Bluetooth earphone mainboard surface quality detection method and system

Abstract

The invention discloses a Bluetooth earphone mainboard surface quality detection method and system, and relates to the technical field of quality detection, and the method comprises the following steps: collecting a Bluetooth earphone production line mechanical vibration signal to generate a vibration frequency spectrum matrix, and obtaining an anti-vibration imaging sequence based on the vibration frequency spectrum matrix; based on a DTW algorithm, performing time sequence alignment matching on the anti-vibration imaging sequence to obtain vibration compensation coefficients corresponding to different DTW thresholds, and constructing a plurality of detection strategy combinations; respectively using each detection strategy combination to carry out Bluetooth earphone mainboard image acquisition, synchronously obtaining image clear data and processing delay data, and constructing a vibration interference evaluation model; and performing data analysis on the output of the vibration interference evaluation model to obtain an optimal detection strategy combination, and applying the optimal detection strategy combination to the surface quality detection of the Bluetooth earphone mainboard, thereby solving the problem that the accuracy of the surface quality detection of the Bluetooth earphone mainboard is reduced because the optimal detection strategy cannot be dynamically obtained by a traditional method.

Description

Technical Field

[0001] The present invention relates to the technical field of quality inspection, and more specifically, to a method and system for surface quality inspection of a Bluetooth headset main board. Background Art

[0002] With the rapid development of the Bluetooth headset industry, especially its wide application in the fields of consumer electronics and intelligent devices, the quality control and inspection technology of Bluetooth headsets is also constantly advancing and improving. Since a Bluetooth headset is a highly integrated small consumer product, its quality control and inspection not only need to meet strict functional requirements, but also need to cope with complex production environments and external interferences.

[0003] For example, in the invention patent announcement No. CN118858298B, 202411320132.2, a method, system and device for detecting an earphone charging base first obtains detection feature information, where the detection feature information includes the shooting feature information of a camera, the feature information of the earphone charging base and the light source position coordinates, then obtains the position coordinates of the camera shooting according to the shooting feature information, then obtains the detected position coordinates according to the feature information of the earphone charging base, and finally determines whether the detected position coordinates, the light source position coordinates and the position coordinates of the camera shooting are collinear. If they are collinear, the camera angle is adjusted so that the camera and the light source position coordinates are not on the same straight line, thereby avoiding the reflection on the side of the earphone charging base from affecting the shooting of the camera, and thus the taken photo will not have strong reflection spots due to the reflection on the side of the earphone charging base, and will not seriously affect the problem of defect detection of the earphone charging base.

[0004] In the above disclosed technical solution, at least the following technical problems exist: In the traditional quality inspection of Bluetooth headset main boards, under the interference of mechanical vibration on the production line, relying on empirical parameters, an evaluation model of vibration interference, image quality and image detection delay is not constructed, and the best detection strategy cannot be dynamically obtained, resulting in a reduction in the accuracy of surface quality inspection of Bluetooth headset main boards.

[0005] In view of the above problems, the present invention proposes a solution. Summary of the Invention

[0006] In order to overcome the above-mentioned defects of the prior art, an embodiment of the present invention provides a method and system for surface quality inspection of a Bluetooth headset main board, which analyzes data of multiple detection strategy combinations by constructing a vibration interference evaluation model, and effectively screens out the best detection strategy combination to solve the problem that the traditional method relies on empirical parameters and cannot dynamically obtain the best detection strategy, resulting in a reduction in the accuracy of surface quality inspection of Bluetooth headset main boards.

[0007] To achieve the above object, the present invention provides the following technical solutions: A method for detecting the surface quality of a Bluetooth headset main board includes the following steps: collecting mechanical vibration signals of a Bluetooth headset production line to generate a vibration spectrum matrix, and obtaining an anti-vibration imaging sequence based on the vibration spectrum matrix; based on the DTW algorithm, performing temporal alignment matching on the anti-vibration imaging sequence to obtain vibration compensation coefficients corresponding to different DTW thresholds; constructing several detection strategy combinations based on the vibration compensation coefficients corresponding to different DTW thresholds; respectively applying each detection strategy combination to collect images of the Bluetooth headset main board, and simultaneously obtaining image clarity data and processing delay data to construct a vibration interference evaluation model; performing data analysis on the output of the vibration interference evaluation model to obtain the optimal detection strategy combination and applying it to the surface quality detection of the Bluetooth headset main board.

[0008] In a preferred embodiment, the step of collecting mechanical vibration signals of a Bluetooth headset production line to generate a vibration spectrum matrix specifically includes: collecting mechanical vibration signals of a Bluetooth headset production line, performing Fourier transform on the mechanical vibration signals of the production line to obtain a frequency domain map; based on the frequency domain map, dividing the frequency band according to a preset bandwidth to obtain several first frequency bands; calculating the energy fluctuation variance of each first frequency band and screening several first frequency bands based on a preset first threshold to obtain several screened first frequency bands; extracting features from the several screened first frequency bands to obtain a frequency domain feature vector; arranging the frequency domain feature vectors in a time window sequence to obtain a vibration spectrum matrix.

