A method and system for detecting quality of firework inner tubes used in firework production

Abstract

The present invention discloses a method and system for detecting the quality of fireworks inner tubes used in fireworks production, and relates to the technical field of fireworks inner tube quality detection. The method for detecting the quality of fireworks inner tubes used in fireworks production obtains image data of a plurality of image channels of the fireworks inner tube to be detected, performs comprehensive analysis, obtains comprehensive image data of the fireworks inner tube to be detected, and inputs the image data into a pre-trained neural network model for defect recognition analysis, and performs comprehensive analysis after defect recognition analysis to obtain the image quality index of the fireworks inner tube to be detected, and performs judgment analysis. If the image quality index meets the expected standard, the comprehensive quality strength index of the fireworks inner tube is calculated, and the judgment analysis is performed again. If the image quality index does not meet the expected standard, it is marked as unqualified. The present invention incorporates the appearance defects of the fireworks inner tube and the physical properties of the inner tube into the quality evaluation system, so as to more comprehensively and accurately evaluate the quality of the fireworks inner tube.

Description

Technical Field

[0001] The present invention relates to the technical field of quality detection of firework inner tubes, and in particular to a quality detection method and system for firework inner tubes used in firework production. Background Art

[0002] Fireworks are a common festive entertainment product, widely used in various celebrations, festivals, and public performances. With the advancement of technology and continuous improvements in production processes, modern fireworks are not only becoming increasingly diverse in appearance and effects, but their safety and quality standards have also become key concerns in the industry's development. The quality of fireworks directly affects the display effect, viewing experience, and safety. The quality of the fireworks' inner tube, in particular, is crucial to the overall performance and safety of the fireworks.

[0003] The inner tube of a firework is the container used to hold gunpowder in fireworks. Its quality requirements include material density, wall thickness, hardness, and overall structural integrity. During the firing process, the inner tube will be subjected to various mechanical effects such as internal pressure, external impact, heat, and vibration. If its quality is unqualified, it may cause explosions or poor fireworks effects, and even cause safety accidents. In order to ensure the safety and reliability of fireworks, quality inspection of the inner tube of fireworks is particularly important.

[0004] Existing methods for inspecting the quality of fireworks inner tubes mostly rely on manual inspection and visual testing, relying primarily on the operator's experience to judge the surface condition and external defects of the fireworks inner tubes. However, this method has certain limitations and is easily affected by human factors. It is also difficult to accurately inspect the internal structure, mechanical properties, and potential minor defects of the fireworks inner tubes. With the increasing complexity of fireworks manufacturing processes and the improvement of quality requirements, traditional manual inspection methods can no longer meet the accuracy and efficiency requirements of modern production.

[0005] The limitations of the existing technology include at least the following problems. First, the existing quality inspection method of firework inner tubes relies too much on single image data for defect detection. Although this method can detect surface defects such as holes and cracks, it is difficult to comprehensively evaluate the overall quality and strength of the firework inner tube. Since the quality of the firework inner tube depends not only on surface defects, but also involves multiple factors such as its internal material properties and structural strength, a single image inspection cannot fully reflect its true structural health. Therefore, the existing technology is difficult to accurately predict the damage or failure that may occur in the actual firing process of the firework inner tube, which poses a major safety hazard. Secondly, the existing inspection methods usually only use local or limited physical parameters for evaluation, and it is difficult to accurately reflect the overall quality of the firework inner tube through comprehensive and detailed multi-dimensional data comprehensive analysis. As a result, some firework inner tubes that meet the surface requirements may become structurally unstable or fail in actual use due to the failure to consider the influence of all key parameters, thereby affecting the safety and quality of the fireworks. Summary of the Invention

[0006] In response to the shortcomings of the existing technology, the present invention provides a method and system for detecting the quality of firework inner tubes for use in firework production. This solves the problem that the existing firework inner tube quality detection method relies too much on single image data for defect detection. Although this method can detect surface defects such as holes and cracks, it is difficult to comprehensively evaluate the overall quality and strength of the firework inner tube. Because the quality of a firework inner tube depends not only on surface defects but also involves multiple factors such as its internal material properties and structural strength, a single image detection cannot fully reflect its true structural health. Therefore, the existing technology has difficulty in accurately predicting damage or failure that may occur to the firework inner tube during the actual firing process, thereby posing a major safety hazard. Secondly, the existing detection methods usually only use local or limited physical parameters for evaluation, and it is difficult to accurately reflect the overall quality of the firework inner tube through comprehensive and detailed multi-dimensional data comprehensive analysis. As a result, some firework inner tubes that meet surface requirements may become structurally unstable or fail in actual use due to the failure to consider the influence of all key parameters, thereby affecting the safety and quality of the fireworks.

[0007] To achieve the above objectives, the present invention is implemented through the following technical solutions: A method for detecting the quality of firework inner tubes for firework production, comprising the following steps: acquiring image data of several image channels of the firework inner tube to be detected, and performing comprehensive analysis to obtain comprehensive image data of the firework inner tube to be detected, wherein the image data is specifically the pixel value of each pixel point in the firework inner tube image, and the comprehensive image data is specifically the comprehensive pixel value of each pixel point in the comprehensive image of the firework inner tube; inputting the comprehensive image data of the firework inner tube to be detected into a pre-trained neural network model for defect recognition analysis to obtain several hole areas and several crack areas in the comprehensive image of the firework inner tube of the firework inner tube to be detected, and performing comprehensive analysis to obtain the image quality index of the firework inner tube to be detected; judging whether the image quality index of the firework inner tube to be detected is within a preset image quality threshold range; if the image quality index of the firework inner tube to be detected is within a preset image quality threshold range, If the image quality index of the firework inner tube to be tested is outside the preset image quality threshold range, the firework inner tube to be tested is marked as unqualified; if the image quality index of the firework inner tube to be tested is within the preset image quality threshold range, the physical feature data of the firework inner tube to be tested is obtained, and a comprehensive analysis is performed to obtain the comprehensive quality strength index of the firework inner tube to be tested, wherein the physical feature data includes the material density value, wall thickness value, material hardness value, surface roughness value, surface temperature value, strain value, vibration frequency value, and surface smoothness value of each measurement position point; it is judged whether the comprehensive quality strength index of the firework inner tube to be tested is within the preset comprehensive quality strength threshold range; if the comprehensive quality strength index of the firework inner tube to be tested is within the preset comprehensive quality strength threshold range, the firework inner tube to be tested is marked as qualified; if the comprehensive quality strength index of the firework inner tube to be tested is outside the preset comprehensive quality strength threshold range, the firework inner tube to be tested is marked as unqualified.

[0008] Furthermore, the specific formula for calculating the comprehensive pixel value of each pixel in the comprehensive image of the firework inner tube to be detected is as follows: ;in, The first image of the firework inner tube to be detected is The comprehensive pixel value of each pixel, The first The image of the inner tube of the firework in the image channel The pixel value of each pixel, The first The image of the inner tube of the firework in the image channel The weighting coefficient of the pixel value of each pixel, =1, 2, 3, ..., , is the number of pixels, =1, 2, 3, ..., , is the number of image channels.

[0009] Furthermore, the neural network model is specifically a convolutional neural network model, and the specific steps for obtaining several hole areas and several crack areas in the comprehensive image of the inner tube of the firework to be detected are as follows: adjusting and processing the comprehensive image data of the inner tube of the firework to be detected; converting the adjusted comprehensive image data of the inner tube of the firework to be detected into a tensor, and inputting it into a pre-trained convolutional neural network model for feature extraction; performing defect recognition and classification processing on the feature extraction results to obtain several hole areas and several crack areas in the comprehensive image of the inner tube of the firework to be detected.

[0010] Furthermore, the specific steps for obtaining the image quality index of the firework inner tube to be tested are as follows: counting the number of pixels in each hole area and each crack area in the comprehensive image of the firework inner tube of the firework inner tube to be tested, and using them as the area value of each hole area and the area value of each crack area in the comprehensive image of the firework inner tube of the firework inner tube to be tested; obtaining the cross-sectional area value of the firework inner tube to be tested, and performing a comprehensive analysis based on the area value of each hole area and the area value of each crack area in the comprehensive image of the firework inner tube of the firework inner tube to be tested, to obtain the image hole influence factor and the image crack influence factor of the firework inner tube to be tested; obtaining the reference bearing capacity value of the firework inner tube to be tested, and performing a comprehensive analysis based on the image hole influence factor and the image crack influence factor of the firework inner tube to be tested, to obtain the image quality index of the firework inner tube to be tested.

