vector as the weight to calibrate and adjust the first echo signal time series.

[0066] Specifically, perform matrix vectorization processing on the dynamic ultrasonic influence structure graph to express the multiple influencing factors and their mutual relationships in the graph in numerical form. Convert the dynamic ultrasonic influence structure graph into a matrix form and convert the data therein into a vector form through vectorization processing. Specifically, each influencing factor (such as probe frequency, temperature, humidity, etc.) and the relationship between influencing factors (such as positive correlation or negative correlation) in the dynamic ultrasonic influence structure graph will be converted into a matrix. For example, assuming the influencing factors include temperature, humidity, and probe frequency, these factors are encoded into a matrix form according to their influencing degrees, where each element in the matrix represents the contribution of a certain factor to the ultrasonic detection result.

[0067] The dynamic ultrasonic influence vector is a vector form obtained by vectorizing the influence factor data in the dynamic ultrasonic influence structure diagram through matrix processing. It includes the degree of influence, direction (positive or negative correlation), and intensity of influence of each factor (such as temperature, humidity, probe frequency, etc.) on the ultrasonic signal. The dimension of the dynamic ultrasonic influence vector usually depends on the number of influence factors.

[0068] The dynamic vector eigenvalues are extracted from the dynamic ultrasonic influence vector, which reflect the properties of the vector and are usually used to represent certain important statistics or the intensity of influence on the signal. Calculating the eigenvalues of a vector usually involves the concepts of eigenvalues and eigenvectors in linear algebra. Usually, a characteristic equation is constructed based on the matrix converted from the dynamic ultrasonic influence structure diagram, and a polynomial equation is solved to obtain the eigenvalues. For multiple found eigenvalues, the eigenvectors are calculated through a system of linear equations. The factors with larger eigenvalues usually have a greater influence on the signal, while the factors with smaller eigenvalues have a smaller influence on the signal. The dynamic vector eigenvalues of the dynamic ultrasonic influence vector are used as weights assigned to each influence factor. Usually, the factors with larger eigenvalues correspond to higher weights, meaning that this factor has a greater influence on the calibration of the echo signal timing. The larger the eigenvalue, the greater the weight, indicating that this factor has a stronger influence on the adjustment of the signal timing.

[0069] According to the weights, the timing of the first echo signal is calibrated and adjusted, and the signal is adjusted according to the weights of each influence factor. Since the environmental conditions or detection settings during the detection process may affect the propagation and reflection of the signal, calibration and adjustment are required to reduce deviations and errors. The timing of the echo signal is adjusted to make it more accurate and meet the actual detection requirements. By performing matrix vectorization processing and eigenvalue analysis on the dynamic ultrasonic influence structure diagram, the influence of each influence factor on the timing of the echo signal is quantified, and then efficient calibration and adjustment are carried out to reduce errors and improve the accuracy of the echo signal.

[0070] Furthermore, the present application further includes the following steps: S260: Obtain the first echo scatter plot of the timing of the first echo signal; S270: Perform polynomial regression fitting on the first echo scatter plot to obtain the first echo fitting formula; S280: Use the first echo fitting curve of the first echo fitting formula as the first echo spline curve.

[0071] Specifically, the first echo signal timing refers to the time series data of the echo signal associated with the first detection point during the ultrasonic detection process, which records important characteristics such as the propagation time and reflection intensity of the sound wave. The data in the echo signal timing are plotted as scatter points on a scatter plot, with the horizontal axis representing time (e.g., milliseconds) and the vertical axis representing the echo intensity (e.g., decibels). Each data point represents the intensity of the echo signal at a certain moment. Based on the time and echo intensity in the dataset, the first echo scatter plot is drawn. The scatter plot can help quickly identify whether there are obvious patterns or trends in the signal. The scatter plot is a method of data visualization used to show the relationship between two variables. The first echo scatter plot plots each signal data point in the first echo signal timing as a scatter point, where the abscissa represents time (or other relevant parameters) and the ordinate represents the intensity or amplitude of the echo signal.

[0072] Perform polynomial regression fitting on the first echo scatter plot, and use the polynomial regression analysis method to fit these scatter points. Polynomial regression is a statistical method of fitting data through a polynomial function, aiming to find an optimal fitting curve to minimize the fitting error. Fit the data through a polynomial function, select an appropriate polynomial order (such as a quadratic or cubic polynomial) to minimize the error between the data points and the fitting curve. The goal of the fitting curve is to find an optimal trend line representing the law of change of the echo signal over time. The selection of the polynomial regression order needs to consider the fitting accuracy and model complexity. If the order is too low, it may not be able to capture the changes in the signal; if the order is too high, it may overfit, resulting in the curve being sensitive to noise. Therefore, cross-validation or other methods are usually used to select the optimal order. The steps of regression fitting include selecting an appropriate polynomial order, using optimization algorithms such as the least squares method to calculate the regression coefficients, and obtaining the fitting curve equation.

