An intelligent conveyor belt damage monitoring system and its multi-dimensional data analysis method

Abstract

The present invention discloses an intelligent conveyor belt damage monitoring system and a multi-dimensional data analysis method thereof, which relate to the technical field of industrial equipment maintenance, and solve the problems of low image clarity, poor damage detection accuracy and insufficient multi-dimensional data fusion of existing conveyor belt damage monitoring systems; the present invention comprises a data acquisition module, an image enhancement module, a multi-scale feature extraction module, a damage type classification and positioning module and a multi-dimensional data fusion and analysis module; the working status, image and environmental parameters of the conveyor belt are collected through a sensor network; the image is subjected to dynamic range compression, noise suppression and detail enhancement through the image enhancement module; the detailed features of the conveyor belt damage are captured through the multi-scale feature extraction module; the damage type is accurately classified through the damage type classification and positioning module; and a comprehensive fusion analysis is performed through the multi-dimensional data fusion and analysis module; the present invention greatly improves the accuracy and reliability of conveyor belt damage detection.

Description

Technical Field

[0001] The present invention relates to the technical field of industrial equipment maintenance, and more specifically to an intelligent conveyor belt damage monitoring system and a multi-dimensional data analysis method thereof. Background Art

[0002] With the advancement of Industry 4.0, smart manufacturing and Industrial Internet of Things technologies have gained widespread application. As a key piece of equipment in industrial production, the operating status of conveyor belts is directly related to the stability and efficiency of production lines. In recent years, with the advancement of sensor technology, big data analytics, and artificial intelligence, intelligent conveyor belt damage monitoring systems have gradually become a focus of industry attention. By monitoring the operating status of conveyor belts in real time and promptly identifying and warning of potential damage, they can effectively reduce downtime and repair costs, while improving production efficiency.

[0003] Existing intelligent conveyor belt damage monitoring systems primarily rely on a variety of sensors and data analysis technologies. Common sensors include visual sensors, vibration sensors, temperature sensors, and acoustic sensors. These sensors can collect real-time operating data on the conveyor belt. This data is then processed and analyzed using big data analytics and machine learning algorithms to identify and predict conveyor belt damage. For example, Chinese patent CN117274733A proposes inputting conveyor belt damage images into a well-trained conveyor belt damage detection model to perform conveyor belt damage detection, thereby improving the detection accuracy of the conveyor belt damage detection model. Chinese patent CN108510488B proposes a residual network-based conveyor belt damage detection method for detecting four types of conveyor belt damage. This method trains the residual network on image samples extracted from conveyor belt operation videos to achieve real-time detection of four types of conveyor belt damage: "tears," "scratches," "chipped edges," and "potholes." However, in practical applications, such as in underground coal mines, where weak and uneven lighting, dust, and high humidity are common, the captured images are low in clarity or even blurred, making tear damage segmentation difficult and affecting the accuracy of damage identification. Furthermore, existing damage detection algorithms may have inconsistent recognition effects for different types of damage (such as cracks, wear, and tears). In particular, when damage patterns are complex, they cannot integrate multiple dimensions of data, such as time series data, spatial distribution information, and physical properties, to generate valuable insights. This makes it difficult to accurately distinguish normal wear from potential faults, resulting in a high false alarm rate when processing complex damage patterns. Summary of the Invention

[0004] In response to the shortcomings of the existing technology, the present invention discloses an intelligent monitoring system for conveyor belt damage and a multi-dimensional data analysis method thereof, aiming to solve the problems of low image clarity, inaccurate damage type identification, and high false alarm rate caused by insufficient multi-dimensional data fusion in the existing intelligent monitoring system for conveyor belt damage.

[0005] In order to achieve the above technical effects, the present invention adopts the following technical solutions:

[0006] An intelligent conveyor belt damage monitoring system, comprising:

[0007] The data acquisition module is used to collect various types of conveyor belt data through the sensor network, including the working status of the conveyor belt, conveyor belt images, and environmental parameters;

[0008] An image enhancement module is configured to determine image clarity based on the conveyor belt image data collected by the data acquisition module using standard deviation and entropy. If the image clarity is lower than a preset threshold, the module performs dynamic range compression, noise suppression, and detail enhancement on the image using an adaptive Retinex algorithm and an adaptive threshold histogram equalization method, and outputs the enhanced image to the multi-scale feature extraction module. If the image clarity is higher than or equal to the preset threshold, the module directly outputs the image to the multi-scale feature extraction module.

[0009] A multi-scale feature extraction module is used to capture detailed features of conveyor belt damage, including cracks, wear, and tears, based on the output image of the image enhancement module through a pyramid pooling convolution structure; the pyramid pooling convolution structure includes an input layer, a depthwise separable convolution layer, a multi-scale pooling layer, a feature fusion layer, and an output layer;

[0010] a damage type classification and localization module, configured to classify and localize damage based on the output data of the multi-scale feature extraction module using an attention-based target detection algorithm; the attention-based target detection algorithm calculates the weight of each region in the image using the attention mechanism, and classifies the extracted feature vectors using a multi-classification Softmax layer to determine the damage type;

[0011] The multi-dimensional data fusion and analysis module is used to integrate the damage development rate, damage location distribution, material stress and material temperature data based on the collected data of the data acquisition module and the output data of the multi-scale feature extraction module through the graph neural network and the long short-term memory network, perform fusion analysis, and output a damage risk assessment report, including the damage type, degree, and development trend.

[0012] As a further technical solution of the present invention, the working method of the image enhancement module for performing dynamic range compression, noise suppression and detail enhancement on the image is as follows: first, the brightness perception characteristics of the human visual system are simulated by the adaptive Retinex algorithm to perform global and local contrast enhancement on the image, while suppressing the influence of uneven lighting; based on the enhanced image data, the histogram distribution of each small area is dynamically adjusted by the adaptive threshold histogram equalization method to adjust the local dark or bright areas; during the image processing process, the adaptive Retinex algorithm and the adaptive threshold histogram equalization method automatically adjust the parameters according to the actual lighting and noise conditions of the image to ensure the image enhancement effect under different lighting conditions.

[0013] As a further technical solution of the present invention, the working principle of the adaptive Retinex algorithm is as follows: based on the conveyor belt image data of the data acquisition module, the standard deviation and entropy of the image are calculated through the image processing library and mathematical statistical methods to quantify the clarity of the image; if the image clarity is lower than the preset threshold, the processing flow of the adaptive Retinex algorithm is triggered; if the image clarity meets or exceeds the preset threshold, the processing flow of the adaptive Retinex algorithm is skipped; when the image clarity is lower than the preset threshold, the image is decomposed into brightness components at multiple scales through the color restoration multi-scale method; the color restoration multi-scale method is based on the Gaussian pyramid model, and the image is smoothed by Gaussian filters of different scales to obtain brightness information at different scales; then, each scale is smoothed. The luminance components at different scales are logarithmically transformed. After the transformation, the adaptive Retinex algorithm fuses the logarithmic luminance components at each scale through a weighted averaging mechanism. During the fusion process, the contribution of different scales to image details is considered, and different weight values ​​are assigned to the luminance components at different scales through an adaptive gain adjustment mechanism to achieve global and local contrast enhancement of the image; wherein the weight value implements adaptive gain adjustment based on the local statistical characteristics of the image; the weight gain coefficient is automatically adjusted according to the local luminance and contrast information of each pixel in the image based on the local statistical characteristics of the image; after the gain adjustment, the adaptive Retinex algorithm recombines the enhanced luminance component with the color information of the original image through color space conversion to generate an enhanced image.

[0014] As a further technical solution of the present invention, the adjustment method of the adaptive threshold histogram equalization method is:

[0015] Step 401: Divide the image into a number of overlapping or non-overlapping small windows using a sliding window, with each window serving as an independent processing unit;

[0016] Step 402: Calculate the frequency distribution of pixel grayscale values ​​in each small area using a histogram statistical method to generate a local histogram;

[0017] Step 403: Calculate the brightness mean and variance of each small area based on the local histogram of each small area. If the grayscale value distribution in the small area is relatively concentrated, lower the preset threshold. If the grayscale value distribution in the small area is relatively dispersed, increase the preset threshold.