[0009] In a preferred embodiment, the step of obtaining an anti-vibration imaging sequence based on the vibration spectrum matrix specifically includes: dividing the vibration spectrum matrix according to a time window and calculating the vibration main frequency offset within each time window; constructing a vibration phase angle sequence according to the vibration main frequency offset within each time window; obtaining the anti-vibration imaging sequence by interpolating the vibration phase angle sequence.

[0010] In a preferred embodiment, the step of performing temporal alignment matching on the anti-vibration imaging sequence based on the DTW algorithm to obtain vibration compensation coefficients corresponding to different DTW thresholds specifically includes: constructing a standard anti-vibration imaging sequence library for the surface of a Bluetooth headset main board and dividing the anti-vibration imaging sequence according to a preset time window to obtain several segmented sequences; using the DTW algorithm to calculate the path offset of each segmented sequence and the standard anti-vibration imaging sequence library for the surface of a Bluetooth headset main board respectively to obtain a DTW distance matrix; based on the DTW distance matrix, dividing the DTW distance into several initial DTW thresholds according to a preset ratio; calculating the vibration compensation coefficients corresponding to different DTW thresholds based on a preset vibration compensation coefficient calculation formula.

[0011] In a preferred embodiment, several detection strategy combinations are constructed based on vibration compensation coefficients corresponding to different DTW thresholds, specifically as follows: Obtain the historical mechanical vibration data of the Bluetooth headset production line, and determine the initial detection frame rate and camera exposure time based on the historical mechanical vibration data of the Bluetooth headset production line; randomly increase the initial detection frame rate and exposure time based on preset constraint conditions to obtain several detection frame rates and exposure times; respectively arrange and combine the vibration compensation coefficients corresponding to different DTW thresholds with several detection frame rates and exposure times to obtain several detection strategy combinations.

[0012] In a preferred embodiment, the image clarity data includes an image clarity influence coefficient, and the specific method for obtaining the image clarity influence coefficient is as follows: Continuously collect N frames of images of the surface of the Bluetooth headset main board for each detection strategy combination, and calculate the mean value of the Sobel edge gradient amplitude in the PCB solder joint area of each frame of the image of the surface of the Bluetooth headset main board; use the mean value of the Sobel edge gradient amplitude as the first image clarity value to obtain N first image clarity values; calculate the average value and variance of the N first image clarity values, and use the ratio of the average value to the variance as the image clarity evaluation coefficient; calculate the standard deviation of the vibration angular velocity within the corresponding time window according to the vibration phase angle sequence; input the vibration compensation coefficient, image clarity evaluation coefficient, and vibration angular velocity standard deviation of the current detection strategy combination into a preset image clarity influence coefficient calculation formula to obtain the image clarity influence coefficient.

[0013] In a preferred embodiment, the processing delay data includes a processing delay influence coefficient, and the specific method for obtaining the processing delay influence coefficient is as follows: Obtain the camera parameters, detection frame rate, and exposure time of the current detection camera, and input the camera parameters, detection frame rate, and exposure time of the current detection camera into a preset frame rate delay calculation formula to obtain a frame rate delay factor; obtain the timing data from the acquisition of the Bluetooth headset main board image to the end of quality detection, and the timing data includes image acquisition time, image preprocessing time, and defect recognition time; perform statistical analysis on the timing data, calculate the ratio of the standard deviation to the mean value of the timing data to obtain a delay fluctuation coefficient; input the frame rate delay factor and the delay fluctuation coefficient into a preset processing delay influence coefficient calculation formula to obtain the processing delay influence coefficient.

[0014] In a preferred embodiment, data analysis is performed on the output of the vibration interference evaluation model to obtain the optimal detection strategy combination, which is then applied to the surface quality detection of the Bluetooth headset main board. Specifically: based on the vibration interference evaluation model, the vibration interference evaluation value corresponding to each detection strategy combination is obtained; the multi-dimensional parameters of each detection strategy combination are mapped to a detection strategy combination identification code according to a hash function, and the multi-dimensional parameters include detection frame rate, exposure time, and vibration compensation coefficient; a two-dimensional curve is constructed with the detection strategy combination identification code as the x-axis and the vibration interference evaluation value as the y-axis; data analysis is performed on the two-dimensional curve, and the detection strategy combination identification code corresponding to the lowest point of the two-dimensional curve is used as the optimal detection strategy combination; the optimal detection strategy combination is applied to the surface quality detection of the Bluetooth headset main board.

