Luggage production data identification processing system

Abstract

The invention discloses a luggage production data identification processing system, and relates to the field of identification processing. The luggage production data identification processing system comprises a data acquisition subsystem, a data analysis subsystem, a judgment subsystem and a classification processing subsystem, the data acquisition subsystem is used for acquiring production data, and the production data comprises the number of produced bags and the label information, appearance data and quality data of each bag; and the data analysis subsystem is used for respectively extracting the appearance data and the quality data of each luggage and carrying out comprehensive analysis according to the extraction result. The judgment subsystem can automatically judge whether the appearance characteristic index and the quality characteristic index of each luggage meet the set production standard or not; therefore, automatic production standard inspection is realized, the production efficiency can be greatly improved, the time and cost required by manual inspection are reduced, and the possibility of subjective errors is reduced.

Description

[0001] This application is a divisional application of the application filed on June 20, 2024, with application number 202410797732.1 and invention name “A luggage production data identification and processing system”. Technical Field

[0002] The present invention relates to the field of identification and processing, and in particular to a bag production data identification and processing system. Background Art

[0003] The current luggage production industry is also facing a series of complex challenges. First, as the global market's demand for sustainable development continues to increase, luggage production needs to better integrate sustainable materials and manufacturing methods to reduce dependence on traditional resources, reduce environmental pollution, and achieve sustainable development. Secondly, the rapid development of e-commerce has led to a rapid increase in demand for luggage products, which has increased the challenges of production management, such as the uncertainty of order changes and the pressure of production cycles. Finally, consumers' demand for personalized and customized products continues to increase, which requires production systems to have higher flexibility and customization capabilities. Traditional production processes and technologies often cannot meet this demand, resulting in low production efficiency and waste of resources. Therefore, designing a luggage production data recognition and processing system that can achieve intelligent, flexible and sustainable is an urgent problem that the current industry needs to solve.

[0004] In the prior art, in traditional production management, the analysis of luggage production data often requires manual operation, resulting in low efficiency. At the same time, since the compliance with production standards often depends on manual observation and judgment, subjective deviations are prone to occur. Summary of the invention

[0005] In view of the shortcomings of the prior art, the present invention provides a luggage production data identification and processing system, which solves the problem that in traditional production management, the analysis of luggage production data often requires manual operation, resulting in low efficiency. At the same time, since the compliance with production standards often depends on manual observation and judgment, subjective deviations are prone to occur.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solutions: a luggage production data identification and processing system, comprising: a data acquisition subsystem, a data analysis subsystem, a judgment subsystem, and a classification processing subsystem; the data acquisition subsystem is used to acquire production data, and the production data includes the number of luggage produced and the label information, appearance data, and quality data of each luggage; the data analysis subsystem is used to extract the appearance data and quality data of each luggage, respectively, and conduct a comprehensive analysis based on the extraction results to obtain the appearance feature index and quality feature index of each luggage, and fit the label information, appearance feature index, and quality feature index of each luggage to obtain the label data of each luggage; the judgment subsystem is used to judge whether the label data of each luggage meets the luggage production standards, respectively, and mark the label data of luggage that does not meet the luggage production standards; the classification processing subsystem is used to identify luggage with marked label data and classify and transport them.

[0007] Furthermore, the label information of each bag is specifically the production serial number information of each bag, the appearance data is specifically the color data of the bag, the pattern data is specifically the image data and shape data containing the bag pattern, the color data is specifically the pixel value of each pixel point of the bag, the shape data is specifically the three-dimensional pixel image data set of the pixel point information data of each surface of the bag, the quality data is specifically the weight value and size data of the bag, and the size data is specifically the pixel point information data of each surface of the bag and the three-dimensional pixel coordinate image data set of the position coordinates of each pixel point.

[0008] Furthermore, the specific process of obtaining the appearance characteristic index of each bag is as follows: the color data, pattern data, and shape data of each bag are read, and the color index, pattern index, and shape index of each bag are analyzed and processed respectively to obtain the color index, pattern index, and shape index of each bag; the color index, pattern index, and shape index of each bag are normalized, and the color index, pattern index, and shape index of each bag after normalization are weighted summed to obtain the appearance characteristic index of each bag. The calculation formula is as follows: Among them, WG a is the appearance characteristic index of the a-th bag, YS a is the color index of the ath bag after normalization, YS′ is the color reference index of the bag, TZ a is the pattern index of the a-th bag after normalization, TZ′ is the set pattern reference index of the bag, XZ ais the shape index of the ath bag after normalization, XZ′ is the set shape reference index of the bag, α1 is the proportional coefficient of the set color index, α2 is the proportional coefficient of the set pattern index, α3 is the proportional coefficient of the set shape index, α1+α2+α3=1, a=1,2,3,…,A, A is the total number of bags produced.

[0009] Furthermore, the color data of each bag is read and analyzed to obtain the color index of each bag as follows: the pixel value of each pixel of each bag is read, and the pixel value is expressed in the form of RGB value; the color data of each bag is converted into color space to obtain the HSV value of each pixel of each bag; each pixel of each bag is screened based on the set color interval; the number of pixels in each color interval is counted for each bag; feature vectors are established based on the number of pixel values ​​corresponding to each color interval of each bag; the feature vectors of each color interval of each bag are weighted to obtain the color index of each bag.

