Packaging control method and system for LED lamp beads

Abstract

The invention relates to the technical field of LED lamp bead production, and discloses a packaging control method and system for LED lamp beads, and the method comprises the steps: determining a target support image, determining a comprehensive dust value of the target support image based on an analysis result, determining the cleaning prediction time of an LED support according to the comprehensive dust value and a dust packaging model, and carrying out the packaging control of the LED lamp beads. Comparing the cleaning prediction time with a historical data set, when it is judged that the cleaning prediction time is corrected, determining a time similarity according to the cleaning prediction time and the historical data set, determining a first time correction coefficient or a cleaning set according to the time similarity, and deleting the historical data set based on the cleaning set, and a second time correction coefficient is determined according to the suspected correction coefficient and the correction weight, and the LED support is cleaned according to the adjusted cleaning prediction time. According to the invention, dependence on artificial experience is avoided, dynamic adjustment is carried out according to actual conditions, and packaging stability is ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of LED lamp bead production. Specifically, it relates to a packaging control method and system for LED lamp beads. Background Art

[0002] With the development of society, the number of street stores is increasing day by day, which poses certain challenges to the production of LED lamp beads in store signs. During the production process, it is necessary to package the LED lamp beads to avoid the electronic chips from being interfered with and damaged by the external environment, so as to provide electrical connection and signal transmission functions. The LED bracket is the base for carrying the LED lamp beads in the packaging process, and its surface cleanliness directly affects the packaging quality and light efficiency of the LED lamp beads.

[0003] Currently, when using a plasma instrument to clean the LED bracket, the cleaning time is usually manually set according to the experience of the operator. Due to the different skill levels and experience degrees of different operators, the cleaning time is easily affected by the operator's skill level and judgment. Moreover, it cannot be dynamically adjusted according to the actual situation, resulting in unstable cleaning quality, which in turn affects the accuracy of the packaging process and ultimately leads to insufficient production efficiency of the LED lamp beads.

[0004] Therefore, how to provide a packaging control method and system for LED lamp beads is an urgent technical problem to be solved by those skilled in the art. Summary of the Invention

[0005] In view of this, the present invention proposes a packaging control method and system for LED lamp beads, aiming to solve the problem that the cleaning time is easily affected by the operator's skill level and judgment, and cannot be dynamically adjusted according to the actual situation, resulting in unstable cleaning quality, which in turn affects the accuracy of the packaging process and ultimately leads to insufficient production efficiency of the LED lamp beads.

[0006] On the one hand, the present invention proposes a packaging control method for LED lamp beads, including: Obtain the initial bracket image of the LED bracket, process the initial bracket image to determine the target bracket image, analyze the target bracket image, determine the comprehensive dust value of the target bracket image based on the analysis result, and determine the cleaning prediction time of the LED bracket according to the comprehensive dust value and the dust packaging model; Compare the cleaning prediction time with the historical data set, and judge whether to correct the cleaning prediction time according to the comparison result; When it is determined to correct the cleaning prediction time, determine the time similarity according to the cleaning prediction time and the historical data set, determine the first time correction coefficient or the cleaning set according to the time similarity, determine the suspected correction coefficient according to the cleaning set, delete the historical data set based on the cleaning set, determine the correction weight of the suspected correction coefficient, and determine the second time correction coefficient according to the suspected correction coefficient and the correction weight; Adjust the cleaning prediction time according to the first time correction coefficient or the second time correction coefficient, and clean the LED bracket according to the adjusted cleaning prediction time.

[0007] Further, when processing the initial bracket image to determine the target bracket image, it includes: Preprocess the initial bracket image, and the preprocessing includes image denoising, color equalization, and normalization of pixel values; Substitute the preprocessed initial bracket image into the neural network model, and based on the neural network model, output the image compliance value of the preprocessed initial bracket image; When the image compliance value is greater than or equal to the preset image compliance value, determine the initial bracket image as the target bracket image; When the image compliance value is less than the preset image compliance value, re-acquire the initial bracket image of the LED bracket.

[0008] Further, when analyzing the target bracket image and determining the comprehensive dust value of the target bracket image based on the analysis result, it includes: Extract all the bracket pixel points of the target bracket image, obtain the bracket pixel values corresponding to all the bracket pixel points, obtain the adapted target bracket image corresponding to the target bracket image, extract all the adapted bracket pixel points of the adapted target bracket image, and obtain the adapted bracket pixel values corresponding to all the adapted bracket pixel points; Compare all the bracket pixel values with all the adapted bracket pixel values, and divide all the bracket pixel points into the dust set according to the comparison result; When the bracket pixel value is greater than the adapted bracket pixel value, divide the bracket pixel point into the first dust set; When the bracket pixel value is equal to the adapted bracket pixel value, do not divide the bracket pixel point; When the bracket pixel value is less than the adapted bracket pixel value, divide the bracket pixel point into the second dust set; Count the first number of the bracket pixel points in the first dust set, count the second number of the bracket pixel points in the second dust set, obtain the first average bracket pixel value of the first dust set, and obtain the second average bracket pixel value of the second dust set; Determine the comprehensive dust value based on the first quantity, the second quantity, the first average bracket pixel value, and the second average bracket pixel value.