[0018] Step 404: Based on the dynamically modified threshold, the grayscale histogram is divided into low-grayscale regions below the threshold and high-grayscale regions above the threshold. The low-grayscale regions below the threshold and the high-grayscale regions above the threshold are subjected to histogram equalization using a local histogram equalization method. The local histogram equalization method maps the local histogram of each small region through local contrast stretching and grayscale value redistribution to generate a new grayscale value distribution.

[0019] Step 405: During the equalization process, if the grayscale value variation range in a certain area is small, grayscale merging is performed to merge some grayscale levels in the area into one grayscale level; if the grayscale value variation range in a certain area is large, grayscale subdivision is performed to further subdivide the grayscale levels in the area;

[0020] Step 406: stitch together the equalization results of each small area through image stitching to generate a final enhanced image.

[0021] As a further technical solution of the present invention, the working steps of the pyramid pooling convolution structure include:

[0022] Step 501: Receive the output image of the image enhancement module through the input layer;

[0023] Step 502: Perform preliminary feature extraction on the input image through the depthwise separable convolution layer; the depthwise separable convolution independently processes the spatial information and channel information of the image through depthwise convolution and pointwise convolution operations respectively;

[0024] Step 503: Perform a multi-scale pooling operation on the initially extracted feature map through the multi-scale pooling layer; the multi-scale pooling layer uses pooling windows of different sizes to downsample the feature map to obtain feature representations at different scales;

[0025] Step 504: The multi-scale features output by the multi-scale pooling layer are fused through the feature fusion layer; the feature fusion layer integrates the feature maps of different scales into a unified feature representation by using an element-by-element addition method;

[0026] Step 505: Output the feature map integrated by the feature fusion layer through the output layer as the final feature representation.

[0027] As a further technical solution of the present invention, the depth-wise separable convolution layer performs preliminary feature extraction on the input image as follows: first, an independent convolution kernel is assigned to each channel of the input image through spatial convolution to capture the spatial features within the channel and generate an intermediate feature map; then, the intermediate feature map generated by the spatial convolution is subjected to inter-channel feature fusion through a point-by-point convolution mechanism, and the point-by-point convolution uses a 1x1 convolution kernel to perform weighted summation of the feature values ​​of all channels at each position through linear transformation, and adds a bias term; in the process of depth-wise convolution and point-by-point convolution, each channel of the feature map is normalized through a batch normalization mechanism to normalize the distribution of the feature map and reduce the influence of internal covariate shift; on the feature map after point-by-point convolution, each element of the feature map is operated element-by-element through the nonlinear activation function ReLU. If the element value is greater than 0, it remains unchanged; if the element value is less than or equal to 0, it is set to 0.

[0028] As a further technical solution of the present invention, the working method of the target detection algorithm based on the attention mechanism is:

[0029] Step 701: pre-process the real-time conveyor belt image data collected by the data acquisition module through image enhancement, denoising and normalization processing;

[0030] Step 702: Perform multi-layer convolution operations on the pre-processed input image through a convolutional neural network to extract multi-level features of the image;

[0031] Step 703: Calculate the weight of each region in the image through the multi-scale attention mechanism to generate an attention map. The calculation formula is:

[0032]

[0033] In formula (1), n ​​represents the feature map output by the feature extraction network; G i (x) and G i (m) represents the global average pooling and global maximum pooling operations, respectively, used to capture global information; X i and M i Represent local average pooling and local maximum pooling operations respectively; i represents different pooling window sizes, which are used to capture multi-scale features; θ represents the inter-channel relationship vector, which is used to capture the correlation between channels; Represents the Sigmoid activation function, which is used to limit the value of the attention map to between 0 and 1; P i represents the splicing operation, which is used to integrate information between different scales and channels; μ cRepresents the convolution operation, which is used to further extract features and generate attention maps;

[0034] Step 704: Based on the attention map output in step 703, multiply the attention weight vector by the eigenvalue of the corresponding position in the feature map through an element-by-element multiplication operation to generate a weighted feature map;

[0035] Step 705: Based on the weighted feature map, the extracted feature vector is classified through a multi-classification Softmax layer to determine the damage type. The multi-classification Softmax layer maps the feature vector of the weighted feature map to a fixed-length vector space through a fully connected layer to generate a classification feature vector. The classification feature vector is normalized using an exponential function and a normalization operation to generate a probability distribution for each damage category.

[0036] Step 706: During the training process of the algorithm, the difference between the predicted result and the true label is calculated through the cross entropy loss function, and the back propagation algorithm is applied to pass the error back to each parameter in the network layer by layer.

[0037] As a further technical solution of the present invention, the attention mechanism is also used to guide the region proposal network to generate candidate regions, thereby improving the accuracy of damage localization. The specific working method is as follows: first, the weight of each region in the image is calculated through the attention mechanism, and the image features are weighted. Based on the feature map after weighted processing by the attention mechanism, a series of candidate regions are generated on the feature map by sliding several predefined anchor boxes on the feature map. In the process of sliding the anchor box, the foreground / background classification and bounding box regression are predicted for each anchor box through the convolution layer. In the process of generating candidate regions, a multi-scale feature map is also constructed through the feature pyramid network to strengthen the key feature association across scales.

[0038] As a further technical solution of the present invention, the working method of the multi-dimensional data fusion and analysis module is: using a graph neural network to process damage location distribution information, constructing a topological structure of conveyor belt damage, capturing the spatial correlation between damages, and processing the damage development rate through a long short-term memory network to capture the change trend of damage over time. Finally, the outputs of the graph neural network and the long short-term memory network are fused through a fully connected layer to generate a damage risk assessment report, including the damage type, degree, and development trend.

[0039] As a further technical solution of the present invention, a multi-dimensional data analysis method for a conveyor belt damage intelligent monitoring system includes the following steps:

[0040] S1. Process damage location distribution information through a graph neural network. The graph neural network constructs a topological structure of conveyor belt damage through nodes and edges, where each node represents a damage point and the edges represent the spatial relationship between damage points. The graph neural network also aggregates the features of each node through a graph convolution operation, so that the feature representation of each node not only includes its own attributes but also integrates the influence of adjacent nodes to form a topological structure diagram of the conveyor belt damage. The formula for constructing the topological structure diagram of the conveyor belt damage through the graph neural network is:

[0041]

[0042] S2. In formula (2), represents the feature representation of the kth node in the l+1 layer; σ represents the ReLU nonlinear activation function; N(k) represents the neighbor set of node k; c kj represents the normalization coefficient, which is used to balance the contributions of different neighbor nodes; W l Represents the weight matrix of the lth layer, which is used to learn the relationship between node features; represents the feature representation of the kth neighbor node in the lth layer; b (l) is the bias term of the lth layer, which is used to adjust the baseline of feature representation;

[0043] S3. Processing the damage progression rate through a long short-term memory network to capture the damage change trend over time; the long short-term memory network captures the damage change trend over time through input gates, forget gates, output gates, and cell states, and dynamically adjusts the degree of memory of historical information;

[0044] S4. After completing spatial correlation capture and temporal trend analysis, the spatial features output by the graph neural network and the temporal features output by the long short-term memory network are integrated through the weight matrix and nonlinear activation function of the fully connected layer to generate a comprehensive feature vector z.

[0045] S5. Based on the comprehensive feature vector z output by the fully connected layer, the comprehensive feature vector is normalized by the Softmax function to generate the probability distribution of each damage category. The formula expression of the Softmax function is:

[0046]

[0047] In formula (3), TAU k is the probability of the kth category; z k represents the kth element of the comprehensive feature vector; C is the total number of categories; z j Represents the j-th element of the comprehensive feature vector.

[0048] Positive beneficial effects:

[0049] 1. Through the image enhancement module, utilizing an adaptive Retinex algorithm and adaptive threshold histogram equalization, this invention can effectively improve the clarity of conveyor belt images captured in harsh environments such as low light and high dust. Detailed features in the image, such as cracks, wear, and tears, can be clearly displayed in the enhanced image, even if they are blurred in the original image. This directly solves the difficulty in tear damage segmentation caused by low image clarity and significantly improves the accuracy of damage identification.