[0011] Furthermore, the specific formulas for calculating the image hole impact factor, image crack impact factor, and image quality index of the firework inner tube to be detected are as follows: ;in, is the image hole impact factor of the inner tube of the firework to be detected, The first image of the firework inner tube to be detected is The area value of the hole area, is the cross-sectional area of ​​the inner tube of the firework to be tested, is the image crack impact factor of the firework inner tube to be detected, =1, 2, 3, ..., , is the number of hole areas, is the crack influence coefficient stored in the database, is a natural constant, The first image of the firework inner tube to be detected is The length of the crack area, The first image of the firework inner tube to be detected is The area value of the crack region, =1, 2, 3, ..., , is the number of crack areas, is the image quality index of the inner tube of the firework to be detected, is the reference load-bearing capacity value of the firework inner tube to be tested.

[0012] Furthermore, the specific steps for obtaining the comprehensive quality strength index of the firework inner tube to be tested are as follows: reading the physical characteristic data of the firework inner tube to be tested and performing a comprehensive analysis to obtain the basic quality strength index and mechanical strength index of the firework inner tube to be tested; performing a comprehensive analysis on the basic quality strength index and mechanical strength index of the firework inner tube to be tested to obtain the comprehensive quality strength index of the firework inner tube to be tested; wherein, the specific formula for calculating the comprehensive quality strength index of the firework inner tube to be tested is as follows: ;in, is the comprehensive quality strength index of the firework inner tube to be tested, is the basic mass strength index of the firework inner tube to be tested, is the basic quality coefficient stored in the database, is the mechanical strength index of the inner tube of the firework to be tested, is the mechanical strength coefficient stored in the database, , is a natural constant.

[0013] Furthermore, the specific steps for obtaining the basic quality strength index of the inner tube of the firework to be tested are as follows: obtaining the material density reference value, wall thickness reference value, material hardness reference value, and surface roughness reference value of the inner tube of the firework to be tested; comprehensively analyzing the material density value, wall thickness value, material hardness value, and surface roughness value of each measuring position point of the inner tube of the firework to be tested in combination with the material density reference value, wall thickness reference value, material hardness reference value, and surface roughness reference value of the inner tube of the firework to be tested to obtain the basic quality strength index of the inner tube of the firework to be tested.

[0014] Furthermore, the specific formula for calculating the basic mass strength index of the firework inner tube to be tested is as follows: ;in, is the basic mass strength index of the firework inner tube to be tested, For each of the fireworks inner tubes to be tested The material density value at each measurement point, Reference value of the material density of the inner tube of the firework to be tested, is the material density coefficient stored in the database, For each of the fireworks inner tubes to be tested The wall thickness value of each measuring point, Reference value of the wall thickness of the inner tube of the fireworks to be tested, is the wall thickness coefficient stored in the database, For each of the fireworks inner tubes to be tested The material hardness value at each measuring point, is the reference value of the material hardness of the inner tube of the firework to be tested, is the material hardness coefficient stored in the database, For each of the fireworks inner tubes to be tested The surface roughness value of each measuring point, is the reference value of the surface roughness of the inner tube of the firework to be tested, is the surface roughness coefficient stored in the database, , =1, 2, 3, ..., , The number of measurement position points.

[0015] Furthermore, the specific steps for obtaining the mechanical strength index of the inner tube of the firework to be tested are as follows: obtaining the surface temperature reference value, strain reference value, vibration frequency reference value, and surface smoothness reference value of the inner tube of the firework to be tested; comprehensively analyzing the surface temperature value, strain value, vibration frequency value, and surface smoothness value of each measurement position point of the inner tube of the firework to be tested in combination with the surface temperature reference value, strain reference value, vibration frequency reference value, and surface smoothness reference value of the inner tube of the firework to be tested, to obtain the mechanical strength index of the inner tube of the firework to be tested; wherein the specific formula for calculating the mechanical strength index of the inner tube of the firework to be tested is as follows: ;in, is the mechanical strength index of the inner tube of the firework to be tested, For each of the fireworks inner tubes to be tested The surface temperature value of each measurement point, is the reference value of the surface temperature of the inner tube of the firework to be tested, is the surface temperature coefficient stored in the database, For each of the fireworks inner tubes to be tested The strain value of each measurement point, is the strain reference value of the inner tube of the firework to be tested, is the gauge factor stored in the database, For each of the fireworks inner tubes to be tested The vibration frequency value of each measurement point, is the reference value of the vibration frequency of the inner tube of the firework to be tested, is the vibration frequency coefficient stored in the database, For each of the fireworks inner tubes to be tested The surface smoothness value of each measurement point, is the reference value of the surface smoothness of the inner tube of the firework to be tested, is the surface smoothness coefficient stored in the database, , is a natural constant, =1, 2, 3, ..., , The number of measurement position points.

[0016] A fireworks inner tube quality inspection system for fireworks production includes: an image intelligent analysis module, a defect recognition module, an image quality intelligent judgment module, a comprehensive quality analysis module, and a comprehensive quality judgment module; the image intelligent analysis module is used to obtain image data of several image channels of the fireworks inner tube to be inspected, and perform comprehensive analysis to obtain comprehensive image data of the fireworks inner tube to be inspected, wherein the image data is specifically the pixel value of each pixel point in the fireworks inner tube image, and the comprehensive image data is specifically the comprehensive pixel value of each pixel point in the comprehensive image of the fireworks inner tube; the defect recognition module is used to input the comprehensive image data of the fireworks inner tube to be inspected into a pre-trained neural network model for defect recognition analysis, obtain several hole areas and several crack areas in the comprehensive image of the fireworks inner tube to be inspected, and perform comprehensive analysis to obtain the image quality index of the fireworks inner tube to be inspected; the image quality intelligent judgment module is used to judge whether the image quality index of the fireworks inner tube to be inspected is within a preset image quality threshold range, and if the fireworks inner tube to be inspected is within a preset image quality threshold range, ... If the image quality index of the inner tube is outside the preset image quality threshold range, the firework inner tube to be tested is marked as unqualified; the comprehensive quality analysis module is used to obtain physical feature data of the firework inner tube to be tested when the image quality index of the firework inner tube to be tested is within the preset image quality threshold range, and perform comprehensive analysis to obtain the comprehensive quality strength index of the firework inner tube to be tested, wherein the physical feature data includes the material density value, wall thickness value, material hardness value, surface roughness value, surface temperature value, strain value, vibration frequency value, and surface smoothness value of each measurement position point; the comprehensive quality judgment module is used to judge whether the comprehensive quality strength index of the firework inner tube to be tested is within the preset comprehensive quality strength threshold range. If the comprehensive quality strength index of the firework inner tube to be tested is within the preset comprehensive quality strength threshold range, the firework inner tube to be tested is marked as qualified; if the comprehensive quality strength index of the firework inner tube to be tested is outside the preset comprehensive quality strength threshold range, the firework inner tube to be tested is marked as unqualified.

[0017] The present invention has the following beneficial effects:

[0018] (1) This quality inspection method for fireworks inner tubes used in fireworks production not only analyzes the appearance defects of fireworks inner tubes, but also incorporates the physical characteristics of the inner tubes into the quality assessment system. Since traditional inspection methods often rely only on visual inspection of surface defects or a single physical parameter, it is easy to make it difficult to fully reflect the actual quality of fireworks inner tubes. By combining image data with physical characteristic data, the quality of fireworks inner tubes can be evaluated more comprehensively and accurately, ensuring their safety and reliability during production, transportation and firing.

[0019] (2) This method for quality inspection of fireworks inner tubes used in fireworks production can significantly improve the accuracy of quality inspection by combining image data with a variety of physical parameter data for comprehensive analysis. In particular, the image data combined with the defect recognition capability of the deep learning model can automatically and efficiently detect small defects that may be difficult to identify using traditional manual inspection methods. Furthermore, the comprehensive analysis of physical feature data makes the inspection process more comprehensive, covering different performance indicators of fireworks inner tubes, improving the accuracy of quality assessment, and reducing missed inspections or misjudgments caused by human factors or a single inspection method.

[0020] (3) The fireworks inner tube quality inspection system used for fireworks production realizes a fully automated quality inspection process by integrating multiple modules such as image intelligent analysis, defect recognition, image quality judgment, and comprehensive quality analysis. The system can quickly obtain image data of the fireworks inner tube and perform defect recognition in real time. It analyzes the hole and crack areas through a pre-trained neural network model, further calculates the image quality index and performs intelligent judgment to automatically determine whether the fireworks inner tube is qualified. In addition, the system also combines the analysis of physical feature data to perform a more detailed comprehensive quality strength assessment. This highly automated quality inspection system not only reduces manual intervention and improves production efficiency, but also provides real-time feedback on quality problems, shortens the inspection cycle, and ensures the continuous operation of the production line. Through automation and real-time optimization, it not only improves overall production efficiency, but also greatly reduces the errors that may occur during manual inspection, ensures the quality reliability of each fireworks inner tube, and thus improves the stability and safety of the production process.