[0009] Further, when determining the cleaning prediction time of the LED bracket according to the comprehensive dust value and the dust encapsulation model, it includes: Obtain a comprehensive dust value data set, divide the comprehensive dust value data set into a training set and a test set, use grid search to find the hyperparameters of the random forest model, and establish a random forest model; Iteratively train the random forest model according to the training set, substitute the test set into the iteratively trained random forest model and verify it; When the verification value reaches the verification threshold, determine the iteratively trained random forest model as the dust encapsulation model, and obtain the cleaning prediction time according to the comprehensive dust value.

[0010] Further, when comparing the cleaning prediction time with the historical data set and judging whether to correct the cleaning prediction time according to the comparison result, it includes: The historical data set includes several historical cleaning prediction times, several historical time correction coefficients, several historical comprehensive dust values, and the historical qualified data mean value, and each historical comprehensive dust value corresponds to a historical cleaning prediction time and a historical time correction coefficient; Compare the cleaning prediction time with the historical qualified data mean value, and judge whether to correct the cleaning prediction time according to the comparison result; When the cleaning prediction time is less than the historical qualified data mean value, it is determined that the cleaning prediction time is corrected. Otherwise, the cleaning prediction time is not corrected, and the LED bracket is cleaned according to the cleaning prediction time.

[0011] Further, when determining the time similarity according to the cleaning prediction time and the historical data set, and determining the first time correction coefficient or the cleaning set according to the time similarity, it includes: Compare the cleaning prediction time with each historical cleaning prediction time in the historical data set to determine the time similarity; Compare the time similarity with a preset time similarity threshold, and determine the first time correction coefficient or the cleaning set according to the comparison result; When there is data in the historical data set with a time similarity greater than or equal to the time similarity threshold, determine the first time correction coefficient according to the historical data set. Otherwise, determine the cleaning set according to the historical data set; When determining the first time correction coefficient according to the historical data set, if the data greater than or equal to the time similarity threshold in the historical data set is unique, the historical time correction coefficient corresponding to the data is used as the first time correction coefficient; If the data greater than or equal to the time similarity threshold in the historical data set is not unique, the historical time correction coefficient corresponding to the data with the maximum time similarity is used as the first time correction coefficient; If the data greater than or equal to the time similarity threshold in the historical data set is not unique, and there are several data with the maximum time similarity, the mean value of the historical time correction coefficients corresponding to each data with the maximum time similarity is used as the first time correction coefficient.

[0012] Further, when determining the suspected correction coefficient according to the cleaning set, it includes: Taking the comprehensive dust value, the cleaning prediction time, and the historical data set as the data set to be clustered, extracting the historical time correction coefficient corresponding to each data in the data set to be clustered, determining that the expected number of clusters k is 3, and initializing the parameters of the Gaussian distribution; According to the probability that each data in the data set to be clustered belongs to each Gaussian distribution, obtaining the responsibility value, obtaining the data set corresponding to the comprehensive dust value and the cleaning prediction time according to the responsibility value, and using the data set as the cleaning set; Obtaining the correction mean value of the historical time correction coefficient in the cleaning set, and determining the correction mean value as the suspected correction coefficient.

[0013] Further, when deleting the historical data set based on the cleaning set, determining the correction weight of the suspected correction coefficient, and determining the second time correction coefficient according to the suspected correction coefficient and the correction weight, it includes: Deleting the data in the historical data set that is the same as the cleaning set, recording the deletion quantity, and counting the quantity of the data in the historical data set that has not been deleted; Obtaining the quantity difference between the deletion quantity and the data quantity, obtaining the absolute value of the quantity difference, and recording the ratio of the absolute value of the difference to the deletion quantity as the deviation dispersion degree; Taking the result of the power operation with the natural constant as the base and the deviation dispersion degree as the exponent as the correction weight of the suspected correction coefficient, and the second time correction coefficient is the product value of the suspected correction coefficient and the correction weight.

[0014] Further, when adjusting the cleaning prediction time according to the first time correction coefficient or the second time correction coefficient, it includes: The first time correction coefficient is directly proportional to the cleaning prediction time, and the second time correction coefficient is directly proportional to the cleaning prediction time.

[0015] Compared with the prior art, the beneficial effects of the present invention are as follows: By processing and analyzing the initial bracket image of the LED bracket, the cleaning prediction time is determined based on the comprehensive dust value and the dust encapsulation model, avoiding the risk that the cleaning time does not match the actual pollution situation due to differences in manual experience, improving the reliability and stability of the cleaning prediction time. The cleaning prediction time is compared with the historical data set to dynamically determine whether to make a correction, and the accuracy of the cleaning prediction time is verified with the help of historical experience, providing a direction for precise correction. When correction is required, the first time correction coefficient or the cleaning set is determined according to the time similarity, making full use of the historical data set, and the cleaning prediction time is flexibly corrected according to different situations, so that the corrected cleaning prediction time fits the actual pollution situation of the current LED bracket. By determining the suspected correction coefficient from the cleaning set, and then determining the correction weight based on the cleaning set, and further obtaining the second time correction coefficient, various factors are comprehensively considered, reducing the packaging quality problems caused by improper cleaning time, ensuring the stability of the cleaning quality, and thus improving the production efficiency of the LED lamp beads.