[0010] Furthermore, the pattern data of each bag is read and analyzed to obtain the pattern index of each bag. The specific process is as follows: read the image data containing the bag pattern of each bag and perform grayscale processing; train the image data containing the bag pattern of each bag after grayscale processing based on the convolutional neural network learning model to obtain the pattern texture features and pattern edge features of each bag; normalize the pattern texture features and pattern edge features of each bag, and perform weighted sum calculation based on the pattern texture features and pattern edge features of each bag after normalization to obtain the pattern index of each bag.

[0011] Furthermore, the image data containing the luggage pattern of each luggage after grayscale processing is trained based on the convolutional neural network learning model to obtain the pattern texture features and pattern edge features of each luggage. The specific process is as follows: the image data containing the luggage pattern of each luggage is loaded and input into the convolution layer of the convolutional neural network learning model; the image data containing the luggage pattern of each luggage is convolved based on the convolution kernel in the convolution layer to generate a feature map of the image data containing the luggage pattern of each luggage, wherein each pixel in the image data represents the feature response value of the position corresponding to the convolution kernel; the ReLU activation function is applied to the output of the convolution layer to set the negative value to zero; the feature map of the image data containing the luggage pattern of each luggage after ReLU activation is pooled to reduce the dimension of the feature map of the image data containing the luggage pattern of each luggage; the feature map output by the pooling layer is flattened as the input of the fully connected layer; prediction is performed in the output layer to output the prediction results of the edge features and texture features of the pattern.

[0012] Furthermore, the shape data of each bag is read and analyzed to obtain the shape index of each bag. The specific process is as follows: read the three-dimensional pixel image data set of each bag; based on the analysis of the three-dimensional pixel image data set of each bag, the number of pixel points on each surface of each bag is obtained, and marked as the area value of each surface of each bag; the area value of each surface of each bag is summed up and calculated to obtain the surface area value of each bag, and the surface area value is regarded as the shape index.

[0013] Furthermore, the specific process of obtaining the quality characteristic index of each bag is as follows: reading the weight value and size data of each bag, and analyzing and processing the size data of each bag to obtain the size index of each bag; normalizing the weight value and size index of each bag, and performing weighted summation based on the normalized weight value and size index of each bag to obtain the quality characteristic index of each bag, and the calculation formula is as follows: Among them, ZT a is the quality characteristic index of the a-th bag, ZL a is the weight value of the ath bag after normalization, ZL′ is the set weight reference value of the bag, CZ a is the size index of the ath bag after normalization, CZ′ is the set reference size index of the bag, β1 is the proportional coefficient of the set weight value, β2 is the proportional coefficient of the set size index, β1+β2=1, a=1,2,3,…,A, A is the total number of bags produced.

[0014] Furthermore, the size data of each bag is analyzed and processed to obtain the size index of each bag as follows: read the three-dimensional pixel coordinate image data set of each bag and perform preprocessing; perform edge detection on the three-dimensional image data of each bag after preprocessing based on Canny edge detection to obtain the boundary information of each surface of each bag, and the boundary information is specifically a set of position coordinates of edge pixel points of each surface of the bag; analyze and compare the boundary information of each surface of each bag respectively to obtain the pixel points with the smallest and largest horizontal and vertical coordinates in the boundary information of each surface of each bag; and record the difference in horizontal coordinates of the pixel points with the smallest and largest horizontal coordinates of each surface of each bag as the initial length value of each surface; record the difference in vertical coordinates of the pixel points with the smallest and largest vertical coordinates of each surface of each bag as the initial width value of each surface; and judge the size of the initial length value and the initial width value of each surface of each bag respectively. If the initial length value is smaller than the initial width value, mark the initial length value as the actual width value of the corresponding surface. , the initial width value is marked as the actual length value of the corresponding surface. If the initial length value is greater than the initial width value, the initial length value is marked as the actual length value of the corresponding surface, and the initial width value is marked as the actual width value of the corresponding surface; and the actual length value and the actual width value of each surface of each bag are compared with each other, and the maximum actual length value of each bag is marked as the length value of the corresponding bag, and the maximum actual width value is marked as the width value of the corresponding bag; the Z axis of each bag is marked as the height direction axis, and the boundary information of each surface of each bag is analyzed and compared respectively to obtain the pixel points with the minimum and maximum Z axis coordinates in the boundary information of each surface of each bag; and the Z axis coordinate difference of the pixel points with the minimum and maximum Z axis coordinates of each surface of each bag is recorded as the initial height value of each surface; the initial height values ​​of each surface of each bag are compared with each other, and the maximum initial height value is marked as the height value of each bag; the length value, width value, and height value of each bag are read and weighted to obtain the size index of each bag.

[0015] Furthermore, the specific process of judging whether the label data of each bag meets the bag production standards respectively and marking the label data of bags that do not meet the bag production standards is as follows: judging whether the appearance feature indicators and quality feature indicators of each bag meet the set appearance feature threshold indicators and quality feature threshold indicators respectively; if there are bags whose appearance feature indicators do not meet the set appearance feature threshold indicators, judging whether the color indicators, pattern indicators and shape indicators of the bags meet the set color threshold indicators, pattern threshold indicators and shape threshold indicators respectively, and marking bags whose color indicators do not meet the color threshold indicators as the first production fault of the appearance category, bags whose pattern indicators do not meet the pattern threshold indicators as the second production fault of the appearance category, bags whose color indicators do not meet the shape threshold indicators as the third production fault of the appearance category, and bags whose color indicators and pattern indicators do not meet the set color threshold indicators and pattern threshold indicators as the first production fault of the appearance category. If the luggage, color index and shape index do not meet the set color threshold index and shape threshold index, it will be marked as the second production fault of the second category in appearance; if the luggage, pattern index and shape index do not meet the set pattern threshold index and shape threshold index, it will be marked as the third production fault of the third category in appearance; if only the quality characteristic index of the luggage does not meet the set quality characteristic threshold index, then judge whether the weight index and size index of the luggage meet the set weight threshold index and size threshold index respectively, and mark the luggage whose weight index does not meet the weight threshold index as the first production fault of the first quality category, mark the luggage whose size index does not meet the size threshold index as the second production fault of the second quality category, and mark the luggage whose weight index and size index do not meet the set weight threshold index and size threshold index as the double production fault of quality; if the appearance characteristic index and quality characteristics of the luggage do not meet the set appearance characteristic threshold index and quality characteristic threshold index, then the luggage will be directly marked as a double production fault.