[0021] 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

[0022] Figure 1 This is a flow chart of a method for detecting the quality of fireworks inner tubes used in fireworks production according to the present invention.

[0023] Figure 2 The present invention is a flowchart of the specific steps of obtaining the image quality index of the firework inner tube to be tested in a method for testing the quality of firework inner tubes used in firework production.

[0024] Figure 3 This is a block diagram of a fireworks inner tube quality detection system for fireworks production according to the present invention. DETAILED DESCRIPTION

[0025] The embodiments of the present application address the problem of existing methods for inspecting the quality of fireworks inner tubes, which rely too much on single image data for defect detection, by providing a method and system for inspecting the quality of fireworks inner tubes for use in fireworks production. Although this method can detect surface defects such as holes and cracks, it is difficult to comprehensively assess the overall quality and strength of the fireworks inner tube. Because the quality of a fireworks inner tube depends not only on surface defects but also on multiple factors such as its internal material properties and structural strength, a single image inspection cannot fully reflect its true structural health. Therefore, existing technologies have difficulty accurately predicting damage or failure that may occur to the fireworks inner tube during actual firing, posing a significant safety hazard. Secondly, existing inspection methods typically only use local or limited physical parameters for evaluation, making it difficult to accurately reflect the overall quality of the fireworks inner tube through comprehensive and detailed multi-dimensional data analysis. This can easily lead to some fireworks inner tubes that meet surface requirements becoming structurally unstable or failing in actual use due to the failure to consider the influence of all key parameters, thereby affecting the safety and quality of the fireworks.

[0026] The overall approach to the problems in the embodiments of this application is as follows:

[0027] First, image data of several image channels of the firework inner tube to be inspected are obtained through image acquisition technology. These image data contain the pixel value of each pixel point. Through comprehensive analysis, these image channel data are fused into a comprehensive image data to facilitate subsequent defect identification and quality assessment. The obtained comprehensive image data is input into the pre-trained convolutional neural network model for defect identification analysis. The neural network model can effectively identify the holes and crack areas in the firework inner tube image, and analyze the area of ​​the defect area to obtain the image quality index. The results of this stage help screen out the firework inner tubes with surface defects. When the image quality meets the requirements, the physical characteristic data of the firework inner tube (such as material density, wall thickness, material hardness, surface roughness, surface temperature, strain, vibration frequency and surface smoothness, etc.) are further obtained and comprehensively analyzed. The comprehensive quality strength index of the firework inner tube is calculated through the comprehensive physical characteristic data. Finally, the qualification of the firework inner tube is judged according to the preset quality threshold and marked as qualified or unqualified.

[0028] See also Figure 1, an embodiment of the present invention provides a technical solution: a method for detecting the quality of fireworks inner tubes for fireworks production, comprising the following steps: acquiring image data of several image channels of the fireworks inner tube to be detected after production, and performing comprehensive analysis to obtain comprehensive image data of the fireworks inner tube to be detected, wherein the image data is specifically the pixel value of each pixel point in the fireworks inner tube image, and the comprehensive image data is specifically the comprehensive pixel value of each pixel point in the comprehensive image of the fireworks inner tube; inputting the comprehensive image data of the fireworks inner tube to be detected into a pre-trained neural network model for defect recognition analysis, obtaining several hole areas and several crack areas in the comprehensive image of the fireworks inner tube to be detected, and performing comprehensive analysis to obtain an image quality index of the fireworks inner tube to be detected; judging whether the image quality index of the fireworks inner tube to be detected is within a preset image quality threshold range; if the image quality index of the fireworks inner tube to be detected is within a preset image quality threshold range, If the image quality index of the inner tube of the firework to be tested is outside the preset image quality threshold range, the inner tube of the firework to be tested will be marked as unqualified and classified into the preset unqualified quality area; if the image quality index of the inner tube of the firework to be tested is within the preset image quality threshold range, the physical feature data of the inner tube of the firework to be tested will be obtained, and a comprehensive analysis will be performed to obtain the comprehensive quality strength index of the inner tube of the firework to be tested; it will be determined whether the comprehensive quality strength index of the inner tube of the firework to be tested is within the preset comprehensive quality strength threshold range; if the comprehensive quality strength index of the inner tube of the firework to be tested is within the preset comprehensive quality strength threshold range, the inner tube of the firework to be tested will be marked as qualified and classified into the preset qualified quality area; if the comprehensive quality strength index of the inner tube of the firework to be tested is outside the preset comprehensive quality strength threshold range, the inner tube of the firework to be tested will be marked as unqualified and classified into the preset unqualified quality area.

[0029] The specific formula for calculating the comprehensive pixel value of each pixel in the comprehensive image of the firework inner tube to be detected is as follows: ;in, The first image of the firework inner tube to be detected is The comprehensive pixel value of each pixel, The first The image of the inner tube of the firework in the image channel The pixel value of each pixel, The first The image of the inner tube of the firework in the image channel The weighting coefficient of the pixel value of each pixel, =1, 2, 3, ..., , is the number of pixels, =1, 2, 3, ..., , is the number of image channels.

[0030] Specifically, the neural network model is a convolutional neural network model, and the specific steps for obtaining several hole areas and several crack areas in the comprehensive image of the inner tube of the firework to be detected are as follows: adjusting the comprehensive image data of the inner tube of the firework to be detected, including: normalizing the pixel values ​​of the comprehensive image data to the input range of the model (such as [0,1] or [-1,1]), if the size of the comprehensive image is inconsistent with the input size of the model, using the interpolation method to adjust the image size; converting the adjusted comprehensive image data of the inner tube of the firework to be detected into a tensor (because the input and output of its neural network are usually tensors, tensors can have different dimensions. For image data, there are usually three dimensions: height (H), width (W) and number of channels (C). For example, for RGB images, the tensor dimension of the image is (H, W, 3), that is, height * width * 3 (RGB three color channels). In the convolutional neural network (CNN), the image data is input as a tensor and passes through a series of convolutions. After operations such as multiplication and pooling, the output is also a tensor, which usually represents the feature map of the image) and is input into the pre-trained convolutional neural network model for feature extraction (including: convolution operation (the convolution layer of the model will extract high- and low-level features of the image layer by layer, for example: low-level features: edges, textures; high-level features: hole shapes, crack connection areas, etc.), pooling operation: (reducing the spatial resolution of the image and retaining key features)); defect identification and classification processing is performed on the feature extraction results (including: the model classifies each pixel in the input image as "hole", "crack" or "normal area", and outputs a mask image (mask), for example: pixels in the hole area are marked as 1; pixels in the crack area are marked as 2; pixels in the defect-free area are marked as 0. According to the classification results, the marked hole and crack areas are extracted to generate a set of area coordinates (such as contours, polygonal areas, etc.)), and several hole areas and several crack areas in the comprehensive image of the firework inner tube to be detected are obtained.

[0031] The convolutional neural network model includes: an input layer for receiving comprehensive image data of the firework inner tube. The input data is usually a multi-channel two-dimensional image matrix; an input layer representing the comprehensive information of multiple image channels; a convolutional layer responsible for extracting image features, such as the edges and shapes of holes and cracks, whose weights are learned from pre-training tasks and capable of identifying defect features; a pooling layer for reducing the dimensionality of the data while retaining important features and improving the model's computational efficiency; a fully connected layer for mapping the extracted features to specific output categories, such as "hole region," "crack region," or "defect-free region"; and an output layer for providing the model's prediction results, the output of which is a mask that marks the holes and cracks in the image.

[0032] The pre-training process of the convolutional neural network model is as follows:

[0033] Obtain a dataset containing a large amount of labeled data (such as images of firework inner tubes with holes and cracks labeled) to train the convolutional neural network model. Each sample in the dataset includes the input comprehensive image data and the corresponding target area mask.

[0034] During the training process, a loss function (such as cross entropy loss or Dice loss) is used to measure the difference between the predicted results and the true annotations, and the convolutional neural network model is gradually optimized.

[0035] The convolutional neural network model parameters are updated through optimization algorithms (such as Adam or SGD) to gradually improve the recognition accuracy.

[0036] After the training is completed, the weights and structure of the convolutional neural network model are saved for actual detection tasks.