[0015] Technical effects and advantages of a method and system for detecting the surface quality of a Bluetooth headset main board according to the present invention: 1. By using the DTW algorithm, the present invention calculates the path offset between each segmented sequence and the standard anti-vibration imaging sequence library respectively, and generates a DTW distance matrix. Based on the threshold division and compensation coefficient calculation of the DTW distance matrix, the accuracy of the surface quality detection of the Bluetooth headset main board in different vibration environments can be improved, the interference of vibration on the detection result can be reduced, and the production quality control can be optimized.

[0016] 2. By constructing multiple detection strategy combinations based on the vibration compensation coefficients corresponding to different DTW thresholds, the present invention can adapt to the detection requirements in different vibration environments and improve the anti-interference ability of the system. During the actual detection process, each detection strategy combination is applied to collect images of the Bluetooth headset main board, and the image clarity data and processing delay data are obtained synchronously to ensure the comprehensive evaluation of the image quality and the system response speed under different detection conditions; secondly, by constructing a vibration interference evaluation model, the detection effects of different strategy combinations are quantitatively analyzed, the vibration interference effects of each combination are accurately identified, and the optimal detection strategy combination is determined through data analysis; finally, the optimal detection strategy combination is applied to the surface quality detection of the Bluetooth headset main board, which not only effectively reduces the influence of vibration interference on the detection accuracy, but also improves the stability and reliability of the detection system, thereby improving the production quality of the Bluetooth headset. Brief Description of the Drawings

[0017] Figure 1 It is a schematic flow chart of a method for detecting the surface quality of a Bluetooth headset main board according to the present invention.

[0018] Figure 2 It is a schematic structural diagram of a system for detecting the surface quality of a Bluetooth headset main board according to the present invention. Detailed Embodiments

[0019] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0020] Embodiment 1 Figure 1 A method for detecting the surface quality of a main board of a Bluetooth headset according to the present invention is provided, including the following steps: S1, Collect the mechanical vibration signals of the Bluetooth headset production line to generate a vibration spectrum matrix, and obtain an anti-vibration imaging sequence based on the vibration spectrum matrix; In this example, collecting the mechanical vibration signals of the Bluetooth headset production line to generate a vibration spectrum matrix is specifically as follows; Collect the mechanical vibration signals of the Bluetooth headset production line, and perform Fourier transform on the mechanical vibration signals of the production line to obtain a frequency-domain spectrum; Based on the frequency-domain spectrum, divide the frequency band according to a preset bandwidth to obtain a number of first frequency bands; Calculate the energy fluctuation variance of each first frequency band and based on a preset first threshold, screen a number of first frequency bands to obtain a number of screened first frequency bands; Extract features from a number of screened first frequency bands to obtain a frequency-domain feature vector; Arrange the frequency-domain feature vectors in a time window sequence to obtain a vibration spectrum matrix.

[0021] It should be noted that first, the mechanical vibration signals of the Bluetooth headset production line are collected through a high-precision vibration sensor, and the time-domain signals are converted into a frequency-domain spectrum by using Fourier transform, reducing the randomness influence of the time-domain signals and improving the feasibility and accuracy of data analysis.

[0022] After obtaining the frequency-domain spectrum, the frequency band is divided based on a preset bandwidth, so that the vibration signals can be classified and processed according to different frequency intervals, thereby reducing data redundancy and improving calculation efficiency. Secondly, calculate the energy fluctuation variance of each first frequency band to quantify the fluctuation of the vibration signals in each frequency band, and screen out the frequency bands mainly affected by vibration based on a preset first threshold to ensure the effectiveness and pertinence in the subsequent data analysis process and avoid the interference of irrelevant frequency band signals on the detection results.

[0023] Finally, feature extraction is performed on the filtered first frequency band to obtain a frequency-domain feature vector that can characterize the vibration features. The extracted frequency-domain feature vectors are arranged according to the time window sequence to construct a vibration spectrum matrix, thereby maintaining the integrity of the vibration data in the time series dimension, enabling the detection system to more comprehensively analyze the vibration change trend, and providing high-quality input data for subsequent anti-vibration imaging, vibration compensation, and detection strategy optimization.

[0024] Furthermore, compared with the traditional single-point vibration measurement method, it can characterize the global features of the vibration signal in matrix form, enabling the detection system to more comprehensively adapt to vibration interference under different working conditions. By adopting the methods of frequency-domain analysis and matrix construction, not only is the utilization rate of the vibration signal improved, but also the recognizability of the vibration features is enhanced, ensuring clear and stable detection data can still be obtained in a complex vibration environment, thereby effectively improving the detection accuracy, reducing the problems of false detection and missed detection, and improving the production quality control ability.