[0016] The present invention has the following beneficial effects:

[0017] (1) The luggage production data recognition and processing system can automatically determine whether the appearance characteristic indicators and quality characteristic indicators of each luggage meet the set production standards through the judgment subsystem, thereby realizing automated production standard inspection, which can greatly improve production efficiency, reduce the time and cost required for manual inspection, and reduce the possibility of subjective errors.

[0018] (2) The luggage production data identification and processing system can mark luggage that does not meet production standards based on the results of the judgment subsystem and classify them according to different fault types. This accurate fault identification and classification helps production managers to discover and solve problems in a timely manner, thereby improving the efficiency and reliability of the production line.

[0019] (3) The luggage production data identification and processing system can conduct a comprehensive analysis of the appearance data and quality data of each luggage through the data analysis subsystem to obtain appearance characteristic indicators and quality characteristic indicators. These indicators reflect the appearance and quality of the luggage, which helps production management personnel to monitor and control the production process, discover potential problems in a timely manner and take measures, thereby improving the quality control level.

[0020] Of course, any product implementing the present invention does not necessarily need to achieve all of the advantages described above at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 This is a block diagram of a bag production data identification and processing system of the present invention.

[0022] Figure 2 This is a flowchart of the specific steps of obtaining the appearance feature index of each bag in a bag production data identification and processing system of the present invention.

[0023] Figure 3 This is a flowchart of the specific steps for obtaining the color index of each bag in a bag production data identification and processing system of the present invention.

[0024] Figure 4 The present invention is a flowchart of the specific steps for obtaining the pattern index of each bag in a bag production data recognition and processing system.

[0025] Figure 5 This is a flowchart of the specific steps for obtaining the shape index of each bag in a bag production data recognition and processing system of the present invention.

[0026] Figure 6 This is a flowchart of the specific steps for obtaining the quality characteristic index of each bag in a bag production data identification and processing system of the present invention. DETAILED DESCRIPTION

[0027] The embodiment of the present application solves the problem that in traditional production management, the analysis of luggage production data often requires manual operation, resulting in low efficiency. At the same time, since the compliance with production standards often depends on manual observation and judgment, subjective deviations are prone to occur.

[0028] The overall idea of ​​the problem in the embodiment of this application is as follows:

[0029] First, the data acquisition subsystem is responsible for obtaining the number of bags and suitcases from production, as well as the label information, appearance data and quality data of each bag and suitcase. Secondly, the data analysis subsystem analyzes and processes the appearance data and quality data of each bag and obtains appearance characteristic indicators and quality characteristic indicators. Then, the judgment subsystem judges whether the label data of each bag and suitcase meets the standards according to the set production standards, and marks the bags and suitcases that do not meet the standards. Finally, the classification processing subsystem classifies the marked bags and suitcases to further process or correct problems encountered in production, thereby improving production efficiency and product quality.

[0030] See also Figure 1 The embodiment of the present invention provides a technical solution: a bag production data identification and processing system, comprising: a data acquisition subsystem, a data analysis subsystem, a judgment subsystem, and a classification processing subsystem; the data acquisition subsystem is used to acquire production data, the production data includes the number of bags produced and the label information, appearance data, and quality data of each bag; the data analysis subsystem is used to extract the appearance data and quality data of each bag respectively, and conduct a comprehensive analysis based on the extraction results to obtain the appearance feature index and quality feature index of each bag, and fit the label information, appearance feature index, and quality feature index of each bag to obtain the label data of each bag; the judgment subsystem is used to judge whether the label data of each bag meets the bag production standard respectively, and mark the label data of bags that do not meet the bag production standard; the classification processing subsystem is used to identify bags with marked label data and classify and transport them.

[0031] The label information of each bag is specifically the production serial number information of each bag, the appearance data is specifically the color data of the bag, the pattern data is specifically the image data and shape data containing the bag pattern, the color data is specifically the pixel value of each pixel point of the bag, the shape data is specifically the three-dimensional pixel image data set of the pixel point information data of each surface of the bag, the quality data is specifically the weight value and size data of the bag, the size data is specifically the pixel point information data of each surface of the bag and the three-dimensional pixel coordinate image data set of the position coordinates of each pixel point.

[0032] Specifically, Figure 2 As shown, the specific process of obtaining the appearance characteristic index of each bag is as follows: read the color data, pattern data, and shape data of each bag, and analyze and process them respectively to obtain the color index, pattern index, and shape index of each bag; normalize the color index, pattern index, and shape index of each bag, and perform weighted summation based on the color index, pattern index, and shape index of each bag after normalization to obtain the appearance characteristic index of each bag. The calculation formula is as follows: Among them, WG ais the appearance characteristic index of the a-th bag, YS a is the color index of the a-th bag after normalization, YS′ is the color reference index of the bag, TZ a is the pattern index of the ath bag after normalization, TZ′ is the set pattern reference index of the bag, XZ a is the shape index of the ath bag after normalization, XZ′ is the set shape reference index of the bag, α1 is the proportional coefficient of the set color index, α2 is the proportional coefficient of the set pattern index, α3 is the proportional coefficient of the set shape index, α1+α2+α3=1, a=1,2,3,…,A, A is the total number of bags produced.