[0073] After polynomial regression fitting, the first echo fitting formula is obtained, representing the law of change of the echo signal. Generate the first echo fitting curve according to the first echo fitting formula, representing the change trend of the echo signal over time, and use the first echo fitting curve as the first echo spline curve. The spline curve is a piecewise smooth polynomial function, usually used to fit a smooth curve through discrete data points. Through the spline curve, a finer signal trend can be obtained. Especially when there is noise in the signal or the number of data points is small, the spline curve can smooth the data, reduce the volatility, and obtain a more stable trend prediction. Through polynomial regression and spline curve fitting, the echo signal data can be effectively smoothed, the noise can be eliminated, and a clearer signal change trend can be obtained. The spline curve can better reflect the true change of the signal rather than being disturbed by noise.

[0074] Furthermore, the S400 of the present application further includes: S410: Extract the preset laser influence structure diagram from the preset influence structure diagram; S420: Form a laser influence factor group according to the preset laser influence structure diagram; S430: Perform dynamic monitoring of the laser scanning detection based on the laser influence factor group to obtain a laser influence factor parameter set; S440: Render the laser influence factor parameter set to the preset laser influence structure diagram to obtain a dynamic laser influence structure diagram; S450: Calibrate and adjust the first laser record according to the dynamic laser influence structure diagram.

[0075] Specifically, extract the preset laser influence structure diagram from the preset influence structure diagram. The preset influence structure diagram not only includes the preset ultrasonic influence structure diagram but also the preset laser influence structure diagram. The preset laser influence structure diagram is a structure diagram specifically used to describe and analyze the influencing factors of laser scanning detection, including factors that may affect the laser scanning detection results, such as laser emission angle, distance, surface reflectivity, etc., and the relationships between these factors. Further, under the guidance of the laser influence factor group, dynamically monitor the laser scanning detection process to obtain dynamic data of the factors affecting the laser detection quality, that is, obtain the laser influence factor parameter set. Then, render and update the factor parameters monitored in real time to the preset laser influence structure diagram, and calibrate and adjust the first laser record through the updated dynamic laser influence structure diagram. Correct potential errors caused by environmental, equipment, or external conditions during the laser scanning process.

[0076] According to the preset laser influence structure diagram, identify which factors may affect the laser scanning result, such as the laser emission angle, the reflectivity of the target surface, etc. According to the identified influencing factors, correct the first laser record. This process is logically similar to the process of calibrating and adjusting the first echo signal timing through the preset ultrasonic influence structure diagram. The difference is that one is ultrasonic detection and the other is laser scanning, so it will not be elaborated here. By considering the environmental influence factors of laser scanning, the measurement error can be corrected more precisely. Various interference factors (such as surface reflectivity, distance change, etc.) during the laser scanning process may cause errors in the scanning result. By calibration, the influence of these factors can be reduced, so as to ensure that the measurement data is closer to the real situation.

[0077] In summary, the method for detecting the appearance defects of an automotive steering wheel based on machine vision provided by this application has the following technical effects: The high-frequency acoustic wave detection is performed on the target steering wheel by activating the ultrasonic probe, and the echo signal record is obtained; the time series of the first echo signal at the first detection point in the echo signal record is curve-processed to obtain the first echo spline curve; it is judged whether the first curve deviation obtained by comparing the first echo spline curve with the preset spline curve conforms to the curve deviation threshold; if the first curve deviation does not conform to the curve deviation threshold, the laser emitter is activated to perform laser scanning detection on the first detection point to obtain the first laser record; based on the first laser record, local three-dimensional modeling is performed on the first detection point of the target steering wheel to obtain the first local point cloud model; the first local point cloud model is compared with the target preset point cloud model of the target steering wheel to obtain the first point cloud deviation set; the first point cloud deviation set is analyzed to obtain the target appearance defect of the target steering wheel. That is to say, by performing high-frequency acoustic wave detection on the steering wheel through ultrasonic detection, curve-processing the echo signal, comparing it with the preset spline curve, judging whether there is a deviation, if there is a deviation, activating the laser emitter for laser scanning, performing local three-dimensional modeling, comparing with the target preset point cloud model, analyzing the point cloud deviation, and identifying the appearance defect, the accuracy of the appearance defect detection of the automotive steering wheel is improved.

[0078] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0079] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalent technologies, the present application is also intended to include these changes and modifications.