[0037] In this embodiment, image analysis using a convolutional neural network can automatically, quickly, and accurately detect defects in firework tubes. First, after pre-training, the convolutional neural network model can effectively extract high-level features in the image, such as the shape of the hole and the connection of the cracks, by learning from a large amount of labeled data (including images of firework tubes labeled with defects such as holes and cracks). This enables the system to not only identify the type of defect when processing complex image data, but also accurately locate the specific area. Secondly, through data preprocessing and adjustment (such as pixel value normalization and resizing), the system ensures that the image data is input into the neural network in the optimal format, avoiding errors caused by data inconsistency. The robustness and accuracy of the model are enhanced. Furthermore, the combination of convolutional layers and pooling layers enables the system to extract low-level features (such as edges and textures) to high-level features (such as the shape and position of holes and cracks) from the image layer by layer, and reduce computational complexity and improve processing efficiency through pooling operations. Through the above process, the system can automatically identify and classify defects (holes, cracks, etc.) in the image of the firework inner tube, and generate a mask image to mark the defective area. This not only improves the accuracy of defect detection, but also greatly reduces the error and time cost of manual inspection. Ultimately, real-time and accurate quality inspection can be achieved on the production line to ensure that the quality of each firework inner tube meets the standards, thereby improving product consistency and production efficiency.

[0038] Specifically, if Figure 2 As shown, the specific steps for obtaining the image quality index of the firework inner tube to be tested are as follows: counting the number of pixels in each hole area and each crack area in the comprehensive image of the firework inner tube to be tested, and using these as the area value of each hole area and the area value of each crack area in the comprehensive image of the firework inner tube to be tested; obtaining the cross-sectional area value of the firework inner tube to be tested, and performing a comprehensive analysis based on the area value of each hole area and the area value of each crack area in the comprehensive image of the firework inner tube to be tested, to obtain the image hole influence factor and the image crack influence factor of the firework inner tube to be tested; obtaining the reference bearing capacity value of the firework inner tube to be tested (representing the maximum bearing capacity value of the firework inner tube material under ideal conditions, i.e., without holes, cracks or other defects, and under standard environmental parameters), and performing a comprehensive analysis based on the image hole influence factor and the image crack influence factor of the firework inner tube to be tested, to obtain the image quality index of the firework inner tube to be tested.

[0039] The specific formulas for calculating the image hole impact factor, image crack impact factor, and image quality index of the firework inner tube to be inspected are as follows: ;in, is the image hole impact factor of the inner tube of the firework to be detected, The first image of the firework inner tube to be detected is The area value of the hole area, is the cross-sectional area of ​​the inner tube of the firework to be tested, is the image crack impact factor of the firework inner tube to be detected, =1, 2, 3, ..., , is the number of hole areas, is the crack influence coefficient stored in the database, is a natural constant, and in this embodiment, its value is 2.71. The first image of the firework inner tube to be detected is The length of the crack area, The first image of the firework inner tube to be detected is The area value of the crack region, =1, 2, 3, ..., , is the number of crack areas, is the image quality index of the inner tube of the firework to be detected, is the reference load-bearing capacity value of the firework inner tube to be tested.

[0040] It should be explained that the cross-sectional area value of the inner tube of the firework to be tested can be obtained by measuring the outer diameter value and the inner diameter value of the firework inner tube and performing calculation and analysis. Specifically, the cross-sectional area value = (pi)*((outer diameter of the inner tube of the firework / 2) 2 -(Inner diameter of the firework tube / 2) 2 ).

[0041] The reference load-bearing capacity value of the firework inner tube to be tested can be obtained by measuring the internal pressure value generated by the firework inner tube when the firework is ignited, the average value of the inner and outer diameters of the firework inner tube, and the wall thickness of the firework inner tube, and performing calculations and analysis. Specifically, the reference load-bearing capacity value = (internal pressure value generated by the firework inner tube when the firework is ignited * average value of the inner and outer diameters of the firework inner tube) / wall thickness of the firework inner tube.

[0042] The crack influence coefficient stored in the database can be obtained by the following method:

[0043] Experimental testing method: By collecting firework inner tube samples with different crack lengths, areas and distribution characteristics, recording the actual strength or failure point of each sample under stress (such as internal pressure, tensile force, etc.), analyzing the degree of weakening of the firework inner tube strength by crack length, area, etc., extracting its quantitative impact on strength, establishing a relationship model between crack length and strength weakening, and obtaining an experimental data table of crack influence coefficients.

[0044] Simulation calculation method: A three-dimensional model is established based on the geometric structure and material properties of the firework inner tube, and cracks of different lengths and areas are introduced into the model to simulate the stress conditions (such as internal pressure, external pressure, etc.) of the firework inner tube in actual work. By analyzing the stress distribution around the cracks, the degree to which the cracks weaken the structural strength is extracted, and the results are converted into crack influence coefficients.

[0045] The length value of the crack area is obtained based on image processing technology analysis, as follows: use edge detection algorithms (such as Canny and Sobel) to extract the boundaries of the cracks from the image, binarize the image, and generate a binary image with only the cracks, where the crack pixel value is 1 and the other pixel values ​​are 0. Skeletonize the binary crack image and turn the crack area into a 1-pixel wide line. This can eliminate the influence of the crack width and only retain its center line. Traverse all pixel points in the skeleton image and count the total number of all crack pixels, which is the length value of the crack area.

[0046] In this embodiment, by counting the number of pixels in each hole area and crack area, combined with the cross-sectional area and reference bearing capacity value of the firework inner tube, the impact of defects (such as holes and cracks) on the overall quality of the firework inner tube can be quantitatively analyzed. The influencing factors of each defect (such as the image hole influencing factor and the image crack influencing factor) take into account the physical characteristics of the defect such as area and length, and comprehensively calculate them with the bearing capacity of the firework inner tube, making the quality assessment more accurate and comprehensive. The calculation of the image quality index combines multiple factors, such as the influence of holes and cracks, and the reference bearing capacity of the inner tube, so that the final quality index not only depends on the image data itself, but also integrates parameters closely related to the physical properties of the firework inner tube. This comprehensive analysis can more comprehensively reflect the quality strength of the inner tube, especially when there are holes or cracks, its impact on the structural strength can be accurately evaluated through image analysis and bearing capacity model. In order to ensure the accuracy of the crack influencing factor, a combination of experimental testing method and simulation calculation method is adopted. Through experiments, different crack lengths, The strength influence data of factors such as area, etc., and the analysis of the weakening effect of cracks on the structure by establishing a three-dimensional model through simulation calculation can obtain the influence coefficient of cracks in a more scientific and accurate manner. This double verification method not only improves the reliability of the data, but also provides a solid theoretical and practical foundation for future quality assessment. The image processing technology (such as edge detection, binarization, skeletonization, etc.) is used to accurately extract the crack length, making the image analysis results more physically meaningful. These image processing technologies can effectively remove the influence of the crack width and only retain the center line of the crack, thereby accurately calculating the length value of the crack, providing more accurate input data for subsequent quality assessment. Through this comprehensive calculation process, the system can realize automated quality assessment without relying on human intervention. The quality analysis of each firework inner tube is not only fast and accurate, but also helps to improve the automation level of the production line and reduce errors caused by human factors. In addition, this method can be widely used in the production process to monitor and warn of quality problems in real time to ensure that each firework inner tube produced meets safety and quality standards.

[0047] Specifically, the specific steps for obtaining the comprehensive quality strength index of the inner tube of the firework to be tested are as follows: reading the physical characteristic data of the inner tube of the firework to be tested, and performing a comprehensive analysis to obtain the basic quality strength index and mechanical strength index of the inner tube of the firework to be tested; performing a comprehensive analysis on the basic quality strength index and mechanical strength index of the inner tube of the firework to be tested to obtain the comprehensive quality strength index of the inner tube of the firework to be tested.

[0048] The specific formula for calculating the comprehensive quality strength index of the firework inner tube to be tested is as follows: ;in, is the comprehensive quality strength index of the firework inner tube to be tested, is the basic mass strength index of the firework inner tube to be tested, is the basic quality coefficient stored in the database, is the mechanical strength index of the inner tube of the firework to be tested, is the mechanical strength coefficient stored in the database, , is a natural constant and in this embodiment is taken as 2.71.

[0049] It should be explained that the data stored in the database 、 The specific calculation process is as follows: the material density reference value, wall thickness reference value, material hardness reference value, and surface roughness reference value of the fireworks inner tube to be tested after standardization are weighted analyzed with the corresponding coefficients to obtain the basic quality strength reference index.

[0050] The standardized surface temperature reference value, strain reference value, vibration frequency reference value, and surface smoothness reference value of the firework inner tube to be tested are weightedly analyzed with the corresponding coefficients to obtain a mechanical strength reference index.

[0051] The basic quality strength reference index and the mechanical strength reference index are summed and analyzed to obtain the comprehensive quality strength value. Then, the basic quality strength reference index and the mechanical strength reference index are respectively analyzed in proportion to the comprehensive quality strength value. The proportion analysis result is the corresponding coefficient.