[0033] In this implementation scheme, normalization processing is used to ensure that different characteristic indicators are at the same level, which is conducive to comprehensive analysis of the appearance characteristics of luggage. By weighted summation based on the set color reference indicator of the luggage, the set pattern reference indicator of the luggage, the set shape reference indicator of the luggage and the normalized data, the influence of different appearance characteristics on the final result can be relatively balanced, thereby improving the flexibility and applicability of the system. This process realizes the automated calculation of the appearance characteristic indicators of luggage, reduces the time and cost of manual processing, and improves production efficiency. By comprehensively considering multiple appearance characteristic indicators such as color, pattern, and shape, the appearance quality of luggage can be more comprehensively evaluated, which is helpful to improve the product quality on the production line. By setting reference indicators and proportional coefficients, the calculation process of appearance characteristic indicators is more standardized and controllable, which is conducive to the formulation and implementation of unified production standards and improving product consistency and reliability.

[0034] Specifically, Figure 3 As shown, the specific process of reading the color data of each bag and analyzing and processing to obtain the color index of each bag is as follows: read the pixel value of each pixel of each bag, and the pixel value is expressed in the form of RGB value; perform color space conversion on the color data of each bag to obtain the HSV value of each pixel of each bag; screen each pixel of each bag based on the set color interval; count the number of pixels in each color interval for each bag; establish feature vectors based on the number of pixel values ​​corresponding to each color interval of each bag; perform weighted calculation on the feature vectors of each color interval of each bag to obtain the color index of each bag; the specific steps of color space conversion on the color data of each bag are as follows: compare the pixel values ​​of each pixel in the color data of each bag respectively, and obtain the maximum RGB component and the minimum RGB component of each bag, and mark them as Max respectively. (R,G,B) and Min (R,G,B) ; Calculate hue H: If Max (R,G,B)Equal to Min (R,G,B) , then H is 0, which means gray; otherwise, if Max (R,G,B) is equal to R, and G is greater than or equal to B, then H is calculated as follows: Where G is the green component and B is the blue component; if Max (R,G,B) If G is equal to R and G is less than B, then H is calculated as follows: If Max (R,G,B) Equal to G, then H is calculated by the following formula: Where R is the red component; if Max (R,G,B) If B is equal to B, then H is calculated by the following formula: Calculate saturation S: If Max (R,G,B) If it is equal to 0, then S is 0; otherwise, S is calculated by the following formula: Calculate the brightness V: V is Max (R,G,B) .

[0035] In this implementation, RGB values ​​are converted into HSV values ​​through color space conversion, which can further enhance the way of color perception, make it closer to the way humans perceive color, and improve the accuracy of color recognition. By more detailed classification and judgment of colors according to actual needs and set color ranges, the color index is made more accurate. By establishing feature vectors and weighted calculations, the calculation process of color index is made more standardized and controllable, which is conducive to improving the stability and maintainability of the system.

[0036] Specifically, Figure 4 As shown, the specific process of reading the pattern data of each bag and analyzing and processing to obtain the pattern index of each bag is as follows: reading the image data containing the bag pattern of each bag and performing grayscale processing; training the image data containing the bag pattern of each bag after the grayscale processing based on the convolutional neural network learning model to obtain the pattern texture features and pattern edge features of each bag; normalizing the pattern texture features and pattern edge features of each bag, and performing weighted sum calculation based on the pattern texture features and pattern edge features of each bag after the normalization processing to obtain the pattern index of each bag.

[0037] The convolutional neural network learning model includes: an input layer for receiving image data containing bag patterns of bags, a convolution layer for extracting local features of the image, a ReLU activation layer for introducing nonlinearity and increasing the expressive power of the model, a pooling layer for reducing the dimension of the feature map, reducing the amount of calculation, and retaining important features, a fully connected layer for converting the feature map output by the convolution layer into a prediction result, and an output layer for outputting the prediction result of the model. In this embodiment, the image data containing the bag pattern of each bag is read and grayscale processed to convert the color image into a grayscale image, thereby simplifying the processing process and reducing the amount of calculation. Then, a convolutional neural network learning model is established to train the image data containing the bag pattern of each bag to extract the texture features and edge features of the pattern. Then, the image data set is divided into a training set and a validation set for training the model and evaluating the model performance. Then, the established convolutional neural network is tested on the training set. The convolutional neural network learning model is trained, and the model parameters are updated through the back-propagation algorithm so that the model can more accurately predict the texture features and edge features of the pattern. Then the validation set is used to evaluate the model performance, and the model parameters are adjusted according to the validation results until the pattern texture features and edge features predicted by the model are consistent with the actual pattern texture features and edge features. This process realizes the automatic recognition and analysis of luggage patterns, reduces the need for manual intervention, and improves the efficiency and accuracy on the production line. By using the convolutional neural network model, the texture features and edge features of the pattern can be extracted quickly and accurately, greatly improving the processing speed. By training and adjusting the model parameters, the accuracy of pattern recognition can be adjusted and optimized according to actual needs, making the system more flexible and customizable. The use of deep learning technologies such as convolutional neural networks improves the accuracy and efficiency of pattern recognition, and promotes the application and development of related technologies in the field of production data recognition and processing.