[0016] On the other hand, the present application also provides a packaging control system for LED lamp beads, which is used to apply the above-mentioned packaging control method for LED lamp beads, and includes: An acquisition module, configured to acquire an initial bracket image of the LED bracket, process the initial bracket image to determine a target bracket image, analyze the target bracket image, determine the comprehensive dust value of the target bracket image based on the analysis result, and determine the cleaning prediction time of the LED bracket according to the comprehensive dust value and the dust encapsulation model; A judgment module, configured to compare the cleaning prediction time with the historical data set, and judge whether to correct the cleaning prediction time according to the comparison result; A processing module, configured to, when it is determined to correct the cleaning prediction time, determine the time similarity according to the cleaning prediction time and the historical data set, determine the first time correction coefficient or the cleaning set according to the time similarity, determine the suspected correction coefficient according to the cleaning set, delete the historical data set based on the cleaning set, determine the correction weight of the suspected correction coefficient, and determine the second time correction coefficient according to the suspected correction coefficient and the correction weight; A cleaning module, configured to adjust the cleaning prediction time according to the first time correction coefficient or the second time correction coefficient, and clean the LED bracket according to the adjusted cleaning prediction time.

[0017] It is understandable that the above-mentioned encapsulation control method and system for LED lamp beads have the same beneficial effects, which will not be elaborated here. Description of the Drawings

[0018] By reading the following detailed description of the preferred embodiments, various other advantages and benefits will become clear to those of ordinary skill in the art. The drawings are only for the purpose of showing the preferred embodiments and are not considered as limiting the present invention. Moreover, throughout the drawings, the same reference numerals are used to represent the same components. In the drawings: Figure 1 It is a flowchart of a method for encapsulation control of LED lamp beads provided by an embodiment of the present invention; Figure 2 It is a functional block diagram of a system for encapsulation control of LED lamp beads provided by an embodiment of the present invention. Detailed Embodiments

[0019] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the drawings. Although the exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided so that the present disclosure can be more thoroughly understood and the scope of the present disclosure can be fully conveyed to those skilled in the art. It should be noted that, without conflict, the embodiments in the present invention and the features in the embodiments can be combined with each other. The present invention will be described in detail below with reference to the drawings and in combination with the embodiments.

[0020] Refer to Figure 1 As shown, in some embodiments of the present application, this embodiment provides a method for encapsulation control of LED lamp beads, including: S100: Obtain an initial stent image of the LED stent, process the initial stent image to determine a target stent image, analyze the target stent image, determine the comprehensive dust value of the target stent image based on the analysis result, and determine the cleaning prediction time of the LED stent according to the comprehensive dust value and the dust encapsulation model.

[0021] S200: Compare the cleaning prediction time with the historical data set, and judge whether to correct the cleaning prediction time according to the comparison result.

[0022] S300: When it is determined to correct the cleaning prediction time, determine the time similarity according to the cleaning prediction time and the historical data set, determine the first time correction coefficient or the cleaning set according to the time similarity, determine the suspected correction coefficient according to the cleaning set, delete the historical data set based on the cleaning set, determine the correction weight of the suspected correction coefficient, and determine the second time correction coefficient according to the suspected correction coefficient and the correction weight.

[0023] S400: Adjust the cleaning prediction time according to the first time correction coefficient or the second time correction coefficient, and clean the LED bracket according to the adjusted cleaning prediction time.

[0024] Specifically, use acquisition devices such as infrared cameras to obtain the initial bracket image of the LED bracket. The obtained initial bracket image may have interference factors such as noise and ambient brightness, which affect subsequent judgments. Therefore, the initial bracket image is processed first to determine the target bracket image, avoiding subjective errors caused by manual judgment. Analyze the target image, quantitatively calculate the comprehensive dust value on the LED bracket. The comprehensive dust value reflects the true pollution situation on the surface of the LED bracket. Use the trained dust encapsulation model to determine a cleaning prediction time based on the comprehensive dust value obtained from the image analysis. Predict the cleaning time required for the true pollution situation through the big data model, ensuring the consistency between the cleaning prediction time and the comprehensive dust value, thereby improving the encapsulation quality. Using the dust encapsulation model for prediction cannot ensure the accuracy and reliability of the cleaning prediction time. The historical data set covers various data in the past cleaning process. By comparing the cleaning prediction time with the historical data set, it can accurately judge whether the cleaning prediction time conforms to the past actual situation, and continuously optimize the cleaning prediction time during the cleaning process, ensuring that the cleaning prediction time can meet the actual cleaning needs and avoiding problems with the encapsulation quality caused by excessive or insufficient cleaning. Compared with manual experience, through the comparison of the big data model and the historical data set, the accuracy and reliability of the cleaning prediction time are improved, and the risk of unstable cleaning quality caused by differences in the skills and experience of operators is avoided. When the cleaning prediction time needs to be corrected, calculate the time similarity between the cleaning prediction time and the historical data set to measure the closeness between the cleaning prediction time and the historical data set. According to the time similarity, determine the first time correction coefficient or the cleaning set. For the cleaning set, screen out the suspected correction coefficients, and based on this, screen the historical data set to determine the degree of dispersion between the data to improve the pertinence of the data, thereby obtaining the correction weight of the suspected correction coefficient, and finally determining the second time correction coefficient. The dynamic correction mechanism based on historical data makes the adjustment of the cleaning prediction time conform to the actual situation of the LED bracket. Through multi-dimensional analysis and data screening, the stability of the cleaning quality is further guaranteed, and problems with the encapsulation quality caused by inappropriate cleaning time are avoided. It can be understood that the obtained first-time correction coefficient or second-time correction coefficient is applied to the cleaning prediction time to determine the adjusted cleaning prediction time, and the LED bracket is cleaned accordingly. Through analysis and adjustment, the adjusted cleaning prediction time will neither cause waste of resources due to excessive cleaning nor affect the encapsulation quality due to insufficient cleaning, which helps to improve the accuracy of the encapsulation process, reduce the defective rate of LED lamp beads, and thus improve the production efficiency of LED lamp beads.