[0050] 2. Through the pyramid pooling convolutional structure of the multi-scale feature extraction module, this invention can capture the detailed features of conveyor belt damage at different scales. This not only improves the recognition accuracy of single damage types (such as cracks and wear), but also significantly enhances the ability to identify complex damage patterns (such as the coexistence of multiple damage types). Through the combination of depthwise separable convolutional layers and multi-scale pooling layers, the system can extract rich feature information from different scales, providing a solid foundation for subsequent damage classification and location.

[0051] 3. The damage type classification and localization module utilizes an attention-based object detection algorithm. By calculating the weight of each region in the image, it can accurately locate the damage location and determine its type. This solves the problem of inconsistent recognition results of existing damage detection algorithms for complex damage patterns and significantly reduces the false alarm rate. Furthermore, a multi-classification Softmax layer ensures accurate damage type classification, enabling the system to reliably distinguish between normal wear and potential failures.

[0052] 3. The multi-dimensional data fusion and analysis module integrates data from multiple dimensions, including damage growth rate, damage location distribution, material stress, and material temperature, and utilizes graph neural networks and long-short-term memory networks for fusion analysis. This not only enables a comprehensive assessment of conveyor belt damage but also improves its accuracy. By capturing the spatial and temporal trends of damage, the system can more accurately predict damage development trends, providing timely warning information to operators and maintenance personnel, effectively avoiding potential failures. Description of the drawings:

[0053] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. Those skilled in the art can also derive other drawings based on these drawings without inventive work, among which:

[0054] Figure 1 This is an architecture diagram of an intelligent monitoring system for conveyor belt damage according to the present invention;

[0055] Figure 2 A diagram showing the steps of adjusting the histogram equalization method with an adaptive threshold according to the present invention;

[0056] Figure 3 This is a schematic diagram of the processing flow principle of the adaptive Retinex algorithm of the present invention;

[0057] Figure 4 A diagram showing the working method of the target detection algorithm based on the attention mechanism of the present invention;

[0058] Figure 5 This is a flow chart of the working method of the multi-dimensional data fusion and analysis module of the present invention. DETAILED DESCRIPTION

[0059] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0060] like Figure 1 As shown, an intelligent monitoring system for conveyor belt damage includes: a data acquisition module for collecting various types of conveyor belt data through a sensor network, including the working status of the conveyor belt, conveyor belt images, and environmental parameters; in specific implementations, the data acquisition module uses a variety of sensors, first of all, a photoelectric sensor, a pulse-modulated active photoelectric detection system-type electronic sensor, and infrared light detection. The photoelectric sensors are installed on both sides below the conveyor belt, and their lateral positions are arranged to detect the maximum damage on the belt edge. By reflecting and receiving infrared light, the photoelectric sensor can quickly detect edge damage of the conveyor belt in a non-contact and non-destructive manner;

[0061] There are also visual sensors (cameras) used to capture real-time images of the conveyor belt. The vertical distance from the conveyor belt is determined by the camera's focal length, ensuring that the camera area covers the conveyor belt to ensure image continuity and integrity. Lighting equipment is installed on both sides of the conveyor belt at a 30-degree angle from vertical. The cameras use timed shooting or continuous video streaming to capture cracks, wear, tears, and other damage characteristics on the conveyor belt surface. The image data is transmitted in real time to the data processing center via network transmission or wired connection.

[0062] The system also includes a longitudinal tear sensor, which detects longitudinal tears in the conveyor belt. Installed below the belt conveyor's material receiving point, it ensures detection of foreign objects penetrating the conveyor belt. When the conveyor belt is penetrated by foreign objects and the material falls onto the sensor, the sensor sends a signal to the control box, triggering an alarm and a stop signal.

[0063] Second, the system also includes temperature and humidity sensors, pressure sensors, and accelerometers. The temperature and humidity sensors monitor the temperature and humidity of the conveyor belt and its surroundings, assessing thermal and moisture stresses in materials and environmental conditions. They are installed at key locations on the conveyor belt, such as bearings, motors, and the belt surface, as well as at key locations in the environment, such as on the sides and top of the belt. Pressure sensors monitor the load and pressure distribution on the conveyor belt and assess its stress conditions. Pressure sensors are installed on the conveyor belt's support structure, such as idlers and rollers, as well as key bearing points. The pressure sensors periodically read pressure values ​​to collect data on the load and pressure distribution at various locations on the conveyor belt, and transmit this data in real time to a data processing center via wireless or wired means. Accelerometers monitor the vibration and motion of the conveyor belt and assess its operational stability. Accelerometers are installed on the conveyor belt's drive unit, bearings, and support structure. They periodically read acceleration values ​​to collect data on the vibration amplitude and frequency at various locations on the conveyor belt, and transmit this data in real time to a data processing center via wireless or wired means.

[0064] All sensors transmit the collected data to the data processing center in real time via wired or wireless networks (such as Wi-Fi, LoRa, Zigbee, etc.). The data processing center stores and pre-processes the received data to provide a basis for subsequent feature extraction and data analysis. The specific content of the acquired data includes: conveyor belt image data, including high-definition images of the conveyor belt surface, which are used to detect damage characteristics such as cracks, wear, and tears; temperature data, including temperature values ​​of key parts such as the conveyor belt surface, bearings, and motors, which are used to evaluate the thermal stress and environmental conditions of the material; humidity data, including humidity values ​​of key parts such as the conveyor belt surface, bearings, and motors, which are used to evaluate the wet stress and environmental conditions of the material; pressure data, including load and pressure distribution values ​​of the conveyor belt at different positions, which are used to evaluate the force condition of the conveyor belt; acceleration data, including vibration amplitude and frequency values ​​of the conveyor belt at different positions, which are used to evaluate the operational stability of the conveyor belt.

[0065] An image enhancement module is used to judge the image clarity based on the conveyor belt image data collected by the data acquisition module by standard deviation and entropy. If the image clarity is lower than a preset threshold, the image is subjected to dynamic range compression, noise suppression and detail enhancement by an adaptive Retinex algorithm and an adaptive threshold histogram equalization method, and the enhanced image is output to a multi-scale feature extraction module. If the image clarity is higher than or equal to the preset threshold, the image is directly output to the multi-scale feature extraction module. The working method of the image enhancement module for performing dynamic range compression, noise suppression and detail enhancement on the image is as follows: first, the brightness perception characteristics of the human visual system are simulated by an adaptive Retinex algorithm to enhance the global and local contrast of the image while suppressing the influence of uneven illumination. Based on the enhanced image data, the histogram distribution of each small area is dynamically adjusted by an adaptive threshold histogram equalization method to adjust the local dark or bright areas. During the image processing process, the adaptive Retinex algorithm and the adaptive threshold histogram equalization method automatically adjust the parameters according to the actual illumination and noise conditions of the image to ensure the enhancement effect of the image under different illumination conditions.

[0066] In the specific implementation, after starting the processing flow of the adaptive Retinex algorithm, such as Figure 3 As shown, the module uses the adaptive Retinex algorithm to perform global and local contrast enhancement on the image. It then performs global and local contrast enhancement to improve image quality. It then checks illumination and noise, and suppresses any uneven illumination. If the image requires further adjustment, it performs adaptive threshold histogram equalization. Adaptive threshold histogram equalization adjusts the histogram distribution based on the local area of ​​the image. Based on the results of histogram equalization, the image histogram distribution is dynamically adjusted. Finally, the image enhancement is rechecked to see if the desired effect has been achieved. If so, the process ends. Otherwise, it returns to the adaptive Retinex algorithm step and reprocesses.