[0025] In this example, an anti-vibration imaging sequence is obtained based on the vibration spectrum matrix, specifically: The vibration spectrum matrix is divided according to the time window, and the vibration main frequency offset within each time window is calculated; According to the vibration main frequency offset within each time window, a vibration phase angle sequence is constructed; The anti-vibration imaging sequence is obtained by interpolating the vibration phase angle sequence.

[0026] It should be noted that by constructing an anti-vibration imaging sequence based on the vibration spectrum matrix and using the DTW (Dynamic Time Warping) algorithm to perform time series alignment matching on the anti-vibration imaging sequence, the vibration compensation coefficients corresponding to different DTW thresholds are obtained, improving the accuracy and stability of the surface quality detection of the Bluetooth headset mainboard. In this method, first, the vibration spectrum matrix is divided according to the time window, and the vibration main frequency offset within each time window is calculated, so that the vibration signal is accurately segmented in the time series dimension and the main vibration frequency change of each time window is quantified. This process ensures the pertinence of subsequent anti-vibration processing, enabling the system to effectively identify the dynamic characteristics of the production line vibration.

[0027] Subsequently, a vibration phase angle sequence is constructed according to the vibration main frequency offset of each time window to quantify the phase of the vibration signal. Since the vibration signal often has non-linear and complex fluctuation characteristics in the actual production process of Bluetooth headsets, directly using the original signal for anti-vibration processing will result in large errors. However, by constructing a vibration phase angle sequence, the vibration information is constructed into a mathematical model.

[0028] Furthermore, performing temporal alignment matching on the vibration-resistant imaging sequence based on the DTW algorithm can effectively solve the problem of imaging sequence deformation caused by vibration. By calculating the optimal matching path between sequences, the DTW algorithm can automatically adjust the alignment method for different time windows, making the temporal matching of vibration signals more accurate. By setting different DTW thresholds and calculating the corresponding vibration compensation coefficients, the vibration compensation strategy can be dynamically adjusted in different detection environments to ensure the adaptability and robustness of the detection system.

[0029] S2, Based on the DTW algorithm, perform temporal alignment matching on the vibration-resistant imaging sequence to obtain vibration compensation coefficients corresponding to different DTW thresholds; In this example, based on the DTW algorithm, perform temporal alignment matching on the vibration-resistant imaging sequence to obtain vibration compensation coefficients corresponding to different DTW thresholds, specifically: Construct a vibration-resistant imaging sequence library on the surface of the standard Bluetooth headset motherboard and segment the vibration-resistant imaging sequence according to a preset time window to obtain several segmented sequences; Use the DTW algorithm to calculate the path offset of each segmented sequence from the vibration-resistant imaging sequence library on the surface of the standard Bluetooth headset motherboard respectively to obtain a DTW distance matrix; Based on the DTW distance matrix, divide the DTW distance into several initial DTW thresholds according to a preset ratio; Calculate the vibration compensation coefficients corresponding to different DTW thresholds based on a preset vibration compensation coefficient calculation formula.

[0030] It should be noted that constructing a vibration-resistant imaging sequence library on the surface of the standard Bluetooth headset motherboard, as a reference data set, this sequence library contains standard imaging data obtained under different vibration conditions and can provide stable and highly referential control information. By segmenting the vibration-resistant imaging sequence according to a preset time window, several segmented sequences are obtained, thereby decomposing the complex vibration influence into multiple temporal segments, making the data analysis more refined and facilitating subsequent matching calculations.

[0031] On this basis, use the DTW algorithm to calculate the path offset of each segmented sequence from the vibration-resistant imaging sequence library on the surface of the standard Bluetooth headset motherboard respectively and construct a DTW distance matrix. The DTW algorithm can flexibly adjust the data alignment method on the time axis, eliminate the problems of imaging sequence deformation or time misalignment caused by vibration, and thus ensure the accuracy of comparative analysis. The DTW distance matrix provides the degree of time offset between the vibration-resistant imaging sequence and the standard imaging sequence, enabling the system to quantify the changes in imaging quality under different vibration influences.

[0032] Based on the DTW distance matrix, the DTW distance is divided into several initial DTW thresholds according to a preset ratio, enabling the system to perform hierarchical matching according to different vibration levels. Reasonable threshold division helps optimize the vibration compensation strategy and improve the adaptability of the detection system in different vibration environments. Finally, based on the preset vibration compensation coefficient calculation formula, the vibration compensation coefficients corresponding to different DTW thresholds are calculated, thereby establishing a dynamic compensation mechanism for different vibration situations. The introduction of the vibration compensation coefficient enables the detection system to automatically adjust the imaging parameters to offset the impact of vibration on the imaging quality and ensure the clarity and consistency of the acquired images.