[0025] In some embodiments of the present application, when processing the initial bracket image to determine the target bracket image, it includes: preprocessing the initial bracket image, and the preprocessing includes image denoising, color equalization, and normalizing pixel values. Substitute the preprocessed initial bracket image into the neural network model, and based on the output of the neural network model, obtain the image compliance value of the preprocessed initial bracket image. When the image compliance value is greater than or equal to the preset image compliance value, determine the initial bracket image as the target bracket image. When the image compliance value is less than the preset image compliance value, re-acquire the initial bracket image of the LED bracket.

[0026] Specifically, image denoising can remove the random noise introduced during the acquisition of the initial bracket image. The presence of noise will interfere with the subsequent dust analysis of the LED bracket, resulting in deviations in the calculation of the comprehensive dust value. After denoising, the detailed information of the image is retained, and color equalization makes the color distribution of the image more uniform, avoiding local over-brightness or over-darkness of the image caused by factors such as uneven illumination, which affects the judgment of the surface condition of the LED bracket. Normalizing pixel values unifies the image pixel values obtained by different acquisition devices into a specific range, eliminating data deviations caused by device differences. Ensuring the consistency of image data, a neural network model is pre-trained in advance. This model includes multiple convolutional layers, pooling layers, and fully connected layers. The convolutional layer is used to extract the deep features of the image, the pooling layer is used to reduce the feature dimension, and the fully connected layer is used for the final classification judgment. The preset image compliance value is 0.8. The neural network model can learn to distinguish the features of image compliance and output the probability of image compliance or non-compliance through the softmax function, ensuring that the target bracket image can accurately reflect the true pollution situation of the LED bracket.

[0027] In some embodiments of the present application, when analyzing a target stent image and determining the comprehensive dust value of the target stent image based on the analysis result, it includes: extracting all stent pixel points of the target stent image, obtaining the stent pixel values corresponding to all stent pixel points, obtaining an adapted target stent image corresponding to the target stent image, extracting all adapted stent pixel points of the adapted target stent image, obtaining the adapted stent pixel values corresponding to all adapted stent pixel points, comparing all stent pixel values with all adapted stent pixel values, dividing all stent pixel points into a dust set according to the comparison result. When the stent pixel value is greater than the adapted stent pixel value, dividing the stent pixel point into the first dust set. When the stent pixel value is equal to the adapted stent pixel value, the stent pixel point is not divided. When the stent pixel value is less than the adapted stent pixel value, dividing the stent pixel point into the second dust set. Counting the first number of stent pixel points in the first dust set, counting the second number of stent pixel points in the second dust set, obtaining the first average stent pixel value of the first dust set, obtaining the second average stent pixel value of the second dust set, and determining the comprehensive dust value based on the first number, the second number, the first average stent pixel value, and the second average stent pixel value.

[0028] Specifically, the comprehensive dust value is obtained by the following formula:

[0029] Wherein, L represents the comprehensive dust value, A1 represents the first average stent pixel value, p represents the first number, Xi represents the stent pixel value of the i-th stent pixel point in the first dust set, Yi represents the adapted stent pixel value corresponding to the i-th stent pixel point in the first dust set, A2 represents the second average stent pixel value, q represents the second number, Nj represents the adapted stent pixel value corresponding to the j-th stent pixel point in the second dust set, and Mj represents the stent pixel value of the j-th stent pixel point in the second dust set.

[0030] Specifically, the target bracket image is adapted based on the dust-free LED bracket, and the bracket pixel points and the adapted bracket pixel points correspond one by one, ensuring that the bracket pixel values and the adapted bracket pixel values correspond one by one. The dust set includes a first dust set and a second dust set. Although the target bracket image has been processed such as image denoising and color equalization, the illumination during actual shooting will cause the pixel values in different regions of the target bracket image to change. When the dust is in a region with relatively dim illumination while the background is in a region with relatively strong illumination, even if the reflection characteristics of the dust itself should result in a relatively large bracket pixel value under normal illumination, the bracket pixel value will be relatively small due to the influence of the background illumination at this time. On the contrary, when the dust is under strong light irradiation while the background illumination is weak, even if the dust itself is relatively dark, its bracket pixel value may be relatively large. Therefore, dividing the bracket pixel points that are not equal to the adapted bracket pixel values lays the foundation for the calculation of the comprehensive dust value and ensures the calculation accuracy of the comprehensive dust value. In some embodiments of the present application, when determining the cleaning prediction time of the LED bracket according to the comprehensive dust value and the dust encapsulation model, it includes: obtaining the comprehensive dust value data set, dividing the comprehensive dust value data set into a training set and a test set, using grid search to find the hyperparameters of the random forest model, establishing the random forest model, iteratively training the random forest model according to the training set, substituting the test set into the iteratively trained random forest model for verification, and when the verification value reaches the verification threshold, determining the iteratively trained random forest model as the dust encapsulation model, and obtaining the cleaning prediction time according to the comprehensive dust value.