[0067] The working principle of the adaptive Retinex algorithm is as follows: based on the conveyor belt image data of the data acquisition module, the standard deviation and entropy of the image are calculated through the image processing library and mathematical statistical methods to quantify the clarity of the image; if the image clarity is lower than the preset threshold, the processing flow of the adaptive Retinex algorithm is triggered; if the image clarity meets or exceeds the preset threshold, the processing flow of the adaptive Retinex algorithm is skipped; when the image clarity is lower than the preset threshold, the image is decomposed into brightness components at multiple scales through the color restoration multi-scale method; the color restoration multi-scale method is based on the Gaussian pyramid model, and the image is smoothed by Gaussian filters of different scales to obtain brightness information at different scales; then the brightness component at each scale is logarithmically transformed. After the transformation, the adaptive Retinex algorithm is used. The Retinex algorithm fuses the logarithmic brightness components at each scale through a weighted averaging mechanism. During the fusion process, the contribution of different scales to image details is taken into account, and different weight values ​​are assigned to the brightness components at different scales through an adaptive gain adjustment mechanism to enhance the global and local contrast of the image. The weight value is based on the local statistical characteristics of the image to achieve adaptive gain adjustment. The weight gain coefficient is automatically adjusted based on the local brightness and contrast information of each pixel in the image based on the local statistical characteristics of the image. After the gain adjustment, the adaptive Retinex algorithm recombines the enhanced brightness component with the color information of the original image through color space conversion to generate an enhanced image. In a specific implementation, the image clarity judgment is to quantify the image quality by calculating the standard deviation and entropy of the image. The standard deviation reflects the contrast of the image, and the entropy reflects the amount of information in the image. The calculation formula of the standard deviation is:

[0068]

[0069] In formula (1), S i is the gray value of the i-th pixel in the image; μ represents the average gray value of the image; N represents the total number of pixels in the image; the entropy calculation formula is:

[0070]

[0071] Among them, p i is the probability that a pixel with gray value i appears in the image, and L is the number of gray levels of the image.

[0072] In practical applications, the image's standard deviation and entropy are first calculated using an image processing library (such as OpenCV). If these values ​​fall below a preset threshold, the image is deemed insufficiently sharp and requires enhancement. Otherwise, the image is directly output to the multi-scale feature extraction module. This approach quickly determines image quality, ensuring that only images requiring enhancement enter the subsequent processing steps, improving system efficiency.

[0073] The multi-scale color restoration method is an image enhancement technique based on multi-scale analysis. It utilizes a Gaussian pyramid model to decompose an image into brightness components at different scales. Each brightness component reflects the image's brightness information at different resolutions. This method captures detailed image features while suppressing noise and interference. In the intelligent conveyor belt damage monitoring system, the multi-scale color restoration method is used to pre-process collected conveyor belt images. By decomposing images into brightness components at multiple scales, the algorithm can more accurately identify damage features such as cracks and wear.

[0074] The logarithmic transformation is a nonlinear transformation that compresses the brightness values ​​of an image, so that images with a wider brightness range are more evenly distributed within the new brightness range after the transformation. This transformation helps enhance image contrast and make detailed features more prominent. In the adaptive Retinex algorithm, a logarithmic transformation is used to process the brightness component at each scale. This transformation allows the algorithm to better capture detailed features in the image and suppress the effects of uneven lighting.

[0075] The weighted averaging mechanism is a data fusion method based on weight assignment. It assigns a weight to each data point based on its reliability or importance. The data is then fused by calculating the weighted average. This method preserves key information while suppressing noise and interference. In the adaptive Retinex algorithm, the weighted averaging mechanism is used to fuse logarithmic brightness components at different scales. By adaptively adjusting the weight gain coefficients based on the local statistical characteristics of the image, the algorithm can enhance both global and local contrast in the image.

[0076] In practice, the adaptive Retinex algorithm automatically adjusts its parameters based on the actual image illumination and noise conditions, ensuring effective image enhancement under varying lighting conditions. This makes the algorithm more adaptable in complex environments, such as those found in coal mines, where uneven illumination, dust, and humidity are common. Compared to traditional fixed-parameter methods, the adaptive Retinex algorithm is more adaptable to diverse application scenarios and improves image enhancement. Furthermore, through multi-scale decomposition and logarithmic transformation, the adaptive Retinex algorithm captures both global and local features in the image, enhancing the dynamic range and detail. This helps improve the accuracy and reliability of damage detection, particularly for identifying detailed features such as cracks, wear, and tears. Compared to single-scale methods, multi-scale processing captures more comprehensive image features, improving damage detection accuracy. Secondly, the use of a Gaussian filter effectively removes high-frequency noise from the image, preserving key structural information and enhancing image clarity and contrast. This helps reduce false alarms and improves damage detection accuracy. Compared to traditional noise suppression methods, the Gaussian filter more effectively removes noise while preserving image detail, improving image quality. Through color space conversion, the Adaptive Retinex algorithm enhances image brightness and contrast while preserving color information, ensuring a natural and realistic visual effect. Compared to traditional grayscale enhancement methods, color preservation better preserves the visual quality of images, improving user satisfaction. Furthermore, through multi-scale decomposition and weighted average fusion, the Adaptive Retinex algorithm reduces computational effort and memory consumption while maintaining image quality, improving the algorithm's real-time performance and applicability. Compared to complex image enhancement algorithms, the Adaptive Retinex algorithm offers significant advantages in computational efficiency and resource utilization, making it suitable for real-time monitoring and large-scale data processing.

[0077] like Figure 2As shown: the adjustment method of the adaptive threshold histogram equalization method is as follows: step 401, divide the image into several overlapping or non-overlapping small windows through a sliding window, and each window is used as an independent processing unit; step 402, calculate the frequency distribution of the pixel grayscale value in each small area through a histogram statistical method to generate a local histogram; step 403, calculate the brightness mean and variance of each small area according to the local histogram of each small area, if the grayscale value distribution in the small area is relatively concentrated, then lower the preset threshold, if the grayscale value distribution in the small area is relatively dispersed, then increase the preset threshold; step 404, based on the dynamically changed threshold, divide the grayscale histogram into a low grayscale area below the threshold and a high grayscale area above the threshold In step 405, if the grayscale value variation range in a certain area is small, grayscale levels in the area are merged into one grayscale level by grayscale merging; if the grayscale value variation range in a certain area is large, grayscale levels in the area are further subdivided by grayscale subdivision; and in step 406, the equalization results of each small area are stitched together by image stitching to generate a final enhanced image.

[0078] Sliding window partitioning involves dividing the entire image into multiple small regions, which can be overlapping or non-overlapping. Each small region acts as an independent processing unit, enabling more precise processing of different parts of the image, particularly those with varying lighting conditions. This ensures more accurate processing results for each small region, thereby improving overall image quality. For example, in areas with uneven lighting, a smaller window can be selected for more detailed processing. This partitioning method facilitates subsequent local histogram statistics and equalization, resulting in more accurate processing results for each small region. Local histogram statistics involve calculating the grayscale values ​​of pixels within each small window to generate a local histogram for each window. A histogram displays the frequency of occurrence of each grayscale value, reflecting the distribution of pixels within the window. Histogram statistics provide information about brightness and contrast within each small window, providing a basis for subsequent threshold adjustment and histogram equalization. In practical applications, these histograms help us understand the brightness and contrast within each small window, providing important reference information for subsequent processing steps. For example, if the grayscale values ​​within a small window are concentrated, this indicates low contrast in that area and requires enhancement. In practical applications, the threshold is dynamically adjusted by calculating the brightness mean and variance for each small window. For example, if the grayscale values ​​within a small window are concentrated, this indicates low contrast in that area, and the threshold should be lowered to enhance contrast. Conversely, if the grayscale values ​​are dispersed, this indicates high contrast in that area, and the threshold should be raised to avoid over-enhancement. This dynamic adjustment ensures more balanced processing for each small window, improving overall image quality. Local histogram equalization equalizes the histogram within each small window, enhancing image details by expanding local contrast. Specifically, it redistributes the grayscale values ​​within each small window, making the frequency of each grayscale value more uniform, thereby improving local contrast and detail. Grayscale merging and subdivision adjust the grayscale values ​​within each small window during the equalization process based on the grayscale value range within each small window. If the grayscale value variation range in a small window is small, it means that the contrast of the area is low, and the amount of calculation can be reduced by merging some grayscale levels; if the grayscale value variation range is large, it means that the contrast of the area is high, and the detail performance can be improved by subdividing the grayscale levels.