[0033] Among them, the preset vibration compensation coefficient calculation formula is specifically:

[0034] Among them, is the vibration compensation coefficient, is the eigenvalue of the k-th segmented sequence, is the number of segmented sequences, is the eigenvalue of the k-th surface anti-vibration imaging sequence segment of the standard Bluetooth headset mainboard, is the preset maximum DTW distance.

[0035] S3. Construct several detection strategy combinations based on the vibration compensation coefficients corresponding to different DTW thresholds; In this example, several detection strategy combinations are constructed based on the vibration compensation coefficients corresponding to different DTW thresholds, specifically: Obtain the mechanical vibration historical data of the Bluetooth headset production line, and determine the initial detection frame rate and camera exposure time based on the mechanical vibration historical data of the Bluetooth headset production line; Randomly gain the initial detection frame rate and exposure time based on the preset constraint conditions to obtain several detection frame rates and exposure times; Arrange and combine the vibration compensation coefficients corresponding to different DTW thresholds with several detection frame rates and exposure times respectively to obtain several detection strategy combinations.

[0036] It should be noted that by obtaining the mechanical vibration historical data of the Bluetooth headset production line, the system can analyze the vibration characteristics in different time periods of the production environment, and then determine the initial detection frame rate and camera exposure time based on this historical data. The reasonable setting of the initial detection frame rate and exposure time helps to ensure that the imaging system can obtain clear and stable image data under typical vibration conditions, providing basic parameters for subsequent optimization.

[0037] On this basis, the initial detection frame rate and exposure time are randomly increased based on preset constraint conditions to generate multiple different parameter combinations. The introduction of random gain can dynamically adjust the frame rate and exposure time within a certain range to adapt to the changes in different vibration environments, improving the flexibility and robustness of the detection strategy. For the vibration compensation coefficients corresponding to different DTW thresholds, they are respectively combined with multiple detection frame rates and exposure times to generate multiple detection strategy combinations. This combination method can fully consider the influence of different vibration environments on image clarity and processing delay, ensuring that the detection system can select the optimal detection parameters under various working conditions.

[0038] By combining vibration history data, DTW thresholds, and vibration compensation coefficients, the system can dynamically adjust the detection strategy according to different vibration conditions, ensuring the stability and reliability of the detection results. Compared with the traditional fixed parameter setting method, this solution can optimize the detection parameters based on real-time vibration characteristics, thus effectively reducing the impact of vibration interference on imaging quality and improving the accuracy of defect detection. In addition, the establishment of multiple detection strategy combinations enables the system to continuously optimize the detection parameters according to the feedback of the vibration interference evaluation model in practical applications, improving the detection efficiency and reducing the false detection rate.

[0039] S4. Apply each detection strategy combination to collect images of the Bluetooth headset main board, and synchronously obtain image clarity data and processing delay data to construct a vibration interference evaluation model; In this example, the image clarity data includes an image clarity influence coefficient, and the specific method for obtaining the image clarity influence coefficient is as follows: Continuously collect N frames of images of the surface of the Bluetooth headset main board for each detection strategy combination, and calculate the average value of the Sobel edge gradient amplitude in the PCB solder joint area of each frame of the image of the surface of the Bluetooth headset main board; Take the average value of the Sobel edge gradient amplitude as the first image clarity value to obtain N first image clarity values; Calculate the average value and variance of the N first image clarity values, and take the ratio of the average value to the variance as the image clarity evaluation coefficient; Calculate the standard deviation of the vibration angular velocity within the corresponding time window according to the vibration phase angle sequence; Input the vibration compensation coefficient, image clarity evaluation coefficient, and vibration angular velocity standard deviation of the current detection strategy combination into a preset image clarity influence coefficient calculation formula to obtain the image clarity influence coefficient.

[0040] Among them, the preset image clarity influence coefficient calculation formula is specifically:

[0041] Among them, is the image clarity influence coefficient, is the image clarity evaluation coefficient, is the standard deviation of the vibration angular velocity, is the exposure time, is the vibration compensation coefficient of the current detection strategy combination.

[0042] It should be noted that for each detection strategy combination, N frames of images of the surface of the Bluetooth headset main board are continuously collected, and the mean value of the Sobel edge gradient amplitude of each frame of image is calculated in the PCB solder joint area, which is used as the first image clarity value. Since the Sobel edge gradient can effectively measure the edge sharpness of the image, this method can more accurately reflect the clarity of the image.

[0043] After obtaining N first image clarity values, statistical analysis is performed on them, their mean value and variance are calculated, and the ratio of the mean value to the variance is used as the image clarity evaluation coefficient. This process can avoid the influence of the clarity of a single frame of image and comprehensively evaluate the overall contribution of the detection strategy combination to the image clarity. In addition, the vibration factor needs to be combined, and the standard deviation of the vibration angular velocity within the corresponding time window is calculated through the vibration phase angle sequence to quantify the degree of vibration influence. Finally, the vibration compensation coefficient, the image clarity evaluation coefficient, and the standard deviation of the vibration angular velocity of the current detection strategy combination are input into a preset image clarity influence coefficient calculation formula to obtain the image clarity influence coefficient, providing a quantitative basis for the optimization of the detection strategy.