[0052] The following is a specific implementation example of calculating the comprehensive quality strength index of the inner tube of a firework to be tested. The following data is available: material density, wall thickness, material hardness, surface roughness, surface temperature, strain, vibration frequency, and surface smoothness at eight measurement points. The specific data are shown in Tables 1 and 2:

[0053] Table 1 Examples of physical characteristic data of the inner tube of a firework to be tested

[0054] ;

[0055] Table 2 Examples of reference data for the physical characteristics of the inner tube of a firework to be tested

[0056]

[0057] Material density coefficients stored in the database Approximately: 0.272;

[0058] Wall thickness coefficient stored in the database Approximately: 0.248;

[0059] Material hardness coefficients stored in the database Approximately: 0.251;

[0060] Surface roughness coefficients stored in the database Approximately: 0.229;

[0061] Substituting the above coefficients and the data in Tables 1 and 2 into the specific formula for calculating the basic mass strength index of the firework inner tube to be tested, we obtain:

[0062] Basic mass strength index of measurement position 1 of the inner tube of the firework to be tested ;

[0063] Basic mass strength index of measurement position 2 of the inner tube of the firework to be tested ;

[0064] Basic mass strength index of measurement position 3 of the inner tube of the firework to be tested ;

[0065] Basic mass strength index of measurement position 4 of the inner tube of the firework to be tested ;

[0066] Basic mass strength index of measurement position 5 of the inner tube of the firework to be tested ;

[0067] Basic mass strength index of measurement position 6 of the inner tube of the firework to be tested ;

[0068] Basic mass strength index of measurement position 7 of the inner tube of the firework to be tested ;

[0069] Basic mass strength index of measurement point 8 of the inner tube of the firework to be tested ;

[0070] Calculate the basic quality strength index of the firework inner tube to be tested from the above data .

[0071] In addition, the surface temperature coefficient stored in the database Approximately: 0.315;

[0072] Gauge factors stored in the database Approximately: 0.268;

[0073] Vibration frequency coefficients stored in the database Approximately: 0.248;

[0074] Surface smoothness coefficient stored in the database Approximately: 0.169;

[0075] Substituting the above coefficients and the data in Tables 1 and 2 into the specific formula for calculating the mechanical strength index of the inner tube of the firework to be tested, we obtain:

[0076] Mechanical strength index of the firework inner tube at measurement position 1 to be tested ;

[0077] Mechanical strength index of the firework inner tube at measurement position 2 to be tested ;

[0078] Mechanical strength index of the firework inner tube at measurement point 3 to be tested ;

[0079] Mechanical strength index of the firework inner tube at measurement point 4 to be tested ;

[0080] Mechanical strength index of the firework inner tube at measurement point 5 to be tested ;

[0081] Mechanical strength index of the firework inner tube at measurement position 6 to be tested ;

[0082] Mechanical strength index of the firework inner tube at measurement position 7 to be tested ;

[0083] Mechanical strength index of the firework inner tube at measurement point 8 to be tested ;

[0084] Calculate the mechanical strength index of the firework inner tube to be tested based on the above data .

[0085] And the basic quality coefficient stored in the database Approximately: 0.57;

[0086] Mechanical strength coefficients stored in the database Approximately: 0.43;

[0087] Substitute the above coefficients and the calculated basic mass strength index and mechanical strength index of the firework inner tube to be tested into the specific formula for calculating the comprehensive mass strength index of the firework inner tube to be tested to obtain:

[0088] Comprehensive quality strength index of the firework inner tube to be tested .

[0089] In this implementation plan, by combining the physical characteristic data of the firework inner tube (such as material density, wall thickness, material hardness, surface roughness, etc.) with the mechanical properties (such as surface temperature, strain, vibration frequency, surface smoothness, etc.), the basic quality strength index and mechanical strength index are calculated respectively, which can comprehensively evaluate the quality of the firework inner tube. This multi-dimensional analysis method can avoid the misjudgment that may be caused by a single parameter, and ensure that the quality assessment is more accurate and complete. When calculating the basic quality strength index and mechanical strength index, a standardized processing method is adopted to conduct a unified analysis of the various parameters of the firework inner tube to be tested. Through weighted analysis, it is ensured that the influence of each parameter is consistent with its physical characteristics, so that the final index calculation is more in line with the actual situation, and the quality is improved. The setting of weight coefficients can also be adjusted according to experimental data and actual conditions, thereby improving the flexibility and accuracy of the analysis. By calculating the basic quality strength reference index and the mechanical strength reference index respectively and conducting a comprehensive analysis of the two indices, the overall quality condition and mechanical bearing capacity of the fireworks inner tube can be effectively quantified and combined. This method can fully consider the material properties and the mechanical effects in actual applications. The final comprehensive quality strength index can better reflect the stability and safety of the fireworks inner tube in actual production. Through the coefficients and reference values ​​stored in the database, the various parameters of the fireworks inner tube are weighted and the corresponding coefficients are obtained. These database data can be continuously optimized and updated. With the accumulation of more sample data and test results, Accumulation can improve the accuracy and adaptability of the coefficients, so that the quality assessment model of the system can continuously optimize itself and improve long-term detection performance and adaptability. The calculation process of the comprehensive quality strength index combines multiple dimensions of basic quality and mechanical strength, making the overall assessment process more efficient. Through this method, the system can directly obtain the quality strength assessment results through simple physical measurement data without relying on complex external testing equipment, reducing test time and cost, and improving the real-time monitoring capability of the production process. Due to the use of natural constants and adjustable coefficients (such as basic quality coefficient and mechanical strength coefficient) in the formula, this method has good flexibility and adaptability, different production environments, different types of fireworks inner tubes, different The material properties can be adjusted by adjusting the relevant parameters and re-optimizing the model to ensure that it can provide accurate quality assessment results under various conditions. By adopting standardized and weighted mathematical models for quality analysis, the influence of human factors on the test results is reduced. The system can perform quality assessment with consistent standards to ensure that the quality test results of each firework inner tube have high accuracy and consistency, avoiding errors that may be caused by manual operation. Through the comprehensive quality strength index, the production line can quickly identify unqualified firework inner tubes and adjust the production process or raw material ratio in time. This not only helps to improve product quality, but also effectively avoids possible safety problems in the production process, and supports efficient quality control and production process optimization.

[0090] Specifically, the specific steps for obtaining the basic quality strength index of the inner tube of the firework to be tested are as follows: obtain the material density reference value, wall thickness reference value, material hardness reference value, and surface roughness reference value of the inner tube of the firework to be tested; comprehensively analyze the material density value, wall thickness value, material hardness value, and surface roughness value of each measurement position point of the inner tube of the firework to be tested in combination with the material density reference value, wall thickness reference value, material hardness reference value, and surface roughness reference value of the inner tube of the firework to be tested to obtain the basic quality strength index of the inner tube of the firework to be tested.

[0091] The specific formula for calculating the basic mass strength index of the firework inner tube to be tested is as follows: ;in, is the basic mass strength index of the firework inner tube to be tested, For each of the fireworks inner tubes to be tested The material density value at each measurement point, Reference value of the material density of the inner tube of the firework to be tested, is the material density coefficient stored in the database, For each of the fireworks inner tubes to be tested The wall thickness value of each measuring point, Reference value of the wall thickness of the inner tube of the fireworks to be tested, is the wall thickness coefficient stored in the database, For each of the fireworks inner tubes to be tested The material hardness value at each measuring point, is the reference value of the material hardness of the inner tube of the firework to be tested, is the material hardness coefficient stored in the database, For each of the fireworks inner tubes to be tested The surface roughness value of each measuring point, is the reference value of the surface roughness of the inner tube of the firework to be tested, is the surface roughness coefficient stored in the database, , =1, 2, 3, ..., , The number of measurement position points.

[0092] It should be explained that the material density value can be measured and obtained through high-precision density sensors (such as X-ray densitometers). It refers to the mass of the material contained in a unit volume, which directly affects the overall quality, strength and load-bearing capacity of the firework inner tube. The higher the density, the denser and stronger the material is. For the inner tube of a firework, too low a density means that the material is not strong enough and has poor load-bearing capacity; while too high a density can easily make the inner tube itself too heavy, affecting the effect and safety of the firework.

[0093] The material density reference value can be obtained by obtaining the conventional density value of the firework inner tube material from standard data.

[0094] The wall thickness value can be obtained by measuring with an ultrasonic thickness gauge, and the wall thickness can be calculated by the reflection time of the ultrasonic wave. It directly affects the mechanical strength and pressure resistance of the inner tube. The inner tube of the firework needs to have sufficient wall thickness to withstand the internal pressure, vibration and temperature changes. If the wall thickness is too thin, the inner tube is prone to rupture during the ignition process, causing the firework to lose control; if the wall thickness is too thick, it will increase the weight of the firework and affect performance.