[0079] Compared with existing technologies, the adaptive threshold histogram equalization method, through local processing and dynamic threshold adjustment, can more accurately capture detailed features in images while suppressing noise and interference, thereby improving the image enhancement effect. Secondly, this method can automatically adjust parameters based on the actual lighting and noise conditions of the image, ensuring the image enhancement effect under different lighting conditions and improving the robustness and stability of the system. Furthermore, this method further optimizes the distribution of grayscale values ​​through grayscale merging or subdivision, making the image grayscale values ​​more uniform and continuous, which facilitates subsequent damage feature extraction and classification.

[0080] In the practical application of the intelligent conveyor belt damage monitoring system, the hardware working environment of the adaptive Retinex algorithm and the adaptive threshold histogram equalization method includes:

[0081] Computing platform: A high-performance computer or server equipped with a multi-core CPU (such as Intel Xeon E5 series) and high-speed memory (at least 16GB DDR4 RAM) for quickly processing large amounts of image data.

[0082] Graphics Processing Unit: A dedicated GPU (such as the NVIDIA Tesla V100) that is used to accelerate the execution of image processing algorithms, especially for complex image processing tasks such as adaptive Retinex and histogram equalization.

[0083] Storage devices: Large-capacity solid-state drives (SSDs) or RAID arrays are used to store large amounts of raw and processed image data, ensuring fast and reliable data access.

[0084] Image acquisition device: High-resolution industrial camera (such as Basler ace series) is used to capture high-quality images of the conveyor belt surface, ensuring image clarity and rich details.

[0085] Network equipment: Gigabit Ethernet switches are used to connect various components to ensure high-speed and stable data transmission.

[0086] Power supply: Uninterruptible power supply (UPS) ensures that the system can continue to operate normally in the event of power failure and protects data security.

[0087] Based on the above hardware environment, a comparative experiment was designed to verify the effectiveness of the adaptive Retinex algorithm combined with the adaptive threshold histogram equalization method (Group A) compared with the traditional global histogram equalization (Group B) in the conveyor belt damage intelligent monitoring system; in this experiment, the traditional algorithm used by Group B is Global Histogram Equalization (GHE). GHE is a simple image enhancement technology that changes the brightness value distribution of each pixel in the image to make the brightness value distribution of the entire image more uniform, thereby achieving the purpose of enhancing the image contrast. Description of the experimental method

[0088] The experiment first randomly selected five conveyor belt samples with varying degrees of damage from a real-world working environment, capturing 10 images of each sample. These images were then processed using methods from Group A and Group B, with each method repeated five times for each sample to assess the stability and consistency of the methods. Finally, the effectiveness of the two methods was analyzed by comparing image contrast, signal-to-noise ratio, and damage identification accuracy before and after processing. The experimental records are shown in Table 1.

[0089] Table 1 Comparative experimental record

[0090]

[0091]

[0092] Experimental results show that the adaptive Retinex algorithm combined with adaptive threshold histogram equalization (Group A) significantly outperforms traditional global histogram equalization (Group B) in terms of average contrast improvement, average signal-to-noise ratio improvement, and damage identification accuracy. Group A's method better adapts to varying lighting conditions, effectively enhancing local image contrast while reducing noise amplification, improving the accuracy and reliability of damage detection. This demonstrates that Group A's method is more suitable for intelligent conveyor belt damage monitoring systems and can provide more accurate damage detection support for actual production.

[0093] The multi-scale feature extraction module is used to capture the detailed features of conveyor belt damage based on the output image of the image enhancement module through a pyramid pooling convolution structure, and the damage includes cracks, wear and tear; the pyramid pooling convolution structure includes an input layer, a depthwise separable convolution layer, a multi-scale pooling layer, a feature fusion layer and an output layer; the working steps of the pyramid pooling convolution structure include: step 501, receiving the output image of the image enhancement module through the input layer; step 502, performing preliminary feature extraction on the input image through the depthwise separable convolution layer; the depthwise separable convolution performs depthwise convolution and pointwise convolution operations to separate the features. The spatial information and channel information of the image are processed independently; in step 503, a multi-scale pooling operation is performed on the initially extracted feature map through the multi-scale pooling layer; the multi-scale pooling layer uses pooling windows of different sizes to downsample the feature map to obtain feature representations at different scales; in step 504, the multi-scale features output by the multi-scale pooling layer are fused through the feature fusion layer; the feature fusion layer uses an element-by-element addition method to integrate feature maps of different scales into a unified feature representation; in step 505, the feature map integrated by the feature fusion layer is output as the final feature representation through the output layer.

[0094] Among them, the extraction method of the depth-wise separable convolution layer for preliminary feature extraction of the input image is as follows: first, an independent convolution kernel is assigned to each channel of the input image through spatial convolution to capture the spatial features within the channel and generate an intermediate feature map; then, the intermediate feature map generated by the spatial convolution is subjected to inter-channel feature fusion through a point-by-point convolution mechanism, and the point-by-point convolution uses a 1x1 convolution kernel to perform weighted summation of the feature values ​​of all channels at each position through linear transformation, and adds a bias term; in the process of depth-wise convolution and point-by-point convolution, each channel of the feature map is normalized through a batch normalization mechanism to normalize the distribution of the feature map and reduce the influence of internal covariate shift; on the feature map after point-by-point convolution, each element of the feature map is operated element-by-element through the nonlinear activation function ReLU. If the element value is greater than 0, it remains unchanged; if the element value is less than or equal to 0, it is set to 0;

[0095] In its implementation, the depthwise separable convolution layer utilizes independent processing of depthwise convolution and pointwise convolution to effectively reduce computational complexity and improve feature extraction efficiency. Depthwise convolution focuses on extracting spatial features from each input channel, while pointwise convolution combines features from different channels to generate a new feature map. This separate processing approach enables the network to more efficiently capture the spatial and channel information of the image. The multi-scale pooling layer downsamples the feature map using pooling windows of varying sizes to obtain feature representations at different scales. This multi-scale feature extraction method captures detailed features of conveyor belt damage (such as cracks, wear, and tears) at different scales, improving feature richness and robustness. The feature fusion layer uses element-by-element addition to integrate feature maps at different scales into a unified feature representation. This fusion approach not only preserves feature information at each scale but also complements and enhances features, providing more accurate and comprehensive feature input for subsequent damage identification and analysis.

[0096] In practical applications, the pyramid pooling convolution structure can significantly improve the performance of the intelligent conveyor belt damage monitoring system. First, through the efficient feature extraction of the depthwise separable convolution layer, the system can quickly capture key information in the conveyor belt image. Subsequently, the multi-scale pooling layer performs multi-scale downsampling on the feature map to capture the detailed features of the damage at different scales. These features are integrated and enhanced in the feature fusion layer and ultimately output as a unified feature representation. This working method enables the system to more accurately identify and analyze damage on the conveyor belt. Because the pyramid pooling convolution structure can capture feature information at multiple scales, the system has a good ability to recognize damage of different sizes and shapes. In addition, due to the use of the depthwise separable convolution layer, the computational efficiency of the system has also been significantly improved, making real-time monitoring and rapid response possible.

[0097] Compared to existing technologies, the pyramid pooling convolutional architecture captures richer damage details through multi-scale feature extraction and fusion, improving recognition accuracy and robustness. Furthermore, the use of depthwise separable convolutional layers significantly improves the system's computational efficiency, reducing computational costs and power consumption. Furthermore, the pyramid pooling convolutional architecture exhibits strong generalization capabilities, adapting to conveyor belt damage monitoring needs in diverse environments and conditions.

[0098] In the practical application of the intelligent conveyor belt damage monitoring system and its multi-dimensional data analysis method, the hardware working environment of the pyramid pooling convolution structure mainly includes:

[0099] Computing platform: NVIDIA Jetson TX2 embedded computing platform with high-performance GPU and CPU, capable of supporting real-time execution of complex deep learning algorithms.

[0100] Camera: Use an industrial-grade high-definition camera with a resolution of 1080p and a frame rate of 30fps, which can capture clear conveyor belt images.