[0031] Specifically, the comprehensive dust value data set contains key data such as various pollution situations of the LED bracket, environmental dust concentration, and cleaning time under different conditions. The comprehensive dust value data set is divided into a training set and a test set. 50%-90% of the data is used as the training set, and the rest is used as the test set. Ensure that both the training set and the test set contain various data to improve the generalization ability of the model. The training set is used for the random forest model, and the test subset is used to verify the performance of the trained model. Grid search establishes the random forest model by exhaustively searching for hyperparameters in the parameter space. The random forest model includes parameters such as the number of trees and the maximum depth of the trees, aiming to capture the complex relationships in the data. When iteratively training the random forest model using the data in the training set, the model will try to learn the patterns and relationships in the data to improve its prediction or classification ability. After iterative training, the data in the test set is used to verify the model. The verification metrics include accuracy, F1 value, etc. By continuously iteratively training the random forest model, the cleaning prediction time can be stably output. When the verification value reaches the verification threshold, the iteratively trained random forest model is determined as the dust encapsulation model, and the verification threshold is dynamically set according to the verification metrics. For example, the accuracy is set to 80% and the F1 value is set to 0.7.

[0032] It can be understood that the training process of the neural network model is consistent with that of the random forest model. The training of the neural network model and the use of the model to process images are repetitive and redundant, so they are not described in detail. By iteratively training the random forest model, a dust encapsulation model can be determined. Substituting the comprehensive dust value into the dust encapsulation model can obtain the cleaning prediction time, which improves the stability and reliability of the cleaning process.

[0033] In some embodiments of the present application, when comparing the cleaning prediction time with the historical data set and determining whether to correct the cleaning prediction time according to the comparison result, it includes: the historical data set includes several historical cleaning prediction times, several historical time correction coefficients, several historical comprehensive dust values, and the historical qualified data mean value. And each historical comprehensive dust value corresponds to a historical cleaning prediction time and a historical time correction coefficient. Compare the cleaning prediction time with the historical qualified data mean value, and determine whether to correct the cleaning prediction time according to the comparison result. When the cleaning prediction time is less than the historical qualified data mean value, it is determined to correct the cleaning prediction time. Otherwise, do not correct the cleaning prediction time, and clean the LED bracket with the cleaning prediction time.

[0034] Specifically, the historical qualified data mean value is set according to the cleaning requirements. Taking it as a judgment index, its specific value can be changed during actual cleaning. By introducing the historical qualified data mean value as a reference benchmark, the accuracy and automation degree of the correction of the cleaning prediction time are improved. When the cleaning prediction time is less than the historical qualified data mean value, it is determined to correct it because too short a cleaning prediction time may lead to incomplete cleaning of the LED bracket, and the remaining dust will affect the encapsulation quality and light efficiency of the LED lamp beads. By correcting the cleaning prediction time, the interference and error of human judgment are reduced, the adaptability to the actual pollution situation of different LED brackets is improved, thereby effectively ensuring the cleaning quality and improving the production efficiency of LED lamp beads.

[0035] In some embodiments of the present application, when determining the time similarity according to the cleaning prediction time and the historical data set, and determining the first time correction coefficient or the cleaning set according to the time similarity, the following steps are included: comparing the cleaning prediction time with each historical cleaning prediction time in the historical data set to determine the time similarity, comparing the time similarity with a preset time similarity threshold, and determining the first time correction coefficient or the cleaning set according to the comparison result. When there is data in the historical data set with a time similarity greater than or equal to the time similarity threshold, the first time correction coefficient is determined according to the historical data set; otherwise, the cleaning set is determined according to the historical data set. When determining the first time correction coefficient according to the historical data set, if the data with a time similarity greater than or equal to the time similarity threshold in the historical data set is unique, the corresponding historical time correction coefficient of the data is used as the first time correction coefficient; if the data with a time similarity greater than or equal to the time similarity threshold in the historical data set is not unique, the historical time correction coefficient corresponding to the data with the maximum time similarity is used as the first time correction coefficient; if the data with a time similarity greater than or equal to the time similarity threshold in the historical data set is not unique and there are several data with the maximum time similarity, the average value of the historical time correction coefficients corresponding to the data with the maximum time similarity is used as the first time correction coefficient.

[0036] Specifically, the time similarity is obtained by the following formula:

[0037] Where Q represents the time similarity, Lh represents the normalized result of the h-th historical comprehensive dust value in the historical data set, LA represents the normalized result of the comprehensive dust value, Th represents the normalized result of the historical cleaning prediction time corresponding to the h-th historical comprehensive dust value, and TA represents the normalized result of the cleaning prediction time.

[0038] Specifically, by normalizing each data to facilitate operations under the same dimension, data under different cleaning conditions can be compared. By judging the matching degree between the current cleaning condition and the historical successful condition through the time similarity threshold, when historical data with a high time similarity is found, these data can be directly used to determine the first-time correction coefficient, thus ensuring the reliability and consistency of the cleaning process. If the data greater than or equal to the time similarity threshold in the historical dataset is unique, the historical time correction coefficient corresponding to this data can be directly used as the first-time correction coefficient, avoiding complex calculations and judgments and improving the accuracy of the correction process. If the data greater than or equal to the time similarity threshold in the historical dataset is not unique, it will be processed according to different situations: if the data with the maximum time similarity is unique, the historical time correction coefficient corresponding to it will be used as the first-time correction coefficient; if there are several data with the maximum time similarity, the mean value of the historical time correction coefficients corresponding to these data will be used as the first-time correction coefficient. Making full use of the historical data in the historical dataset and finding the most suitable first-time correction coefficient among multiple similar situations ensures the accuracy of the correction.

[0039] It can be understood that for the situation where the current cleaning condition does not fully match the historical data, the cleaning set is determined according to the historical dataset to further analyze the relationship between the historical data, laying a foundation for determining the second-time correction coefficient later.