[0038] Specifically, the image data containing the luggage pattern of each luggage after grayscale processing is trained based on the convolutional neural network learning model to obtain the pattern texture features and pattern edge features of each luggage. The specific process is as follows: the image data containing the luggage pattern of each luggage is loaded and input into the convolution layer of the convolutional neural network learning model; in the convolution layer, the image data containing the luggage pattern of each luggage is convolved based on the convolution kernel to generate a feature map of the image data containing the luggage pattern of each luggage, wherein each pixel in the image data represents the feature response value of the position corresponding to the convolution kernel; this step causes the network to learn the edge features in the pattern, and the ReLU activation function is applied to the output of the convolution layer to convert the negative values ​​into positive ones. Set to zero and introduce nonlinearity. This step helps the network learn complex features better. Pooling operation is performed on the feature map of the image data containing the bag pattern of each bag after ReLU activation. Maximum pooling is usually used to reduce the dimension of the feature map of the image data containing the bag pattern of each bag. The maximum pooling operation will retain the maximum value in each pooling window. This step helps to reduce the amount of calculation and overfitting, and retain important features. Flatten the feature map output by the pooling layer as the input of the fully connected layer. The role of the fully connected layer is to combine the features extracted by the convolutional layer to obtain a more advanced feature representation. Prediction is performed at the output layer to output the prediction results of the edge features and texture features of the pattern.

[0039] In this implementation scheme, the image data containing the pattern of each bag is loaded and input into the convolution layer of the convolutional neural network learning model. This step provides the image data to the network for feature extraction. In the convolution layer, the image data of each bag is convolved using a convolution kernel to generate a feature map. Each pixel in the feature map represents the feature response value of the position corresponding to the convolution kernel. This step helps to learn the edge features in the pattern. The ReLU activation function is applied to the output of the convolution layer, and the negative values ​​are set to zero to introduce nonlinearity, which helps the network to better learn complex features. The feature map after ReLU activation is pooled, and maximum pooling is usually used to retain the maximum value in each pooling window, reduce the dimension of the feature map, reduce the amount of calculation and overfitting, and retain important features. The pooling layer The output feature map is flattened as the input of the fully connected layer. The fully connected layer combines the features extracted by the convolution layer to obtain a more advanced feature representation, and makes predictions in the output layer to output the prediction results of the edge features and texture features of the pattern, completing the extraction and prediction of the pattern texture features and edge features of each bag. Through the convolutional neural network, the texture features and edge features in the pattern can be automatically extracted without manually designing the feature extractor, which simplifies the feature engineering process. The introduction of the ReLU activation function enables the network to learn more complex features, improves the network's expression ability and the extraction effect of pattern features, and reduces the dimension of the feature map through pooling operations, reducing the amount of calculation and the risk of overfitting. The fully connected layer combines the extracted features to obtain a more advanced feature representation, which improves the expression ability of the features.

[0040] Specifically, Figure 5 As shown, the shape data of each bag is read and analyzed to obtain the shape index of each bag. The specific process is as follows: read the three-dimensional pixel image data set of each bag; based on the analysis of the three-dimensional pixel image data set of each bag, the number of pixel points on each surface of each bag is obtained, and marked as the area value of each surface of each bag; the area value of each surface of each bag is summed up to obtain the surface area value of each bag, and the surface area value is regarded as the shape index.

[0041] In this embodiment, by reading the three-dimensional pixel image data set of each bag, these data sets contain the pixel representation of the bag in three-dimensional space, record the shape information of the bag, then analyze the number of pixel points of each face of each bag based on the three-dimensional pixel image data set of each bag, and regard it as the area value of the face. This step converts the shape information of the bag into a specific area value, and finally sums up the area value of each face of each bag to obtain the surface area value of each bag. By adding the area value of each face, the surface area of ​​the entire bag is obtained, so that the surface area is used as a shape index. By reading the three-dimensional pixel image data set and analyzing and processing, the surface area value of each bag is obtained as a shape index, and the objective quantification of the shape of the bag is realized. The whole process is automated, does not require manual intervention, improves processing efficiency, and reduces the error rate in the processing process. Using the surface area as a shape index makes the comparison and analysis of the shape more concise and consistent, which is conducive to subsequent data processing and decision making.

[0042] Specifically, Figure 6 As shown, the specific process of obtaining the quality characteristic index of each bag is as follows: read the weight value and size data of each bag, and analyze and process the size data of each bag to obtain the size index of each bag; normalize the weight value and size index of each bag, and perform weighted summation based on the normalized weight value and size index of each bag to obtain the quality characteristic index of each bag. The calculation formula is as follows: Among them, ZT a is the quality characteristic index of the a-th bag, ZL a is the weight value of the ath bag after normalization, ZL′ is the set weight reference value of the bag, CZ a is the size index of the ath bag after normalization, CZ′ is the set reference size index of the bag, β1 is the proportional coefficient of the set weight value, β2 is the proportional coefficient of the set size index, β1+β2=1, a=1,2,3,…,A, A is the total number of bags produced.

[0043] In this implementation, the quality characteristic index of each bag is obtained through comprehensive analysis of the weight value and size data, which can objectively evaluate the overall quality of the bag. The normalization and weighted summation methods are used to convert bags of different sizes and weights into unified quality characteristic indexes, which is conducive to subsequent comparison and analysis. This process provides quantitative indicators of bag quality, which helps manufacturers and sellers make decisions based on data, thereby optimizing product design, manufacturing and sales strategies.