[0101] Storage device: Equipped with high-speed SSD solid-state hard drive to store large amounts of image data and model parameters.

[0102] Power supply: Use a stable DC power adapter to ensure that the system will not be interrupted by power problems during long-term operation.

[0103] Network communication: Data transmission with the monitoring center via wired or wireless networks (such as Wi-Fi) to achieve remote monitoring and management.

[0104] Based on the above hardware environment, a set of comparative experiments was designed to verify the effectiveness and superiority of the pyramid pooling convolutional structure in the intelligent conveyor belt damage monitoring system. The experiments were divided into two groups: Group A used the pyramid pooling convolutional structure, and Group B used traditional single-scale feature extraction methods (such as SIFT). Both groups of experiments were conducted under the same hardware environment to ensure consistency of experimental conditions. Each group of experiments was conducted five times, and the damage detection accuracy and processing time were recorded for each experiment. The experimental results are shown in Table 2:

[0105] Table 2 Experimental record table

[0106]

[0107] Through comparative experiments, it can be seen that the pyramid pooling convolution structure (Group A) outperforms the traditional SIFT algorithm (Group B) in both accuracy and processing time for conveyor belt damage detection. Specifically, the average accuracy of Group A is 92.8%, while the average accuracy of Group B is 85.6%; the average processing time of Group A is 120 milliseconds, while the average processing time of Group B is 249 milliseconds. This shows that the pyramid pooling convolution structure not only significantly improves detection accuracy, but also shows obvious advantages in processing speed. Therefore, the pyramid pooling convolution structure has a positive and beneficial effect in the conveyor belt damage intelligent monitoring system, and can effectively improve the performance and practicality of the system.

[0108] A damage type classification and positioning module is used to classify and locate damage based on the output data of the multi-scale feature extraction module through an attention-based target detection algorithm; the attention-based target detection algorithm calculates the weight of each area in the image through the attention mechanism, and classifies the extracted feature vectors through a multi-classification Softmax layer to determine the damage type; Figure 4As shown: The working method of the target detection algorithm based on the attention mechanism is as follows: Step 701, pre-processing the real-time conveyor belt image data collected by the data acquisition module through image enhancement, denoising and normalization processing; Step 702, performing multi-layer convolution operations on the pre-processed input image through a convolutional neural network to extract multi-level features of the image; Step 703, calculating the weight of each area in the image through a multi-scale attention mechanism to generate an attention map, and the calculation formula is:

[0109]

[0110] In formula (1), n ​​represents the feature map output by the feature extraction network; G i (x) and G i (m) represents the global average pooling and global maximum pooling operations, respectively, used to capture global information; X i and M i Represent local average pooling and local maximum pooling operations respectively; i represents different pooling window sizes, which are used to capture multi-scale features; θ represents the inter-channel relationship vector, which is used to capture the correlation between channels; Represents the Sigmoid activation function, which is used to limit the value of the attention map to between 0 and 1; P i represents the splicing operation, which is used to integrate information between different scales and channels; μ cRepresents a convolution operation, which is used to further extract features and generate an attention map; Step 704, based on the attention map output by Step 703, multiply the attention weight vector by the eigenvalue of the corresponding position of the feature map through an element-by-element multiplication operation to generate a weighted feature map; Step 705, based on the weighted feature map, classify the extracted feature vector through a multi-classification Softmax layer to determine the damage type; the multi-classification Softmax layer maps the feature vector of the weighted feature map to a vector space of a fixed length through a fully connected layer to generate a classification feature vector; and normalizes the classification feature vector through an exponential function and a normalization operation to generate a probability distribution of each damage category; Step 706, during the training process of the algorithm, the cross entropy loss function is used to calculate the probability distribution of each damage category. The attention mechanism is used to calculate the difference between the predicted result and the true label, and the back-propagation algorithm is used to pass the error back to each parameter in the network layer by layer. The attention mechanism is also used to guide the region proposal network to generate candidate regions, thereby improving the accuracy of damage localization. The specific working method is as follows: first, the weight of each region in the image is calculated through the attention mechanism, and the image features are weighted. Based on the feature map after the weighted processing of the attention mechanism, a series of candidate regions are generated on the feature map by sliding several predefined anchor boxes on the feature map. During the sliding anchor box process, the foreground / background classification and bounding box regression prediction are performed on each anchor box through the convolution layer. During the candidate region generation process, a multi-scale feature map is constructed through the feature pyramid network to strengthen the correlation of key features across scales.

[0111] In its implementation, the attention mechanism captures the global context of the image through global average pooling and global max pooling, while simultaneously focusing on local details in the image through local average pooling and local max pooling. This dual consideration of global and local factors enables the algorithm to more comprehensively understand image content. The introduction of inter-channel relationship vectors further enhances the algorithm's ability to identify key features. By calculating the correlation between different channels, the algorithm automatically adjusts the weights of each channel to optimize the feature extraction process. The sigmoid activation function, a nonlinear mapping tool for the attention map, constrains attention values ​​between 0 and 1, ensuring the algorithm's stability and interpretability during processing. Furthermore, feature extraction networks, particularly convolutional neural networks (CNNs), extract multi-layered image features through multi-layer convolution operations. These features encompass rich visual information, from edges and textures to shapes, providing a solid foundation for subsequent damage classification and localization. The construction of a feature pyramid network (FPN) enhances the algorithm's ability to identify damage regions of different scales by fusing cross-scale features, improving both robustness and accuracy.

[0112] In practical applications, the target detection algorithm based on the attention mechanism first pre-processes the real-time conveyor belt image data collected by the data acquisition module through image enhancement, denoising and normalization. Then, the convolutional neural network is used to perform multi-layer convolution operations on the pre-processed image to extract the multi-level features of the image. Next, the weight of each area in the image is calculated through the attention mechanism to generate an attention map, and the feature map is weighted. Based on the weighted feature map, the algorithm classifies the extracted feature vectors through the multi-classification Softmax layer to determine the damage type. At the same time, the region proposal network generates candidate regions based on the weighted feature map, and constructs a multi-scale feature map through the feature pyramid network to further improve the accuracy of damage location. Finally, the algorithm outputs information such as damage type, location and probability distribution for subsequent multi-dimensional data analysis;

[0113] Compared with existing technologies, the algorithm automatically focuses on damaged areas by weighted image processing through an attention mechanism, reducing interference from background information. Combined with the feature extraction and fusion capabilities of CNN and FPN, the algorithm extracts richer image features and improves recognition accuracy. Furthermore, the combination of the attention mechanism, CNN, and FPN enables the algorithm to flexibly adapt to different types of conveyor belts and damage conditions, maintaining high recognition accuracy and stability even in complex environments. Furthermore, by optimizing the algorithm's structure and parameters, the algorithm is able to improve processing speed while ensuring recognition accuracy, enabling real-time monitoring.

[0114] In practical applications, the method of generating candidate regions using a region proposal network based on an attention mechanism can significantly improve the accuracy of damage localization. The specific steps are as follows:

[0115] 1. Image preprocessing: First, the collected conveyor belt images are preprocessed, including image enhancement, denoising and normalization, to ensure that the image quality meets the requirements of subsequent processing.

[0116] 2. Feature Extraction: A multi-layer convolutional neural network is used to extract features from the preprocessed image, generating multi-level feature maps. These feature maps contain both local and global information about the image, providing a foundation for the subsequent attention mechanism and region proposal network.

[0117] 3. Attention map generation: Generate attention maps through multi-scale pooling and convolution operations. The specific steps are:

[0118] Global average pooling and global maximum pooling operations capture global information.

[0119] Local average pooling and local maximum pooling operations capture local information.

[0120] Capture the correlation between channels by modeling the relationship between channels.

[0121] The Sigmoid activation function is used to clamp the values ​​of the attention map between 0 and 1.

[0122] The information between different scales and channels is integrated through the splicing operation.

[0123] The convolution operation is used to further extract features and generate attention maps.

[0124] 4. Feature weighting: The generated attention map is element-wise multiplied with the feature map to generate a weighted feature map. The weighted feature map can highlight the key areas in the image and suppress irrelevant areas, thereby improving the robustness and discriminability of the features.