[0040] In some embodiments of the present application, when determining the suspected correction coefficient according to the cleaning set, it includes: taking the comprehensive dust value, the cleaning prediction time, and the historical dataset as the dataset to be clustered, extracting the historical time correction coefficient corresponding to each data in the dataset to be clustered, determining that the expected number of clusters k is 3, initializing the parameters of the Gaussian distribution, obtaining the responsibility value according to the probability that each data in the dataset to be clustered belongs to each Gaussian distribution, obtaining the dataset corresponding to the comprehensive dust value and the cleaning prediction time according to the responsibility value, and taking the dataset as the cleaning set, obtaining the correction mean value of the historical time correction coefficient in the cleaning set, and determining the correction mean value as the suspected correction coefficient.

[0041] In some embodiments of the present application, when deleting the historical dataset based on the cleaning set to determine the correction weight of the suspected correction coefficient and determining the second-time correction coefficient according to the suspected correction coefficient and the correction weight, it includes: deleting the data in the historical dataset that is the same as the cleaning set, recording the deletion quantity, and counting the quantity of the data in the historical dataset that has not been deleted, obtaining the quantity difference between the deletion quantity and the data quantity, and obtaining the absolute value of the quantity difference, taking the ratio of the absolute value of the quantity difference to the deletion quantity as the deviation dispersion degree, and taking the power operation result with the natural constant as the base and the deviation dispersion degree as the exponent as the correction weight of the suspected correction coefficient, and the second-time correction coefficient is the product value of the suspected correction coefficient and the correction weight.

[0042] Specifically, if the historical dataset has not been effectively accumulated yet, statistics are conducted according to the industry cleaning standard of the LED lamp bead encapsulation process and the cleaning time when the cleaning operation is successfully completed. The historical data is analyzed through a clustering algorithm to find the cleaning set that is closest to the current cleaning conditions, which improves the accuracy of determining the suspected correction coefficient, reduces the dependence on manual experience and judgment, reduces the human error during the cleaning process, and enhances the overall automation level and reliability. The corrected mean of the historical time correction coefficients in the cleaning set is used as the suspected correction coefficient. This is an effective integration of the similar historical data after clustering. Using the corrected mean as the suspected correction coefficient eliminates the influence of individual data, thereby obtaining a stable suspected correction coefficient that can represent the characteristics of the cleaning set.

[0043] It can be understood that during the clustering process, the cleaning set utilizes some historical data that is closest to the current cleaning conditions. Therefore, there are still some historical data in the historical dataset that have not been clustered. Although these unclustered historical data are not as representative as the data in the cleaning set, they are still part of the historical dataset and can reflect the distribution characteristics and change trends of the overall data, avoiding only focusing on the cleaning set and ignoring the influence of the overall data, making the calculation of the correction weight comprehensive and objective. The deviation dispersion measures the standard of the representativeness of the cleaning set for the historical dataset. If there are more unclustered data, it means that the deletion quantity is less and the data quantity is more, indicating that the representativeness of the cleaning set is limited, then the deviation dispersion is larger, that is, the correction weight is larger. On the contrary, if there are fewer unclustered data, it means that the cleaning set has better representativeness, then the deviation dispersion is smaller, that is, the correction weight is smaller. By comprehensively utilizing a large amount of historical data, rich reference information is provided for the determination of the suspected correction coefficient and the correction weight, and learning and optimization are carried out from historical experience, improving the cleaning efficiency.

[0044] In some embodiments of the present application, when adjusting the cleaning prediction time according to the first time correction coefficient or the second time correction coefficient, it includes: the first time correction coefficient is in a direct proportional relationship with the cleaning prediction time, and the second time correction coefficient is in a direct proportional relationship with the cleaning prediction time.

[0045] Specifically, assume that the first-time correction coefficient or the first-time correction coefficient is R, and the cleaning prediction time is T. Determine the adjusted cleaning prediction time as R*T. By establishing a direct proportional relationship among the first-time correction coefficient, the second-time correction coefficient, and the cleaning prediction time, precise control of the cleaning prediction time is achieved. The initial bracket image of the LED bracket is obtained in real time to determine the cleaning prediction time, and the first-time correction coefficient or the second-time correction coefficient is obtained based on the historical data set. The cleaning prediction time is corrected according to the first-time correction coefficient or the second-time correction coefficient, improving the reliability of the cleaning process, reducing the defective rate of LED lamp beads, and thus increasing the production efficiency of LED lamp beads.

[0046] In summary, the beneficial effects of the present invention are as follows: By processing and analyzing the initial bracket image of the LED bracket, the cleaning prediction time is determined based on the comprehensive dust value and the dust encapsulation model, avoiding the risk of mismatch between the cleaning time and the actual pollution situation caused by differences in manual experience, improving the reliability and stability of the cleaning prediction time. Comparing the cleaning prediction time with the historical data set to dynamically determine whether to make a correction, and verifying the accuracy of the cleaning prediction time with historical experience, providing a direction for precise correction. When correction is required, the first-time correction coefficient or the cleaning set is determined according to the time similarity, making full use of the historical data set, and flexibly correcting the cleaning prediction time according to different situations, so that the corrected cleaning prediction time fits the actual pollution situation of the current LED bracket. By determining the suspected correction coefficient from the cleaning set, and then determining the correction weight based on the cleaning set, and further obtaining the second-time correction coefficient, various factors are comprehensively considered, reducing the packaging quality problems caused by improper cleaning time, ensuring the stability of the cleaning quality, and thus increasing the production efficiency of LED lamp beads.