[0044] Specifically, the size data of each bag is analyzed and processed to obtain the size index of each bag as follows: read the three-dimensional pixel coordinate image data set of each bag and perform preprocessing; perform edge detection on the three-dimensional image data of each bag after preprocessing based on Canny edge detection to obtain the boundary information of each surface of each bag, and the boundary information is specifically a set of position coordinates of edge pixel points of each surface of the bag; analyze and compare the boundary information of each surface of each bag respectively to obtain the pixel points with the smallest and largest horizontal and vertical coordinates in the boundary information of each surface of each bag; and record the difference between the horizontal coordinates of the pixel points with the smallest and largest horizontal coordinates of each surface of each bag as the initial length value of each surface; record the difference between the vertical coordinates of the pixel points with the smallest and largest vertical coordinates of each surface of each bag as the initial width value of each surface; respectively judge the size of the initial length value and the initial width value of each surface of each bag, and if the initial length value is smaller than the initial width value, mark the initial length value as the actual width value of the corresponding surface, The initial width value is marked as the actual length value of the corresponding surface. If the initial length value is greater than the initial width value, the initial length value is marked as the actual length value of the corresponding surface, and the initial width value is marked as the actual width value of the corresponding surface; and the actual length value and the actual width value of each surface of each bag are compared with each other, and the maximum actual length value of each bag is marked as the length value of the corresponding bag, and the maximum actual width value is marked as the width value of the corresponding bag; the Z axis of each bag is marked as the height direction axis, and the boundary information of each surface of each bag is analyzed and compared respectively to obtain the pixel points with the minimum and maximum Z axis coordinates in the boundary information of each surface of each bag; and the Z axis coordinate difference of the pixel points with the minimum and maximum Z axis coordinates of each surface of each bag is recorded as the initial height value of each surface; the initial height values ​​of each surface of each bag are compared with each other, and the maximum initial height value is marked as the height value of each bag; the length value, width value, and height value of each bag are read and weighted to obtain the size index of each bag.

[0045] In this implementation scheme, the three-dimensional pixel coordinate image data set of each bag is read and preprocessed, which may include denoising, smoothing, etc., to improve the effect and accuracy of subsequent processing, and then edge detection is performed on the preprocessed three-dimensional image data of each bag based on Canny edge detection to obtain the boundary information of each surface of each bag, which is the basis for the size analysis of the bag. The boundary information of each surface of each bag is then analyzed and compared to obtain the initial length and width values ​​of each surface, that is, the horizontal and vertical spans of the surface. Then, based on the initial length and width values, the actual length and width values ​​of each surface are judged to ensure that the length and width relationship of the bag is correctly matched. Then, based on the actual length and width values ​​of each surface of each bag, the actual length and width values ​​of each bag are compared to obtain the size of each bag. The length and width values ​​of the bag are obtained, and the Z-axis coordinate information of each face of each bag is obtained to calculate the height value of the bag. Then the length, width and height values ​​of each bag are read and weighted to obtain the size index of each bag. By performing edge detection and analysis on the three-dimensional image data, accurate data on the size of the bag can be obtained, which reduces human errors. The entire process is completed automatically without human intervention, which improves processing efficiency and saves labor costs. Not only the length and width of the bag are considered, but also the height information, and the size characteristics of the bag are comprehensively evaluated from multiple dimensions. This process provides an automated size measurement method based on image data, which provides fast and accurate size data for production and quality control, and helps to improve production efficiency and product quality.

[0046] Specifically, the specific process of judging whether the label data of each bag meets the bag production standards respectively, and marking the label data of bags that do not meet the bag production standards is as follows: judge whether the appearance feature indicators and quality feature indicators of each bag meet the set appearance feature threshold indicators and quality feature threshold indicators respectively; if there are bags whose only appearance feature indicators do not meet the set appearance feature threshold indicators, then judge whether the color indicators, pattern indicators, and shape indicators of the bags meet the set color threshold indicators, pattern threshold indicators, and shape threshold indicators respectively, and mark the bags whose color indicators do not meet the color threshold indicators as the first production fault of the appearance category, the bags whose pattern indicators do not meet the pattern threshold indicators as the second production fault of the appearance category, the bags whose color indicators do not meet the shape threshold indicators as the third production fault of the appearance category, and the bags whose color indicators and pattern indicators do not meet the set color threshold indicators and pattern threshold indicators as the first production fault of the appearance category. If the luggage, color index and shape index do not meet the set color threshold index and shape threshold index, it will be marked as the second production fault of the second category in appearance; if the luggage, pattern index and shape index do not meet the set pattern threshold index and shape threshold index, it will be marked as the third production fault of the third category in appearance; if only the quality characteristic index of the luggage does not meet the set quality characteristic threshold index, then judge whether the weight index and size index of the luggage meet the set weight threshold index and size threshold index respectively, and mark the luggage whose weight index does not meet the weight threshold index as the first production fault of the first quality category, mark the luggage whose size index does not meet the size threshold index as the second production fault of the second quality category, and mark the luggage whose weight index and size index do not meet the set weight threshold index and size threshold index as the double production fault of quality; if the appearance characteristic index and quality characteristics of the luggage do not meet the set appearance characteristic threshold index and quality characteristic threshold index, then the luggage will be directly marked as a double production fault.