[0044] In this example, the processing delay data includes the processing delay influence coefficient, and the specific method for obtaining the processing delay influence coefficient is as follows: Obtain the camera parameters, detection frame rate, and exposure time of the current detection camera, and input the camera parameters, detection frame rate, and exposure time of the current detection camera into a preset frame rate delay calculation formula to obtain the frame rate delay factor; Obtain the timing data of the Bluetooth headset main board image from collection to the end of quality detection, and the timing data includes the image acquisition time, image preprocessing time, and defect recognition time; Perform statistical analysis on the timing data, calculate the ratio of the standard deviation to the mean value of the timing data to obtain the delay fluctuation coefficient; Input the frame rate delay factor and the delay fluctuation coefficient into a preset processing delay influence coefficient calculation formula to obtain the processing delay influence coefficient.

[0045] Among them, the preset frame rate delay calculation formula is specifically:

[0046] Among them, is the delay fluctuation coefficient, is the current detection frame rate, is the exposure time, It is a camera interface constant.

[0047] Among them, the preset calculation formula for the processing delay influence coefficient is specifically as follows:

[0048] Among them, is the processing delay influence coefficient, is the preset weight coefficient, is the preset reference delay, is the delay fluctuation coefficient, is the delay fluctuation coefficient.

[0049] It should be noted that the timing data from image acquisition to quality inspection end for obtaining the image of the Bluetooth headset mainboard includes image acquisition time, image preprocessing time, and defect recognition time. By statistically analyzing these timing data and calculating the ratio of its standard deviation to the mean, the delay fluctuation coefficient is obtained. This coefficient can reflect the stability of the system in different detection batches and avoid the detection efficiency fluctuation caused by uneven processing time.

[0050] Finally, the frame rate delay factor and the delay fluctuation coefficient are input into the preset calculation formula for the processing delay influence coefficient to obtain the overall processing delay influence coefficient, providing data support for the optimization of the detection strategy. The beneficial effect of the present invention is that by accurately calculating and quantifying the processing delay of each link, the real-time optimization of the detection system is achieved, reducing misjudgments caused by time delay fluctuations, and improving the efficiency and reliability of the surface quality inspection of the Bluetooth headset mainboard.

[0051] S5. Perform data analysis on the output of the vibration interference evaluation model to obtain the optimal detection strategy combination and apply it to the surface quality inspection of the Bluetooth headset mainboard.

[0052] In this example, performing data analysis on the output of the vibration interference evaluation model to obtain the optimal detection strategy combination and apply it to the surface quality inspection of the Bluetooth headset mainboard is specifically as follows: Based on the vibration interference evaluation model, obtain the vibration interference evaluation value corresponding to each detection strategy combination; Map the multi-dimensional parameters of each detection strategy combination to a detection strategy combination identification code according to the hash function, and the multi-dimensional parameters include detection frame rate, exposure time, and vibration compensation coefficient; Construct a two-dimensional curve with the detection strategy combination identification code as the x-axis and the vibration interference evaluation value as the y-axis; Perform data analysis on the two-dimensional curve, and use the detection strategy combination identification code corresponding to the lowest point of the two-dimensional curve as the optimal detection strategy combination; Apply the optimal detection strategy combination to the surface quality inspection of the Bluetooth headset mainboard.

[0053] It should be noted that, first, based on the vibration interference evaluation model, the vibration interference evaluation value corresponding to each detection strategy combination is calculated. The vibration interference evaluation value can reflect the stability of different strategy combinations in the vibration interference environment, thus providing data support for screening the optimal strategy.

[0054] Next, the multi-dimensional parameters of each detection strategy combination (including detection frame rate, exposure time, and vibration compensation coefficient) are mapped to a unique detection strategy combination identification code through a hash function. Through this mapping method, not only the efficiency of data processing is improved, but also the uniqueness of each detection strategy is ensured. Subsequently, a two-dimensional curve is constructed with the detection strategy combination identification code as the x-axis and the vibration interference evaluation value as the y-axis. This curve can intuitively display the performance of each detection strategy combination under different vibration interferences, providing a graphical reference for subsequent strategy selection.

[0055] By analyzing the data of the two-dimensional curve, the system can find the detection strategy combination identification code corresponding to the lowest point and determine this combination as the optimal strategy. Finally, this optimal detection strategy combination is applied to the surface quality detection of the Bluetooth headset main board, thereby maximizing the accuracy of the detection process.