[0047] In another preferred embodiment based on the above embodiments, refer to Figure 2 As shown, this embodiment provides a packaging control system for LED lamp beads, which is used to apply the above packaging control method for LED lamp beads, including: An acquisition module, configured to obtain the initial bracket image of the LED bracket, process the initial bracket image to determine the target bracket image, analyze the target bracket image, determine the comprehensive dust value of the target bracket image based on the analysis result, and determine the cleaning prediction time of the LED bracket according to the comprehensive dust value and the dust encapsulation model.

[0048] A judgment module, configured to compare the cleaning prediction time with the historical data set and judge whether to correct the cleaning prediction time according to the comparison result.

[0049] A processing module, configured to, when it is determined to correct the cleaning prediction time, determine a time similarity based on the cleaning prediction time and a historical data set, determine a first time correction coefficient or a cleaning set according to the time similarity, determine a suspected correction coefficient according to the cleaning set, delete the historical data set based on the cleaning set, determine a correction weight of the suspected correction coefficient, and determine a second time correction coefficient according to the suspected correction coefficient and the correction weight.

[0050] A cleaning module, configured to adjust the cleaning prediction time according to the first time correction coefficient or the second time correction coefficient, and clean the LED bracket according to the adjusted cleaning prediction time.

[0051] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0052] The present application is described with reference to the flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or block in the flowcharts and / or block diagrams can be implemented by computer program instructions, and the combination of the processes and / or blocks in the flowcharts and / or block diagrams can also be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate a device for implementing the functions specified in one Figure 1 process or multiple processes and / or blocks Figure 1 or multiple blocks.

[0053] These computer program instructions can also be stored in a computer-readable storage that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable storage generate a manufactured article including an instruction device, and the instruction device implements the functions specified in one Figure 1 process or multiple processes and / or blocks Figure 1 or multiple blocks.

[0054] These computer program instructions can also be loaded onto a computer or other programmable data processing device, so that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process, and thus the instructions executed on the computer or other programmable device provide for implementing the processFigure 1 One process or multiple processes and / or boxes Figure 1 Steps of the functions specified in one box or multiple boxes.

[0055] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the above embodiments, those of ordinary skill in the art should understand that: modifications or equivalent replacements can still be made to the specific embodiments of the present invention. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention shall be covered by the protection scope of the claims of the present invention.

Claims

1. A packaging control method for LED lamp beads, characterized in that, Including: Obtain an initial bracket image of the LED bracket, process the initial bracket image to determine a target bracket image, analyze the target bracket image, determine a comprehensive dust value of the target bracket image based on the analysis result, and determine a cleaning prediction time of the LED bracket according to the comprehensive dust value and a dust encapsulation model; Compare the cleaning prediction time with a historical data set, and determine whether to correct the cleaning prediction time according to the comparison result; When it is determined to correct the cleaning prediction time, determine a time similarity according to the cleaning prediction time and the historical data set, determine a first time correction coefficient or a cleaning set according to the time similarity, determine a suspected correction coefficient according to the cleaning set, delete the historical data set based on the cleaning set, determine a correction weight of the suspected correction coefficient, and determine a second time correction coefficient according to the suspected correction coefficient and the correction weight; Adjust the cleaning prediction time according to the first time correction coefficient or the second time correction coefficient, and clean the LED bracket according to the adjusted cleaning prediction time.

2. The encapsulation control method for LED lamp beads according to claim 1, wherein When processing the initial bracket image to determine a target bracket image, it includes: Preprocess the initial bracket image, and the preprocessing includes image denoising, color equalization, and normalization of pixel values; Substitute the preprocessed initial bracket image into a neural network model, and output an image compliance value of the preprocessed initial bracket image based on the neural network model; When the image compliance value is greater than or equal to a preset image compliance value, determine the initial bracket image as the target bracket image; When the image compliance value is less than the preset image compliance value, re-obtain the initial bracket image of the LED bracket.

3. The encapsulation control method for LED lamp beads according to claim 2, wherein, When analyzing the target bracket image and determining a comprehensive dust value of the target bracket image based on the analysis result, it includes: Extract all bracket pixel points of the target bracket image, obtain bracket pixel values corresponding to all bracket pixel points, obtain an adapted target bracket image corresponding to the target bracket image, extract all adapted bracket pixel points of the adapted target bracket image, and obtain adapted bracket pixel values corresponding to all adapted bracket pixel points; Compare all bracket pixel values with all adapted bracket pixel values, and divide all bracket pixel points into a dust set according to the comparison result; When the bracket pixel value is greater than the adapted bracket pixel value, divide the bracket pixel point into a first dust set; When the bracket pixel value is equal to the adapted bracket pixel value, do not divide the bracket pixel point; When the bracket pixel value is less than the adapted bracket pixel value, divide the bracket pixel point into a second dust set; Count a first quantity of bracket pixel points in the first dust set, count a second quantity of bracket pixel points in the second dust set, obtain a first average bracket pixel value of the first dust set, and obtain a second average bracket pixel value of the second dust set; Determine the comprehensive dust value based on the first quantity, the second quantity, the first average bracket pixel value, and the second average bracket pixel value.