Claims

1. A method for detecting appearance defects of automobile steering wheels based on machine vision, characterized in that: include: Activate the ultrasonic probe to perform high-frequency sound wave detection on the target steering wheel and obtain echo signal records; Curve processing is performed on the time sequence of the first echo signal of the first detection point in the echo signal record to obtain a first echo spline curve; Determining whether a first curve deviation obtained by comparing the first echo spline curve with a preset spline curve meets a curve deviation threshold; If the first curve deviation does not meet the curve deviation threshold, activating the laser transmitter to perform laser scanning detection on the first detection point to obtain a first laser record; Performing local three-dimensional modeling on the first detection point of the target steering wheel based on the first laser record to obtain a first local point cloud model; Comparing the first local point cloud model with the target preset point cloud model of the target steering wheel to obtain a first point cloud deviation set; The first point cloud deviation set is analyzed to obtain target appearance defects of the target steering wheel.

2. The method for detecting appearance defects of automobile steering wheels based on machine vision according to claim 1, characterized in that: Before the first echo signal time sequence of the first detection point in the echo signal record is processed into a curve to obtain a first echo spline curve, the method further includes: Acquire a preset influence structure diagram, wherein the preset influence structure diagram includes a preset ultrasonic influence structure diagram; Establishing an ultrasonic impact factor group according to the preset ultrasonic impact structure diagram; Dynamically monitoring high-frequency sound wave detection based on the ultrasonic impact factor group to obtain an ultrasonic impact factor parameter set; Rendering the ultrasonic impact factor parameter set to the preset ultrasonic impact structure diagram to obtain a dynamic ultrasonic impact structure diagram; The first echo signal timing is calibrated and adjusted according to the dynamic ultrasonic impact structure diagram.

3. The method for detecting appearance defects of automobile steering wheels based on machine vision according to claim 2, characterized in that: Obtaining a preset impact structure diagram, wherein the preset impact structure diagram includes a preset ultrasonic impact structure diagram, including: Access to historical ultrasonic testing database; Extracting a first historical detection data group from the historical ultrasonic detection database, wherein the first historical detection data group includes first historical detection condition information and first historical detection accuracy; Performing a correlation analysis on the first historical detection condition information and the first historical detection accuracy to obtain a first correlation result; The preset ultrasonic influence structure diagram is constructed according to the first correlation result.

4. The method for detecting appearance defects of automobile steering wheels based on machine vision according to claim 3 is characterized in that: The first historical detection condition information includes first historical steering wheel condition information, first historical probe condition information and first historical environmental condition information.

5. The method for detecting appearance defects of automobile steering wheels based on machine vision according to claim 3 is characterized in that: Constructing the preset ultrasonic impact structure diagram according to the first correlation result includes: Get any conditional index; Matching any correlation coefficient corresponding to the any condition indicator in the first correlation result; When the arbitrary correlation coefficient reaches the correlation threshold, a predetermined rendering strategy is called to render the arbitrary condition index into the preset ultrasonic impact structure diagram.

6. The method for detecting appearance defects of automobile steering wheels based on machine vision according to claim 5, characterized in that: Retrieving a predetermined rendering strategy to render the arbitrary condition index to the preset ultrasonic impact structure diagram includes: According to the predetermined rendering strategy, taking the arbitrary condition index as the impact endpoint and taking the arbitrary correlation coefficient as the impact length; Obtaining any influence structure of the arbitrary condition indicator according to the influence endpoint and the influence length; The arbitrary influencing structure is added to the initial influencing structure diagram to obtain the preset ultrasonic influencing structure diagram.

7. The method for detecting appearance defects of automobile steering wheels based on machine vision according to claim 2, characterized in that: Calibration and adjustment of the first echo signal timing according to the dynamic ultrasonic impact structure diagram includes: Performing matrix vectorization processing on the dynamic ultrasonic influence structure diagram to obtain a dynamic ultrasonic influence vector; The first echo signal timing is calibrated and adjusted using the dynamic vector eigenvalue of the dynamic ultrasonic wave influence vector as a weight.

8. The method for detecting appearance defects of automobile steering wheels based on machine vision according to claim 1, characterized in that: include: Acquire a first echo scatter plot of the first echo signal time series; Performing polynomial regression fitting processing on the first echo scatter plot to obtain a first echo fitting formula; The first echo fitting curve of the first echo fitting formula is used as the first echo spline curve.

9. The method for detecting appearance defects of automobile steering wheels based on machine vision according to claim 2, characterized in that: After obtaining the preset impact structure diagram, it also includes: Extracting a preset laser influence structure diagram from the preset influence structure diagram; Establishing a laser influence factor group according to the preset laser influence structure diagram; Dynamically monitoring laser scanning detection based on the laser influence factor group to obtain a laser influence factor parameter set; Rendering the laser impact factor parameter set to the preset laser impact structure diagram to obtain a dynamic laser impact structure diagram; The first laser record is calibrated and adjusted according to the dynamic laser influence structure diagram.

Images