[0056] Embodiment 2 Figure 2 A surface quality detection system for a Bluetooth headset main board according to the present invention is provided, including a data acquisition module, a vibration compensation module, a strategy combination module, an interference evaluation module, and a quality detection module: The data acquisition module is used to collect mechanical vibration signals of the Bluetooth headset production line to generate a vibration spectrum matrix, and obtain an anti-vibration imaging sequence based on the vibration spectrum matrix; The vibration compensation module is used to perform temporal alignment matching on the anti-vibration imaging sequence based on the DTW algorithm to obtain vibration compensation coefficients corresponding to different DTW thresholds; The strategy combination module is used to construct a number of detection strategy combinations based on the vibration compensation coefficients corresponding to different DTW thresholds; The interference evaluation module is used to respectively apply each detection strategy combination to collect images of the Bluetooth headset main board, and synchronously obtain image clarity data and processing delay data to construct a vibration interference evaluation model; The quality detection module is used to perform data analysis on the output of the vibration interference evaluation model to obtain the optimal detection strategy combination and apply it to the surface quality detection of the Bluetooth headset main board.

[0057] The above formulas are all dimensionless and take their numerical values for calculation. The formulas are obtained by collecting a large amount of data for software simulation to obtain a formula closest to the actual situation. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0058] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product.

[0059] Those of ordinary skill in the art will appreciate that the modules and algorithm steps of the examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or in a combination of computer software and electronic hardware. Whether these functions are executed in hardware or software depends on the specific application and design constraints of the technical solution. A person skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of this application.

[0060] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically alone, or two or more modules can be integrated into one module.

[0061] As described above, this is only the specific implementation manner of this application, but the protection scope of this application is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed in this application, and all such changes or substitutions should be covered by the protection scope of this application. Therefore, the protection scope of this application should be subject to the protection scope of the claims.

[0062] Finally: The above description is only the preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A method for detecting the surface quality of a Bluetooth headset motherboard, characterized in that: The following steps are involved: The mechanical vibration signals of the Bluetooth headset production line are collected to generate a vibration spectrum matrix, and the anti-vibration imaging sequence is obtained based on the vibration spectrum matrix; Based on the DTW algorithm, the anti-vibration imaging sequence is aligned and matched in time sequence to obtain the vibration compensation coefficients corresponding to different DTW thresholds; Several detection strategy combinations are constructed based on the vibration compensation coefficients corresponding to different DTW thresholds; Each detection strategy combination is applied to collect images of the Bluetooth headset motherboard, and the image clarity data and processing delay data are obtained simultaneously to build a vibration interference assessment model; The output of the vibration disturbance assessment model is analyzed to obtain the optimal detection strategy combination, which is then applied to the surface quality inspection of Bluetooth headset motherboards.

2. The method for detecting the surface quality of a Bluetooth headset motherboard according to claim 1, characterized in that: The collecting of mechanical vibration signals of the Bluetooth headset production line to generate a vibration spectrum matrix is ​​specifically: Collect the mechanical vibration signal of the Bluetooth headset production line, and perform Fourier transform on the mechanical vibration signal of the production line to obtain the frequency domain spectrum; Based on the frequency domain spectrum, the frequency band is divided according to a preset bandwidth to obtain a plurality of first frequency bands; Calculating the energy fluctuation variance of each first frequency band and screening a plurality of first frequency bands based on a preset first threshold value to obtain a plurality of screened first frequency bands; Performing feature extraction on a plurality of filtered first frequency bands to obtain frequency domain feature vectors; Arrange the frequency domain feature vectors in time window sequence to obtain the vibration spectrum matrix.

3. The method for detecting the surface quality of a Bluetooth headset motherboard according to claim 2, characterized in that: The anti-vibration imaging sequence is obtained based on the vibration spectrum matrix, specifically: Divide the vibration spectrum matrix into time windows and calculate the vibration main frequency offset in each time window; According to the vibration main frequency offset in each time window, a vibration phase angle sequence is constructed; The vibration phase angle sequence is used to obtain the vibration-resistant imaging sequence based on the interpolation algorithm.

4. The method for detecting the surface quality of a Bluetooth headset motherboard according to claim 3, characterized in that: Based on the DTW algorithm, the anti-vibration imaging sequence is aligned and matched in time sequence to obtain the vibration compensation coefficients corresponding to different DTW thresholds, specifically: Construct a standard Bluetooth headset motherboard surface anti-vibration imaging sequence library and segment the anti-vibration imaging sequence according to a preset time window to obtain several segmented sequences; The DTW algorithm is used to calculate the path offset between each segmented sequence and the standard Bluetooth headset motherboard surface anti-vibration imaging sequence library to obtain the DTW distance matrix. Based on the DTW distance matrix, the DTW distance is divided into several initial DTW thresholds according to a preset ratio; The vibration compensation coefficients corresponding to different DTW thresholds are calculated based on a preset vibration compensation coefficient calculation formula.