4. The encapsulation control method for LED lamp beads according to claim 3, wherein, When determining the cleaning prediction time of the LED bracket according to the comprehensive dust value and the dust encapsulation model, it includes: Obtain a comprehensive dust value data set, divide the comprehensive dust value data set into a training set and a test set, use grid search to find the hyperparameters of the random forest model, and establish a random forest model; Iteratively train the random forest model according to the training set, substitute the test set into the iteratively trained random forest model and verify it; When the verification value reaches the verification threshold, determine the iteratively trained random forest model as the dust encapsulation model, and obtain the cleaning prediction time according to the comprehensive dust value.

5. The encapsulation control method for LED lamp beads according to claim 4, characterized in that, When comparing the cleaning prediction time with the historical data set and judging whether to correct the cleaning prediction time according to the comparison result, it includes: The historical data set includes several historical cleaning prediction times, several historical time correction coefficients, several historical comprehensive dust values and the mean value of historical qualified data, and each historical comprehensive dust value corresponds to a historical cleaning prediction time and a historical time correction coefficient; Compare the cleaning prediction time with the mean value of the historical qualified data, and judge whether to correct the cleaning prediction time according to the comparison result; When the cleaning prediction time is less than the mean value of the historical qualified data, it is determined that the cleaning prediction time is corrected. Otherwise, the cleaning prediction time is not corrected, and the LED bracket is cleaned according to the cleaning prediction time.

6. The encapsulation control method for LED lamp beads according to claim 5, wherein When determining the time similarity according to the cleaning prediction time and the historical data set, and determining the first time correction coefficient or the cleaning set according to the time similarity, it includes: Compare the cleaning prediction time with each historical cleaning prediction time in the historical data set to determine the time similarity; Compare the time similarity with a preset time similarity threshold, and determine the first time correction coefficient or the cleaning set according to the comparison result; When there is data in the historical data set with a time similarity greater than or equal to the time similarity threshold, determine the first time correction coefficient according to the historical data set. Otherwise, determine the cleaning set according to the historical data set; When determining the first time correction coefficient according to the historical data set, if the data greater than or equal to the time similarity threshold in the historical data set is unique, use the historical time correction coefficient corresponding to the data as the first time correction coefficient; If the data greater than or equal to the time similarity threshold in the historical data set is not unique, use the historical time correction coefficient corresponding to the data with the maximum time similarity as the first time correction coefficient; If the data greater than or equal to the time similarity threshold in the historical data set is not unique and there are several data with the maximum time similarity, use the mean value of the historical time correction coefficients corresponding to the data with the maximum time similarity as the first time correction coefficient.

7. The encapsulation control method for LED lamp beads according to claim 6, characterized in that, When determining the suspected correction coefficient according to the cleaning set, it includes: Taking the comprehensive dust value, the cleaning prediction time, and the historical data set as the data set to be clustered, extracting the historical time correction coefficient corresponding to each data in the data set to be clustered, determining the expected number of clusters k as 3, and initializing the parameters of the Gaussian distribution; According to the probability that each data in the data set to be clustered belongs to each Gaussian distribution, obtaining the responsibility value, obtaining the data set corresponding to the comprehensive dust value and the cleaning prediction time according to the responsibility value, and using the data set as the cleaning set; Obtaining the corrected mean of the historical time correction coefficient in the cleaning set, and determining the corrected mean as the suspected correction coefficient.

8. The encapsulation control method for LED lamp beads according to claim 7, characterized in that, When deleting the historical data set based on the cleaning set and determining the correction weight of the suspected correction coefficient, and determining the second time correction coefficient according to the suspected correction coefficient and the correction weight, it includes: Deleting the data in the historical data set that is the same as the cleaning set, recording the deletion quantity, and counting the quantity of the data in the historical data set that has not been deleted; Obtaining the quantity difference between the deletion quantity and the data quantity, and obtaining the absolute value of the quantity difference, and recording the ratio of the absolute value of the quantity difference to the deletion quantity as the deviation dispersion degree; Taking the result of the power operation with the natural constant as the base and the deviation dispersion degree as the exponent as the correction weight of the suspected correction coefficient, and the second time correction coefficient is the product value of the suspected correction coefficient and the correction weight.

9. The encapsulation control method for LED lamp beads according to claim 8, characterized in that When adjusting the cleaning prediction time according to the first time correction coefficient or the second time correction coefficient, it includes: The first time correction coefficient is in a direct proportion relationship with the cleaning prediction time, and the second time correction coefficient is in a direct proportion relationship with the cleaning prediction time.

10. A packaging control system for LED lamp beads, which is used to apply the packaging control method for LED lamp beads according to any one of claims 1-9, characterized in that, It includes: A collection module, configured to obtain an initial bracket image of the LED bracket, process the initial bracket image to determine a target bracket image, analyze the target bracket image, determine the comprehensive dust value of the target bracket image based on the analysis result, and determine the cleaning prediction time of the LED bracket according to the comprehensive dust value and the dust encapsulation model; A judgment module, configured to compare the cleaning prediction time with the historical data set, and judge whether to correct the cleaning prediction time according to the comparison result; A processing module, configured to, when it is determined to correct the cleaning prediction time, determine the time similarity according to the cleaning prediction time and the historical data set, determine the first time correction coefficient or the cleaning set according to the time similarity, determine the suspected correction coefficient according to the cleaning set, delete the historical data set based on the cleaning set, determine the correction weight of the suspected correction coefficient, and determine the second time correction coefficient according to the suspected correction coefficient and the correction weight; A cleaning module, configured to adjust the cleaning prediction time according to the first time correction coefficient or the second time correction coefficient, and clean the LED bracket according to the adjusted cleaning prediction time.

Images