[0047] In this implementation scheme, through this process, the system can automatically determine whether the appearance characteristics and quality characteristics of each bag meet the pre-set standards without the need for manual inspection one by one, thereby improving efficiency. Judgment based on pre-set threshold indicators can ensure the accuracy and consistency of the judgment, avoid errors that may be caused by human subjective judgment, and classify and mark bags that do not meet the standards according to different types, such as single-type production failures and double-type production failures, which helps to take corresponding processing and improvement measures for different types of problems, discover problematic bags in production as early as possible, and mark them in time, which helps to improve the efficiency and timeliness of production quality control and prevent unqualified products from entering the market. The marked data can be used for subsequent data analysis to discover potential production problems and improve production processes, further improving production efficiency and product quality.

[0048] In summary, this application has at least the following effects:

[0049] The judgment subsystem can automatically determine whether the appearance characteristic indicators and quality characteristic indicators of each bag meet the set production standards, thereby realizing automated production standard inspection. This can greatly improve production efficiency, reduce the time and cost required for manual inspection, and reduce the possibility of subjective errors.

[0050] Based on the results of the judgment subsystem, bags that do not meet production standards can be marked and classified according to different fault types. This accurate fault identification and classification helps production managers to discover and solve problems in a timely manner, improving the efficiency and reliability of the production line.

[0051] The data analysis subsystem can conduct a comprehensive analysis of the appearance data and quality data of each bag to obtain appearance characteristic indicators and quality characteristic indicators. These indicators reflect the appearance and quality of the bags, which helps production managers monitor and control the production process, discover potential problems in a timely manner and take measures, thereby improving the level of quality control.

[0052] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0053] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A bag production data identification and processing system, characterized in that: include: Data acquisition subsystem, data analysis subsystem, judgment subsystem, classification processing subsystem; The data acquisition subsystem is used to acquire production data, wherein the production data includes the quantity of bags produced and label information, appearance data, and quality data of each bag; The data analysis subsystem is used to extract the appearance data and quality data of each bag respectively, and perform comprehensive analysis based on the extraction results to obtain the appearance characteristic index and quality characteristic index of each bag, and fit the label information, appearance characteristic index and quality characteristic index of each bag to obtain the label data of each bag; The judging subsystem is used to judge whether the label data of each bag meets the bag production standard, and mark the label data of bags that do not meet the bag production standard; The specific process is as follows: Determine whether the appearance characteristic index and quality characteristic index of each bag meet the set appearance characteristic threshold index and quality characteristic threshold index respectively; If there are bags whose only appearance feature indicators do not meet the set appearance feature threshold indicators, then determine whether the color indicators, pattern indicators, and shape indicators of the bags meet the set color threshold indicators, pattern threshold indicators, and shape threshold indicators respectively, and mark the bags whose color indicators do not meet the color threshold indicators as the first production fault of the appearance category, the bags whose pattern indicators do not meet the pattern threshold indicators as the second production fault of the appearance category, the bags whose color indicators do not meet the shape threshold indicators as the third production fault of the appearance category, the bags whose color indicators and pattern indicators do not meet the set color threshold indicators and pattern threshold indicators as the first production fault of the appearance category, the bags whose color indicators and shape indicators do not meet the set color threshold indicators and shape threshold indicators as the second production fault of the appearance category, and the bags whose pattern indicators and shape indicators do not meet the set pattern threshold indicators and shape threshold indicators as the third production fault of the appearance category; If only the quality characteristic index of the luggage does not meet the set quality characteristic threshold index, then determine whether the weight index and size index of the luggage meet the set weight threshold index and size threshold index respectively, and mark the luggage whose weight index does not meet the weight threshold index as the first quality single-category production failure, the luggage whose size index does not meet the size threshold index as the second quality single-category production failure, and the luggage whose weight index and size index do not meet the set weight threshold index and size threshold index as the double quality production failure; If the appearance characteristic index and quality characteristic of the luggage do not meet the set appearance characteristic threshold index and quality characteristic threshold index, the luggage will be directly marked as a double-category production failure; The classification processing subsystem is used to identify bags with label data and classify and transport them; The label information of each bag is specifically the production serial number information of each bag, the appearance data is specifically the color data of the bag, the pattern data is specifically the image data and shape data containing the bag pattern, the color data is specifically the pixel value of each pixel point of the bag, the shape data is specifically the three-dimensional pixel image data set of the pixel point information data of each surface of the bag, the quality data is specifically the weight value and size data of the bag, the size data is specifically the pixel point information data of each surface of the bag and the three-dimensional pixel coordinate image data set of the position coordinates of each pixel point; The specific process of obtaining the appearance characteristic index of each bag is as follows: Read the color data, pattern data, and shape data of each bag, and analyze and process them respectively to obtain the color index, pattern index, and shape index of each bag; The color index, pattern index, and shape index of each bag are normalized, and the weighted sum of the normalized color index, pattern index, and shape index of each bag is performed to obtain the appearance feature index of each bag. The calculation formula is as follows: Among them, WG a is the appearance characteristic index of the a-th bag, YS a is the color index of the ath bag after normalization, YS′ is the color reference index of the bag, TZ a is the pattern index of the a-th bag after normalization, TZ′ is the set pattern reference index of the bag, XZ a is the shape index of the ath bag after normalization, XZ′ is the set shape reference index of the bag, α1 is the set color index proportional coefficient, α2 is the set pattern index proportional coefficient, α3 is the set shape index proportional coefficient, α1+α2+α3=1, a=1,2,3,…,A, A is the total number of bags produced; The specific process of reading the pattern data of each bag and analyzing and processing it to obtain the pattern index of each bag is as follows: Read the image data of each bag including the bag pattern and perform grayscale processing; Based on the convolutional neural network learning model, the image data containing the bag pattern of each bag after grayscale processing is trained to obtain the pattern texture features and pattern edge features of each bag; Normalizing the pattern texture features and pattern edge features of each bag, and performing weighted sum calculation based on the pattern texture features and pattern edge features of each bag after the normalization process to obtain the pattern index of each bag; The construction process of the convolutional neural network learning model is as follows: obtain a number of image data containing luggage patterns to be processed and establish an image data set; divide the image data set into an image training set and an image verification set; read the image training data set to train the constructed convolutional neural network learning model and calculate the image loss function; use the image verification data set to evaluate the model performance, and adjust the model parameters according to the verification results until the pattern texture features and pattern edge features predicted by the model are consistent with the actual pattern texture features and pattern edge features; Based on the convolutional neural network learning model, the image data containing the bag pattern of each bag after grayscale processing is trained to obtain the pattern texture features and pattern edge features of each bag as follows: Loading image data containing bag patterns of each bag into the convolutional layer of the convolutional neural network learning model; Performing a convolution operation on the image data containing the bag pattern of each bag based on the convolution kernel in the convolution layer to generate a feature map of the image data containing the bag pattern of each bag, wherein each pixel in the image data represents a feature response value at a position corresponding to the convolution kernel; Apply the ReLU activation function to the output of the convolutional layer to set negative values ​​to zero; A pooling operation is performed on the feature map of the image data containing the bag pattern of each bag after ReLU activation to reduce the dimension of the feature map of the image data containing the bag pattern of each bag; Flatten the feature map output by the pooling layer and use it as the input of the fully connected layer; Prediction is performed at the output layer, and the prediction results of the edge features and texture features of the pattern are output; The specific process of obtaining the quality characteristic indicators of each bag is as follows: Read the weight and size data of each bag, and analyze and process the size data of each bag to obtain the size index of each bag; The weight value and size index of each bag are normalized, and the weight value and size index of each bag after normalization are weighted and summed to obtain the quality characteristic index of each bag. The calculation formula is as follows: Among them, ZT a is the quality characteristic index of the a-th bag, ZL a is the weight value of the ath bag after normalization, ZL′ is the set weight reference value of the bag, CZ a is the size index of the ath bag after normalization, CZ′ is the set reference size index of the bag, β1 is the proportional coefficient of the set weight value, β2 is the proportional coefficient of the set size index, β1+β2=1, a=1,2,3,…,A, A is the total number of bags produced.