[0125] 5. Candidate Region Generation: Based on the weighted feature map, a region proposal network is used to generate candidate regions. The specific method is as follows: Predefined anchor boxes are slid across the weighted feature map to generate a series of candidate regions. A convolutional layer then performs foreground / background classification and bounding box regression on each anchor box. Foreground / background classification determines whether the anchor box contains damage, while bounding box regression accurately locates the damage.

[0126] 6. Multi-scale Feature Map Construction: A feature pyramid network is used to construct a multi-scale feature map, strengthening the correlation of key features across scales. The specific method is as follows: First, a top-down path is used to transfer high-level semantic information to lower layers. Then, horizontal connections are performed to combine low-level spatial information with high-level semantic information. Finally, a multi-scale feature map is generated to capture damage characteristics at different scales.

[0127] In the practical application of the intelligent monitoring system for conveyor belt damage and its multi-dimensional data analysis method, the hardware working environment of the target detection algorithm based on the attention mechanism includes: Computing platform: Using the NVIDIA Jetson Xavier NX embedded computing platform, it has a high-performance GPU and multi-core CPU, which can support the real-time operation of complex deep learning algorithms. Camera: Using an industrial-grade high-definition camera with a resolution of 4K and a frame rate of 30fps, it can capture high-definition conveyor belt images. Storage device: Equipped with a high-speed NVMe SSD solid-state drive to store large amounts of image data and model parameters. Power supply: Use a stable DC power adapter to ensure that the system will not be interrupted due to power problems during long-term operation. Network communication: Data transmission with the monitoring center is carried out through wired or wireless networks (such as 5G) to achieve remote monitoring and management.

[0128] A set of comparative experiments was designed to verify the effectiveness and superiority of the target detection algorithm based on the attention mechanism in the conveyor belt damage intelligent monitoring system. The experiments were divided into two groups: Group A used the target detection algorithm based on the attention mechanism, and Group B used the traditional single-scale feature extraction method (Faster R-CNN). Both groups of experiments were carried out under the same hardware environment to ensure the consistency of the experimental conditions. Each group of experiments was conducted five times, and the damage detection accuracy and processing time of each experiment were recorded. Faster R-CNN is a classic two-stage target detection algorithm that generates candidate regions through a region proposal network (RPN) and then performs feature extraction and classification through a convolutional neural network. Although Faster R-CNN performs well in many scenarios, its performance may be limited when dealing with complex backgrounds and multi-scale features. The experimental data record table is shown in Table 3:

[0129] Table 3 Comparative experimental record table

[0130]

[0131] Through comparative experiments, it can be seen that the target detection algorithm based on the attention mechanism (Group A) outperforms the traditional Faster R-CNN algorithm (Group B) in both accuracy and processing time for conveyor belt damage detection. Specifically, the average accuracy of Group A is 94.6%, while the average accuracy of Group B is 88.3%; the average processing time of Group A is 150 milliseconds, while the average processing time of Group B is 202 milliseconds. This shows that the target detection algorithm based on the attention mechanism not only significantly improves the detection accuracy, but also shows obvious advantages in processing speed. Therefore, the target detection algorithm based on the attention mechanism has a positive and beneficial effect in the conveyor belt damage intelligent monitoring system, and can effectively improve the performance and practicality of the system.

[0132] A multi-dimensional data fusion and analysis module is used to integrate the damage development rate, damage location distribution, material stress, and material temperature data based on the collected data of the data acquisition module and the output data of the multi-scale feature extraction module through a graph neural network and a long short-term memory network, perform fusion analysis, and output a damage risk assessment report, including the damage type, degree, and development trend. The working method of the multi-dimensional data fusion and analysis module is as follows: using a graph neural network to process the damage location distribution information, constructing the topological structure of the conveyor belt damage, capturing the spatial correlation between the damages, and processing the damage development rate through a long short-term memory network to capture the change trend of the damage over time. Finally, the outputs of the graph neural network and the long short-term memory network are fused through a fully connected layer to generate a damage risk assessment report, including the damage type, degree, and development trend.

[0133] Furthermore, a multi-dimensional data analysis method for an intelligent monitoring system for conveyor belt damage includes the following steps: S1, processing damage location distribution information through a graph neural network; the graph neural network constructs a topological structure of the conveyor belt damage through nodes and edges, where each node represents a damage point and the edge represents the spatial relationship between the damage points; the graph neural network also aggregates the features of each node through a graph convolution operation, so that the feature representation of each node not only includes its own attributes but also integrates the influence of adjacent nodes to form a topological structure diagram of the conveyor belt damage; the formula expression for constructing the topological structure diagram of the conveyor belt damage by the graph neural network is:

[0134]

[0135] S2. In formula (2), represents the feature representation of the kth node in the l+1 layer; σ represents the ReLU nonlinear activation function; N(k) represents the neighbor set of node k; c kj represents the normalization coefficient, which is used to balance the contributions of different neighbor nodes; W l Represents the weight matrix of the lth layer, which is used to learn the relationship between node features; represents the feature representation of the kth neighbor node in the lth layer; b (l) is the bias term of the lth layer, which is used to adjust the baseline of the feature representation; S3, the damage development rate is processed through the long short-term memory network to capture the change trend of the damage over time; the long short-term memory network captures the change trend of the damage over time through the input gate, forget gate, output gate and cell state, and dynamically adjusts the memory degree of historical information; S4, after completing the spatial association capture and time trend analysis, the spatial features output by the graph neural network and the time features output by the long short-term memory network are integrated through the weight matrix and nonlinear activation function of the fully connected layer to generate a comprehensive feature vector z, S5, based on the comprehensive feature vector z output by the fully connected layer, the comprehensive feature vector is normalized by the Softmax function to generate the probability distribution of each damage category; the formula expression of the Softmax function is:

[0136]

[0137] In formula (3), TAU k is the probability of the kth category; z k represents the kth element of the comprehensive feature vector; C is the total number of categories; z j Represents the j-th element of the comprehensive feature vector.

[0138] In specific implementation, in actual applications, the multi-dimensional data analysis method works as follows: first, the damage data of the conveyor belt is obtained through the data acquisition module and the multi-scale feature extraction module; then, the graph neural network is used to process the damage location distribution information and construct the topological structure of the conveyor belt damage; then, the damage development rate is processed through the long short-term memory network to capture the change trend of the damage over time; finally, the two types of information are fused through the fully connected layer to generate a damage risk assessment report.

[0139] In a graph neural network, each node not only represents a damage point but also connects to adjacent nodes via edges, forming a topological structure representing the damage on the conveyor belt. This structure is not static but continuously updates with the input of new data. The feature representation of each node continuously aggregates information from adjacent nodes through graph convolution operations. This dynamic update mechanism enables the network to capture complex spatial correlations between damages. The core of the graph convolution operation is to update the feature representation of a node by aggregating the features of adjacent nodes. This operation not only considers the node's own properties but also incorporates the influence of adjacent nodes, thereby enabling awareness of local topological structure. Through multiple layers of graph convolution operations, the network can capture connections between nodes at greater distances and form a global representation of the topological structure. Long short-term memory networks use three gating mechanisms: input gate, forget gate, and output gate to fine-tune the flow of information. The input gate determines which new information enters the cell state, the forget gate determines which old information is forgotten, and the output gate determines which information is output to the next layer. This fine-tuning mechanism enables the network to dynamically adjust the degree of memory for historical information, thereby enabling accurate processing of time series data. Cell state is a core component of the LSTM network, responsible for storing the network's long-term memory. By updating and transferring cell state, the network maintains a memory of past information and uses it when needed for current task decisions. This long-term memory mechanism enables the network to capture temporal trends in damage, providing powerful support for damage risk assessment.