[0095] Reference values ​​for wall thickness can be obtained from manufacturing standards. For example, the wall thickness of some fireworks inner tubes is specified to be 3mm, 5mm or thicker.

[0096] The hardness value of a material can be directly measured using a portable hardness tester (such as a Brinell hardness tester, Rockwell hardness tester, or Vickers hardness tester). It is an indicator of the material's ability to resist local deformation, scratches, and wear. Materials with higher hardness generally have stronger resistance to compression and scratches. For the inner tube of a firework, appropriate hardness can ensure that the inner tube will not be damaged by external impact or friction during the firing process. Materials with too high a hardness may increase brittleness and affect the toughness of the inner tube; materials with too low a hardness may cause damage or be unresistant to high temperatures.

[0097] The material hardness reference value can use the standard hardness value of the firework inner tube material as a reference value.

[0098] The surface roughness value can be measured using a surface roughness measuring instrument (such as a profilometer, contact roughness meter, optical roughness meter). It refers to the microscopic unevenness of the material surface and is usually expressed by indicators such as Ra (arithmetic average roughness) and Rz (ten-point average roughness).

[0099] Reference values ​​for surface roughness can be obtained from processing standards for relevant materials of fireworks inner tubes or from production process specification manuals.

[0100] Stored in the database 、 、 、 The specific calculation process is as follows: standardize the material density reference value, wall thickness reference value, material hardness reference value, and surface roughness reference value of the firework inner tube to be tested (i.e., remove the unit), and perform sum analysis after standardization to obtain the basic mass strength and value. Then, perform ratio analysis on the material density reference value, wall thickness reference value, material hardness reference value, and surface roughness reference value of the firework inner tube to be tested after standardization with the basic mass strength and value respectively, and the analysis result is the corresponding coefficient.

[0101] In this embodiment, by combining key physical parameters such as material density, wall thickness, material hardness, and surface roughness, the basic quality characteristics of the firework inner tube can be fully reflected. These factors are directly related to important properties such as the load-bearing capacity, durability, and pressure resistance of the firework inner tube, helping to ensure the stability and safety of the firework inner tube during the firing process. The various parameters of the firework inner tube to be tested (such as density, wall thickness, hardness, and roughness) are standardized to eliminate differences in units and ranges, so that the data has a unified comparison standard. This standardization step can eliminate the impact of different measuring equipment and measurement environments on the results, and improve the comparability and reliability of the data. The parameters of each measurement position point are weighted analyzed with the reference value and the corresponding coefficient, and the results can more accurately reflect the actual impact of each parameter on the overall quality strength. For example, material density and hardness The contribution of different degrees to the inner tube strength may be different. Through the weighting coefficient, the influence of each parameter can be flexibly adjusted to ensure that the final basic quality strength index is more in line with actual needs. Through the reference values ​​and coefficients stored in the database, it can be adapted according to the material type and production standard of different fireworks inner tubes. Different types of inner tubes can adjust the coefficients according to specific material density, hardness and wall thickness requirements, making this method highly adaptable and able to handle fireworks inner tubes of different specifications and materials. The calculation process of the basic quality strength index provides a quantifiable standard for quality control in production. By obtaining accurate data at each measurement point and real-time feedback on the quality level of the fireworks inner tube, potential quality problems such as insufficient density and uneven wall thickness can be discovered in a timely manner, and the production process can be adjusted and optimized to improve product consistency and safety.

[0102] Specifically, the specific steps for obtaining the mechanical strength index of the inner tube of the firework to be tested are as follows: obtaining the surface temperature reference value, strain reference value, vibration frequency reference value, and surface smoothness reference value of the inner tube of the firework to be tested; comprehensively analyzing the surface temperature value, strain value, vibration frequency value, and surface smoothness value of each measurement position point of the inner tube of the firework to be tested in combination with the surface temperature reference value, strain reference value, vibration frequency reference value, and surface smoothness reference value of the inner tube of the firework to be tested to obtain the mechanical strength index of the inner tube of the firework to be tested.

[0103] The specific formula for calculating the mechanical strength index of the firework inner tube to be tested is as follows: ;in, is the mechanical strength index of the inner tube of the firework to be tested, For each of the fireworks inner tubes to be tested The surface temperature value of each measurement point, is the reference value of the surface temperature of the inner tube of the firework to be tested, is the surface temperature coefficient stored in the database, For each of the fireworks inner tubes to be tested The strain value of each measurement point, is the strain reference value of the inner tube of the firework to be tested, is the gauge factor stored in the database, For each of the fireworks inner tubes to be tested The vibration frequency value of each measurement point, is the reference value of the vibration frequency of the inner tube of the firework to be tested, is the vibration frequency coefficient stored in the database, For each of the fireworks inner tubes to be tested The surface smoothness value of each measurement point, is the reference value of the surface smoothness of the inner tube of the firework to be tested, is the surface smoothness coefficient stored in the database, , is a natural constant, and in this embodiment, its value is 2.71. =1, 2, 3, ..., , The number of measurement position points.

[0104] It should be explained that the surface temperature value can be measured and obtained through an infrared thermal imager or a thermocouple sensor. It refers to the temperature distribution on the surface of the inner tube of a firework during the firing process or under other working conditions.

[0105] Reference values ​​for surface temperature can be obtained from the material data sheet of the firework inner tube or through standard tests during production (such as thermal stability tests under different environments).

[0106] Strain values ​​can be measured using strain gauges. They refer to the amount of deformation per unit length of a material when subjected to external loads or pressure. In the application of firework inner tubes, the strain value can be used to evaluate the deformation capacity of the inner tube structure when subjected to internal and external pressures (such as shock waves generated by explosions, temperature changes, etc.). Excessive strain may mean that the inner tube structure has undergone permanent deformation or even lost its original strength.

[0107] The strain reference value can usually be determined by the constitutive model of the firework inner tube material (such as elastic modulus, yield stress, etc.) and the design specifications used.

[0108] Vibration frequency, measured using an accelerometer or vibration sensor, refers to the frequency at which the inner tube of a firework vibrates during ignition or when subjected to external forces. This frequency reflects the elastic properties of the tube material under dynamic loads. Excessively high vibration frequencies can cause fatigue damage or even rupture. Accurately measuring vibration frequency helps determine the dynamic stability of a firework's inner tube.

[0109] The reference value of vibration frequency can be obtained through the natural frequency of the material, vibration modal analysis or vibration experiments similar to the inner tube of fireworks, and the vibration analysis in the design stage can also provide a basis for the reference value.

[0110] The surface smoothness value can be measured using an optical surface smoothness meter, laser scanning microscope or profilometer. It refers to the smoothness of the material's surface microstructure. A smooth surface can usually reduce friction, wear and heat accumulation, while materials with higher surface roughness will affect the strength of the firework inner tube, especially during the firing process. Rough surfaces are more likely to produce cracks or stress concentration. Surface smoothness will also affect the emission of airflow and smoke.

[0111] The reference value of surface smoothness can usually be obtained through the material processing technology standard, equipment accuracy, etc., and the smoothness standard can be obtained based on the material coating treatment, processing accuracy, etc.

[0112] Stored in the database 、 、 、 The specific calculation process is as follows: the surface temperature reference value, strain reference value, vibration frequency reference value, and surface smoothness reference value of the inner tube of the firework to be tested are standardized (i.e., unit-removed), and summed up and analyzed after the standardization to obtain the mechanical strength sum value. Then, the surface temperature reference value, strain reference value, vibration frequency reference value, and surface smoothness reference value of the inner tube of the firework to be tested after the standardization are respectively compared with the mechanical strength sum value. The analysis result is the corresponding coefficient.