2. The bag production data identification and processing system according to claim 1, characterized in that: The specific process of reading the color data of each bag and analyzing and processing it to obtain the color index of each bag is as follows: Read the pixel value of each pixel point of each bag, and the pixel value is expressed in the form of RGB value; Perform color space conversion on the color data of each bag to obtain the HSV value of each pixel of each bag; Filter each pixel of each bag based on the set color range; For each bag, count the number of pixels in each color interval; Establish feature vectors based on the pixel values ​​corresponding to each color interval of each bag; The characteristic vectors of each color interval of each bag are weighted and calculated respectively to obtain the color index of each bag.

3. The bag production data identification and processing system according to claim 1, characterized in that: The specific process of reading the shape data of each bag and analyzing and processing it to obtain the shape index of each bag is as follows: Read the 3D pixel image dataset of each bag; Based on the analysis of the three-dimensional pixel image data set of each bag, the number of pixel points on each surface of each bag is obtained, and marked as the area value of each surface of each bag; The area value of each face of each bag is summed up and calculated to obtain the surface area value of each bag, and the surface area value is regarded as the shape index.

4. The bag production data identification and processing system according to claim 1, characterized in that: The specific process of analyzing and processing the size data of each bag and obtaining the size index of each bag is as follows: Read the three-dimensional pixel coordinate image dataset of each bag and perform preprocessing; Performing edge detection on the three-dimensional image data of each bag after preprocessing based on Canny edge detection to obtain boundary information of each surface of each bag, wherein the boundary information is specifically a position coordinate set of edge pixel points of each surface of the bag; Analyze and compare the boundary information of each surface of each bag respectively, and obtain the pixel points with the smallest and largest horizontal and vertical coordinates in the boundary information of each surface of each bag; The difference between the horizontal coordinates of the pixel point with the smallest horizontal coordinate and the pixel point with the largest horizontal coordinate on each surface of each bag is recorded as the initial length value of each surface; The difference between the ordinate of the pixel point with the smallest ordinate and the pixel point with the largest ordinate on each face of each bag is recorded as the initial width value of each face; Determine the initial length value and initial width value of each surface of each bag respectively. If the initial length value is smaller than the initial width value, mark the initial length value as the actual width value of the corresponding surface, and mark the initial width value as the actual length value of the corresponding surface. If the initial length value is larger than the initial width value, mark the initial length value as the actual length value of the corresponding surface, and mark the initial width value as the actual width value of the corresponding surface. The actual length value and the actual width value of each surface of each bag are compared with each other, and the maximum actual length value of each bag is marked as the length value of the corresponding bag, and the maximum actual width value is marked as the width value of the corresponding bag; Mark the Z axis of each bag as the height axis, and analyze and compare the boundary information of each surface of each bag to obtain the pixel points with the minimum and maximum Z axis coordinates in the boundary information of each surface of each bag; The Z-axis coordinate difference between the pixel point with the minimum Z-axis coordinate and the pixel point with the maximum Z-axis coordinate on each surface of each bag is recorded as the initial height value of each surface; Compare the initial height values ​​of each surface of each bag with each other, and mark the maximum initial height value as the height value of each bag; The length, width and height of each bag are read and weighted to obtain the size index of each bag.

Images