5. The method for detecting the surface quality of a Bluetooth headset motherboard according to claim 4, characterized in that: The vibration compensation coefficients corresponding to different DTW thresholds are used to construct several detection strategy combinations, specifically: Obtain the historical data of mechanical vibration of the Bluetooth headset production line, and determine the initial detection frame rate and camera exposure time based on the historical data of mechanical vibration of the Bluetooth headset production line; Performing random gain on the initial detection frame rate and exposure time based on preset constraints to obtain several detection frame rates and exposure times; The vibration compensation coefficients corresponding to different DTW thresholds are arranged and combined with several detection frame rates and exposure times to obtain several detection strategy combinations.

6. The method for detecting the surface quality of a Bluetooth headset motherboard according to claim 5, characterized in that: The image clarity data includes an image clarity influence coefficient, and a specific method for obtaining the image clarity influence coefficient is as follows: For each detection strategy combination, N frames of Bluetooth headset motherboard surface images are continuously collected, and the mean value of the Sobel edge gradient amplitude in the PCB solder joint area of ​​each frame of the Bluetooth headset motherboard surface image is calculated; Taking the mean value of the Sobel edge gradient amplitude as the first image clarity value, and obtaining N first image clarity values; Calculate the average value and variance of the N first image clarity values, and use the ratio of the average value to the variance as an image clarity evaluation coefficient; Calculate the standard deviation of the vibration angular velocity in the corresponding time window according to the vibration phase angle sequence; The vibration compensation coefficient, image clarity evaluation coefficient and vibration angular velocity standard deviation of the current detection strategy combination are input into a preset image clarity influence coefficient calculation formula to obtain the image clarity influence coefficient.

7. The method for detecting the surface quality of a Bluetooth headset motherboard according to claim 6, characterized in that: The processing delay data includes a processing delay influence coefficient, and a specific method for obtaining the processing delay influence coefficient is as follows: Obtain camera parameters, detection frame rate, and exposure time of the current detection camera, and input the camera parameters, detection frame rate, and exposure time of the current detection camera into a preset frame rate delay calculation formula to obtain a frame rate delay factor; Obtaining time series data of the Bluetooth headset motherboard image from acquisition to the end of quality inspection, wherein the time series data includes image acquisition time, image preprocessing time, and defect recognition time; Perform statistical analysis on the time series data, calculate the ratio of the standard deviation to the mean of the time series data, and obtain the delay fluctuation coefficient; The frame rate delay factor and the delay fluctuation coefficient are input into a preset processing delay influence coefficient calculation formula to obtain the processing delay influence coefficient.

8. The method for detecting the surface quality of a Bluetooth headset motherboard according to claim 7, characterized in that: The output of the vibration interference evaluation model is analyzed to obtain the best detection strategy combination, which is applied to the surface quality detection of the Bluetooth headset motherboard, specifically: Based on the vibration interference evaluation model, the vibration interference evaluation value corresponding to each detection strategy combination is obtained; Mapping the multidimensional parameters of each detection strategy combination into a detection strategy combination identification code according to a hash function, wherein the multidimensional parameters include a detection frame rate, an exposure time, and a vibration compensation coefficient; A two-dimensional curve is constructed with the detection strategy combination identification code as the x-axis and the vibration interference assessment value as the y-axis; Perform data analysis on the two-dimensional curve, and use the detection strategy combination identification code corresponding to the lowest point of the two-dimensional curve as the best detection strategy combination; The optimal combination of detection strategies is applied to the surface quality inspection of Bluetooth headset motherboard.

9. A Bluetooth headset motherboard surface quality detection system, applied to a Bluetooth headset motherboard surface quality detection method according to any one of claims 1 to 8, characterized in that: Including data acquisition module, vibration compensation module, strategy combination module, interference assessment module and quality detection module: A data acquisition module is used to collect mechanical vibration signals of a Bluetooth headset production line to generate a vibration spectrum matrix, and obtain an anti-vibration imaging sequence based on the vibration spectrum matrix; The vibration compensation module is used to perform time alignment and matching on the anti-vibration imaging sequence based on the DTW algorithm to obtain vibration compensation coefficients corresponding to different DTW thresholds; A strategy combination module is used to construct several detection strategy combinations based on vibration compensation coefficients corresponding to different DTW thresholds; The interference assessment module is used to apply each detection strategy combination to collect images of the Bluetooth headset motherboard, and simultaneously obtain image clarity data and processing delay data to build a vibration interference assessment model; The quality detection module is used to perform data analysis on the output of the vibration interference assessment model to obtain the best detection strategy combination, and apply it to the surface quality detection of the Bluetooth headset motherboard.

Images