[0113] In this embodiment, through the comprehensive analysis of multiple mechanical performance parameters such as surface temperature, strain, vibration frequency and surface smoothness, it is possible to comprehensively evaluate the various mechanical stresses and loads that the inner tube of a firework may withstand in actual use. These factors are crucial for evaluating the pressure resistance, fatigue resistance and overall stability of the inner tube of a firework, ensuring the safety of the inner tube of a firework under high temperature, high pressure and dynamic loads. The use of advanced sensors and measuring equipment (such as infrared thermal imagers, strain gauges, accelerometers, etc.) can obtain high-precision data, avoiding the errors that may be caused by traditional measurement methods. This precise measurement can more accurately capture the changes in the inner tube of a firework during the firing process. The response of the fireworks inner tube to changes in the process or external environment provides a reliable basis for the evaluation of mechanical strength. This method can monitor the changes in the mechanical properties of the fireworks inner tube under different environmental conditions in real time. For example, under conditions of high temperature and high vibration, the strength of the inner tube may change. Real-time detection of the mechanical strength index helps to evaluate its performance under different working conditions and ensure its adaptability and stability. By standardizing parameters such as surface temperature, strain, and vibration frequency, we can better understand the design defects and potential problems of the fireworks inner tube. For example, too high a vibration frequency may indicate fatigue hazards in the design, and too large a strain value may mean that the inner tube structure is not strong enough. This helps optimize design and processes during production, improving the mechanical performance and safety of the product. These mechanical property assessments help determine whether the inner tube of a firework is prone to cracking or deformation during actual firing. Calculating the mechanical strength index can identify potential weaknesses in the inner tube early on, reducing safety incidents caused by material fatigue, cracking, overheating, or other mechanical issues, and ensuring the stability and reliability of the fireworks during firing. By quantitatively analyzing and weighting various mechanical performance parameters (such as temperature, strain, vibration frequency, and smoothness), they can be converted into operational coefficients. This quantitative analysis makes the calculation of the mechanical strength index more objective and scientific, and the standardized data more consistent, facilitating comparison and trend analysis. Different firework inner tubes may use different materials or production processes. Using the reference values ​​and coefficients stored in the database, the calculation model can be adjusted based on the physical properties of different materials, making this method highly flexible and applicable to a wide range of firework inner tube types. During production, real-time measurement and analysis of the mechanical strength index of firework inner tubes allows for immediate identification and adjustment of substandard inner tubes. This provides strong support for quality control, enabling real-time monitoring and assessment of product quality on the production line, thereby preventing substandard products from entering the market.

[0114] See also Figure 3, an embodiment of the present invention provides a technical solution: a fireworks inner tube quality inspection system for fireworks production, comprising: an image intelligent analysis module, a defect recognition module, an image quality intelligent judgment module, a comprehensive quality analysis module, and a comprehensive quality judgment module; the image intelligent analysis module is used to obtain image data of several image channels of the fireworks inner tube to be inspected, and perform comprehensive analysis to obtain comprehensive image data of the fireworks inner tube to be inspected, the image data specifically being the pixel value of each pixel point in the fireworks inner tube image, and the comprehensive image data specifically being the comprehensive pixel value of each pixel point in the comprehensive image of the fireworks inner tube; the defect recognition module is used to input the comprehensive image data of the fireworks inner tube to be inspected into a pre-trained neural network model for defect recognition analysis, obtain several hole areas and several crack areas in the comprehensive image of the fireworks inner tube to be inspected, and perform comprehensive analysis to obtain the image quality index of the fireworks inner tube to be inspected; the image quality intelligent judgment module is used to judge whether the image quality index of the fireworks inner tube to be inspected is within a preset image quality threshold range, and if If the image quality index of the firework inner tube to be tested is outside the preset image quality threshold range, the firework inner tube to be tested is marked as unqualified; the comprehensive quality analysis module is used to obtain the physical feature data of the firework inner tube to be tested when the image quality index of the firework inner tube to be tested is within the preset image quality threshold range, and perform comprehensive analysis to obtain the comprehensive quality strength index of the firework inner tube to be tested, the physical feature data including the material density value, wall thickness value, material hardness value, surface roughness value, surface temperature value, strain value, vibration frequency value, and surface smoothness value of each measurement position point; the comprehensive quality judgment module is used to judge whether the comprehensive quality strength index of the firework inner tube to be tested is within the preset comprehensive quality strength threshold range. If the comprehensive quality strength index of the firework inner tube to be tested is within the preset comprehensive quality strength threshold range, the firework inner tube to be tested is marked as qualified; if the comprehensive quality strength index of the firework inner tube to be tested is outside the preset comprehensive quality strength threshold range, the firework inner tube to be tested is marked as unqualified.

[0115] In summary, this application has at least the following effects:

[0116] Not only is the appearance of the inner tube of a firework analyzed for defects, but the physical properties of the inner tube are also incorporated into the quality assessment system. Traditional inspection methods often rely solely on visual inspection of surface defects or a single physical parameter, making it difficult to fully reflect the actual quality of the inner tube of a firework. By combining image data with physical characteristic data, the quality of the inner tube of a firework can be assessed more comprehensively and accurately, ensuring its safety and reliability during production, transportation, and firing.

[0117] By combining image data with a variety of physical parameter data for comprehensive analysis, the accuracy of quality inspection can be significantly improved. In particular, image data combined with the defect recognition capabilities of deep learning models can automatically and efficiently detect small defects that may be difficult to identify with traditional manual inspection methods. Furthermore, the comprehensive analysis of physical feature data makes the inspection process more comprehensive, covering the different performance indicators of the firework inner tube, improving the accuracy of quality assessment, and reducing missed detections or misjudgments due to human factors or a single inspection method.

[0118] By integrating multiple modules such as image intelligent analysis, defect recognition, image quality judgment, and comprehensive quality analysis, a fully automated quality inspection process is achieved. The system can quickly obtain image data of the firework inner tube and perform defect recognition in real time. It analyzes the hole and crack areas through a pre-trained neural network model, further calculates the image quality index and performs intelligent judgment to automatically determine whether the firework inner tube is qualified. In addition, the system also combines the analysis of physical feature data to perform a more detailed comprehensive quality strength assessment. This highly automated quality inspection system not only reduces manual intervention and improves production efficiency, but also provides real-time feedback on quality problems, shortens the inspection cycle, and ensures the continuous operation of the production line. Through automation and real-time optimization, it not only improves overall production efficiency, but also greatly reduces the errors that may occur during manual inspection, ensuring the quality reliability of each firework inner tube, thereby improving the stability and safety of the production process.

[0119] Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional 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.

[0120] Obviously, those skilled in the art may make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if such changes and modifications fall within the scope of the claims and their equivalents, the present invention is intended to include such changes and modifications.

Claims

1. A method for detecting the quality of fireworks inner tubes used in fireworks production, characterized in that: The following steps are involved: Acquire image data of several image channels of the inner tube of the firework to be detected, and perform comprehensive analysis to obtain comprehensive image data of the inner tube of the firework to be detected, wherein the image data is specifically the pixel value of each pixel point in the image of the inner tube of the firework, and the comprehensive image data is specifically the comprehensive pixel value of each pixel point in the comprehensive image of the inner tube of the firework; Inputting the comprehensive image data of the firework inner tube to be inspected into the pre-trained neural network model for defect recognition analysis, obtaining a plurality of hole areas and a plurality of crack areas in the comprehensive image of the firework inner tube to be inspected, and performing comprehensive analysis to obtain the image quality index of the firework inner tube to be inspected; Determine whether the image quality index of the inner tube of the firework to be detected is within a preset image quality threshold range; If the image quality index of the firework inner tube to be detected is outside the preset image quality threshold range, the firework inner tube to be detected is marked as unqualified; If the image quality index of the inner tube of the firework to be detected is within the preset image quality threshold range, then the physical characteristic data of the inner tube of the firework to be detected is obtained, and a comprehensive analysis is performed to obtain the comprehensive quality strength index of the inner tube of the firework to be detected, wherein the physical characteristic data includes the material density value, wall thickness value, material hardness value, surface roughness value, surface temperature value, strain value, vibration frequency value, and surface smoothness value of each measuring position point; Determine whether the comprehensive quality strength index of the firework inner tube to be tested is within a preset comprehensive quality strength threshold range; If the comprehensive quality strength index of the firework inner tube to be tested is within the preset comprehensive quality strength threshold range, the firework inner tube to be tested is marked as qualified; If the comprehensive quality strength index of the firework inner tube to be tested is outside the preset comprehensive quality strength threshold range, the firework inner tube to be tested will be marked as unqualified.

2. The method for detecting the quality of fireworks inner tubes used in fireworks production according to claim 1, characterized in that: The specific formula for calculating the comprehensive pixel value of each pixel in the comprehensive image of the firework inner tube of the firework inner tube to be detected is as follows: ； in, is the first in the comprehensive image of the fireworks inner tube to be detected The comprehensive pixel value of each pixel, The first The image of the inner tube of the firework in the image channel The pixel value of a pixel, The first The image of the inner tube of the firework in the image channel The weighting coefficient of the pixel value of each pixel, =1, 2, 3, ..., , is the number of pixels, =1, 2, 3, …, , is the number of image channels.

3. The method for detecting the quality of fireworks inner tubes used in fireworks production according to claim 1, characterized in that: The neural network model is specifically a convolutional neural network model. The specific steps of obtaining a plurality of hole regions and a plurality of crack regions in the firework inner tube comprehensive image of the firework inner tube to be detected are as follows: Adjusting and processing the comprehensive image data of the inner tube of the fireworks to be inspected; The adjusted and processed comprehensive image data of the inner tube of the firework to be detected is converted into a tensor and input into a pre-trained convolutional neural network model for feature extraction; Defect recognition and classification processing is performed on the feature extraction results to obtain a number of hole areas and a number of crack areas in the comprehensive image of the firework inner tube of the firework inner tube to be detected.

4. The method for detecting the quality of fireworks inner tubes used in fireworks production according to claim 1, characterized in that: The specific steps for obtaining the image quality index of the inner tube of the firework to be detected are as follows: Counting the number of pixels in each hole region and each crack region in the comprehensive image of the inner tube of the firework to be detected, and using them as the area value of each hole region and the area value of each crack region in the comprehensive image of the inner tube of the firework to be detected; Obtain the cross-sectional area value of the firework inner tube to be detected, and perform comprehensive analysis based on the area value of each hole region and the area value of each crack region in the firework inner tube comprehensive image of the firework inner tube to be detected, to obtain the image hole influence factor and the image crack influence factor of the firework inner tube to be detected; The reference bearing capacity value of the firework inner tube to be tested is obtained, and a comprehensive analysis is performed on the image hole influence factor and the image crack influence factor of the firework inner tube to be tested to obtain the image quality index of the firework inner tube to be tested.

5. The method for detecting the quality of fireworks inner tubes used in fireworks production according to claim 4, characterized in that: The specific formulas for calculating the image hole impact factor, image crack impact factor, and image quality index of the inner tube of the firework to be detected are as follows: ； in, is the image hole impact factor of the inner tube of the firework to be detected, is the first in the comprehensive image of the fireworks inner tube of the fireworks inner tube to be detected The area value of the hole area, is the cross-sectional area of ​​the inner tube of the firework to be tested, is the image crack impact factor of the inner tube of the firework to be detected, =1, 2, 3, …, , is the number of hole areas, is the crack influence coefficient stored in the database, is a natural constant, is the first in the comprehensive image of the fireworks inner tube of the fireworks inner tube to be detected The length of the crack area, is the first in the comprehensive image of the fireworks inner tube of the fireworks inner tube to be detected The area value of the crack region, =1, 2, 3, …, , is the number of crack areas, is the image quality index of the inner tube of the firework to be detected, It is the reference load-bearing capacity value of the inner tube of the firework to be tested.

6. The method for detecting the quality of fireworks inner tubes used in fireworks production according to claim 1, characterized in that: The specific steps for obtaining the comprehensive quality strength index of the inner tube of the firework to be tested are as follows: Read the physical characteristic data of the inner tube of the firework to be tested, and conduct a comprehensive analysis to obtain the basic quality strength index and mechanical strength index of the inner tube of the firework to be tested; Comprehensively analyze the basic quality strength index and mechanical strength index of the inner tube of the firework to be tested, and obtain the comprehensive quality strength index of the inner tube of the firework to be tested; The specific formula for calculating the comprehensive quality strength index of the inner tube of the firework to be tested is as follows: ； in, is the comprehensive quality strength index of the inner tube of the firework to be tested, is the basic quality strength index of the firework inner tube to be tested, is the basic quality coefficient stored in the database, is the mechanical strength index of the inner tube of the firework to be tested, is the mechanical strength coefficient stored in the database, , is a natural constant.

7. The method for detecting the quality of fireworks inner tubes used in fireworks production according to claim 6, characterized in that: The specific steps for obtaining the basic quality strength index of the inner tube of the firework to be tested are as follows: Obtaining the material density reference value, wall thickness reference value, material hardness reference value, and surface roughness reference value of the inner tube of the firework to be tested; The material density value, wall thickness value, material hardness value, and surface roughness value of each measuring position point of the firework inner tube to be tested are combined with the material density reference value, wall thickness reference value, material hardness reference value, and surface roughness reference value of the firework inner tube to be tested for a comprehensive analysis to obtain the basic quality strength index of the firework inner tube to be tested.

8. The method for detecting the quality of fireworks inner tubes used in fireworks production according to claim 7, characterized in that: The specific formula for calculating the basic mass strength index of the inner tube of the firework to be tested is as follows: ； in, is the basic quality strength index of the firework inner tube to be tested, For each of the fireworks inner tubes to be tested The material density value at each measuring point, The reference value of the material density of the inner tube of the firework to be tested, is the material density coefficient stored in the database, For each of the fireworks inner tubes to be tested The wall thickness value of each measuring point, The reference value of the wall thickness of the inner tube of the fireworks to be tested, is the wall thickness coefficient stored in the database, For each of the fireworks inner tubes to be tested The material hardness value at each measuring point, is the reference value of the material hardness of the inner tube of the firework to be tested, is the material hardness coefficient stored in the database, For each of the fireworks inner tubes to be tested The surface roughness value of each measuring point, is the reference value of the surface roughness of the inner tube of the firework to be tested, is the surface roughness coefficient stored in the database, , =1, 2, 3, …, , The number of measurement position points.

9. The method for detecting the quality of fireworks inner tubes used in fireworks production according to claim 6, characterized in that: The specific steps for obtaining the mechanical strength index of the inner tube of the firework to be tested are as follows: Obtaining a reference value of surface temperature, a reference value of strain, a reference value of vibration frequency, and a reference value of surface smoothness of the inner tube of the firework to be tested; The surface temperature value, strain value, vibration frequency value, and surface smoothness value of each measuring position of the inner tube of the firework to be tested are respectively combined with the surface temperature reference value, strain reference value, vibration frequency reference value, and surface smoothness reference value of the inner tube of the firework to be tested to obtain the mechanical strength index of the inner tube of the firework to be tested; The specific formula for calculating the mechanical strength index of the inner tube of the firework to be tested is as follows: ； in, is the mechanical strength index of the inner tube of the firework to be tested, For each of the fireworks inner tubes to be tested The surface temperature value of each measuring point, is the reference value of the surface temperature of the inner tube of the firework to be tested, is the surface temperature coefficient stored in the database, For each of the fireworks inner tubes to be tested The strain value of each measuring point, is the strain reference value of the inner tube of the firework to be tested, is the gauge factor stored in the database, For each of the fireworks inner tubes to be tested The vibration frequency value of each measuring point, is the reference value of the vibration frequency of the inner tube of the firework to be tested, is the vibration frequency coefficient stored in the database, For each of the fireworks inner tubes to be tested The surface smoothness value of each measuring position, is the reference value of the surface smoothness of the inner tube of the firework to be tested, is the surface smoothness coefficient stored in the database, , is a natural constant, =1, 2, 3, …, , The number of measurement position points.

10. A fireworks inner tube quality inspection system for fireworks production, using the fireworks inner tube quality inspection method for fireworks production according to any one of claims 1 to 9, characterized in that: include: Image intelligent analysis module, defect recognition module, image quality intelligent judgment module, comprehensive quality analysis module, comprehensive quality judgment module; The image intelligent analysis module is used to obtain image data of several image channels of the firework inner tube to be detected, and perform comprehensive analysis to obtain comprehensive image data of the firework inner tube to be detected, wherein the image data is specifically the pixel value of each pixel point in the firework inner tube image, and the comprehensive image data is specifically the comprehensive pixel value of each pixel point in the comprehensive image of the firework inner tube; The defect recognition module is used to input the comprehensive image data of the firework inner tube to be detected into the pre-trained neural network model for defect recognition analysis, obtain a plurality of hole areas and a plurality of crack areas in the comprehensive image of the firework inner tube to be detected, and perform comprehensive analysis to obtain the image quality index of the firework inner tube to be detected; The image quality intelligent judgment module is used to judge whether the image quality index of the firework inner tube to be detected is within a preset image quality threshold range, and if the image quality index of the firework inner tube to be detected is outside the preset image quality threshold range, the firework inner tube to be detected is marked as unqualified; The comprehensive quality analysis module is used to obtain the physical characteristic data of the firework inner tube to be detected when the image quality index of the firework inner tube to be detected is within a preset image quality threshold range, and perform comprehensive analysis to obtain the comprehensive quality strength index of the firework inner tube to be detected, wherein the physical characteristic data includes the material density value, wall thickness value, material hardness value, surface roughness value, surface temperature value, strain value, vibration frequency value, and surface smoothness value of each measurement position point; The comprehensive quality judgment module is used to judge whether the comprehensive quality strength index of the firework inner tube to be tested is within a preset comprehensive quality strength threshold range. If the comprehensive quality strength index of the firework inner tube to be tested is within the preset comprehensive quality strength threshold range, the firework inner tube to be tested is marked as qualified; if the comprehensive quality strength index of the firework inner tube to be tested is outside the preset comprehensive quality strength threshold range, the firework inner tube to be tested is marked as unqualified.

Images