[0140] During application implementation, the multi-dimensional data analysis method adopts a data-driven approach in actual applications, continuously updating the model's parameters and status based on newly collected data. This dynamic update mechanism enables the system to adapt to changes in the conveyor belt environment and maintain stable performance. When processing real-time data, the system needs to maximize processing speed while ensuring accuracy. To achieve this goal, the system uses efficient algorithms and hardware acceleration technologies, such as GPU parallel computing, to ensure a balance between real-time performance and accuracy. In addition, to facilitate users to view and use damage risk assessment reports, the system provides a user-friendly interactive interface. Users can intuitively understand the damage status, development trends, and potential risk points of the conveyor belt through the interface, providing timely decision-making support for industrial safety production.

[0141] The beneficial effects of this working method are mainly reflected in the following aspects: First, the accuracy of damage identification is improved. By integrating multi-dimensional data, the system can more comprehensively understand the damage status of the conveyor belt. Second, the robustness of the system is enhanced. Even in the face of complex and changing conveyor belt environments, the system can maintain stable performance. Third, the timeliness of damage risk assessment is improved. The system can update the damage risk assessment report in real time, providing timely warning and decision-making support for industrial safety production.

[0142] Compared with existing technologies, the multidimensional data analysis method has the following advantages and characteristics: First, higher accuracy. By integrating and deeply analyzing multidimensional data, the system can more accurately assess the damage status of conveyor belts. Second, greater robustness. The system can adapt to complex and changing conveyor belt environments and maintain stable performance. Third, a more comprehensive damage risk assessment. The system can not only provide basic information such as damage type and extent, but also predict the development trend of damage, providing more comprehensive decision-making support for industrial safety production. Fourth, greater timeliness. The system can update damage risk assessment reports in real time, providing timely warnings and responses for industrial safety production. These advantages and characteristics make the multidimensional data analysis method have broad application prospects and important practical value in conveyor belt damage intelligent monitoring systems.

[0143] In specific implementation, the hardware working environment of the multi-dimensional data analysis method in the conveyor belt damage intelligent monitoring system mainly includes: High-performance computing server: used to run complex graph neural network and long short-term memory network models to ensure real-time processing and analysis of a large amount of collected conveyor belt damage data. Data acquisition equipment: including sensors, cameras, etc., used to monitor the damage of the conveyor belt in real time and transmit the data to the computing server for processing. Data storage device: such as hard disk array or cloud storage service, used to store historical data and model parameters for long-term analysis and model optimization. Visual terminal: used to display damage risk assessment reports and real-time monitoring results to facilitate user decision-making and response.

[0144] Based on the above hardware environment, a comparative experiment was conducted using a multi-dimensional data analysis method (Group A) and a traditional algorithm (Group B). The traditional algorithm uses a rule-based approach, which determines the damage of the conveyor belt according to preset rules and thresholds. This method is simple and intuitive, but lacks flexibility and accuracy. During the experiment, both groups were conducted under the same experimental conditions, with each group conducting five experiments. The damage identification accuracy, processing time, and system stability indicators of each experiment were recorded. The experimental data record table is shown in Table 4:

[0145] Table 4. Multi-dimensional data analysis method experimental comparison data record table

[0146]

[0147]

[0148] Comparative experiments show that the multidimensional data analysis method (Group A) outperforms the traditional algorithm (Group B) in terms of damage identification accuracy, processing time, and system stability. Group A achieved an average damage identification accuracy of 94.6%, compared to only 83.6% for Group B. The average processing time for Group A was only 2.4 seconds, significantly lower than Group B's 3.7 seconds. Furthermore, Group A's system stability score remained consistently at 5 points, while Group B's score fluctuated. Therefore, the multidimensional data analysis method offers greater accuracy and efficiency in intelligent conveyor belt damage monitoring systems, better meeting the needs of industrial safety production.

[0149] Although specific embodiments of the present invention have been described above, those skilled in the art will appreciate that these specific embodiments are merely illustrative, and that those skilled in the art may omit, substitute, and modify the details of the methods and systems described above without departing from the principles and spirit of the present invention. For example, combining the above method steps to perform substantially the same functions in substantially the same manner to achieve substantially the same results falls within the scope of the present invention. Accordingly, the scope of the present invention is limited solely by the appended claims.

Claims

1. An intelligent conveyor belt damage monitoring system; characterized by: The system comprises: The data acquisition module is used to collect various types of conveyor belt data through the sensor network, including the working status of the conveyor belt, conveyor belt images, and environmental parameters; An image enhancement module is configured to determine image clarity based on the conveyor belt image data collected by the data acquisition module using standard deviation and entropy. If the image clarity is lower than a preset threshold, the module performs dynamic range compression, noise suppression, and detail enhancement on the image using an adaptive Retinex algorithm and an adaptive threshold histogram equalization method, and outputs the enhanced image to the multi-scale feature extraction module. If the image clarity is higher than or equal to the preset threshold, the module directly outputs the image to the multi-scale feature extraction module. A multi-scale feature extraction module is used to capture detailed features of conveyor belt damage, including cracks, wear, and tears, based on the output image of the image enhancement module through a pyramid pooling convolution structure; the pyramid pooling convolution structure includes an input layer, a depthwise separable convolution layer, a multi-scale pooling layer, a feature fusion layer, and an output layer; a damage type classification and localization module, configured to classify and localize damage based on the output data of the multi-scale feature extraction module using an attention-based target detection algorithm; the attention-based target detection algorithm calculates the weight of each region in the image using the attention mechanism, and classifies the extracted feature vectors using a multi-classification Softmax layer to determine the damage type; The multi-dimensional data fusion and analysis module is used to integrate the damage development rate, damage location distribution, material stress and material temperature data based on the collected data of the data acquisition module and the output data of the multi-scale feature extraction module through the graph neural network and the long short-term memory network, perform fusion analysis, and output a damage risk assessment report, including the damage type, degree, and development trend.

2. The intelligent conveyor belt damage monitoring system according to claim 1, characterized in that: The image enhancement module performs dynamic range compression, noise suppression, and detail enhancement on images as follows: first, an adaptive Retinex algorithm is used to simulate the brightness perception characteristics of the human visual system to enhance the global and local contrast of the image while suppressing the influence of uneven lighting; based on the enhanced image data, an adaptive threshold histogram equalization method is used to dynamically adjust the histogram distribution of each small area to adjust local areas that are too dark or too bright; during the image processing process, the adaptive Retinex algorithm and the adaptive threshold histogram equalization method automatically adjust parameters according to the actual lighting and noise conditions of the image to ensure the image enhancement effect under different lighting conditions.

3. The intelligent conveyor belt damage monitoring system according to claim 2, characterized in that: The working principle of the adaptive Retinex algorithm is as follows: based on the conveyor belt image data of the data acquisition module, the standard deviation and entropy of the image are calculated through the image processing library and mathematical statistical methods to quantify the clarity of the image; if the image clarity is lower than the preset threshold, the processing flow of the adaptive Retinex algorithm is triggered; if the image clarity meets or exceeds the preset threshold, the processing flow of the adaptive Retinex algorithm is skipped; when the image clarity is lower than the preset threshold, the image is decomposed into brightness components at multiple scales through the color restoration multi-scale method; the color restoration multi-scale method is based on the Gaussian pyramid model, and the image is smoothed by Gaussian filters of different scales to obtain brightness information at different scales; then the brightness components at each scale are smoothed. A logarithmic transformation is performed. After the transformation, the adaptive Retinex algorithm fuses the logarithmic brightness components at each scale through a weighted averaging mechanism. During the fusion process, the contribution of different scales to image details is considered, and different weight values ​​are assigned to the brightness components at different scales through an adaptive gain adjustment mechanism to achieve global and local contrast enhancement of the image; wherein the weight value is based on the local statistical characteristics of the image to achieve adaptive gain adjustment; the weight gain coefficient is automatically adjusted according to the local brightness and contrast information of each pixel in the image based on the local statistical characteristics of the image; after the gain adjustment, the adaptive Retinex algorithm recombines the enhanced brightness component with the color information of the original image through color space conversion to generate an enhanced image.

4. The intelligent conveyor belt damage monitoring system according to claim 2, characterized in that: The adjustment method of the adaptive threshold histogram equalization method is: Step 401: Divide the image into a number of overlapping or non-overlapping small windows using a sliding window, with each window serving as an independent processing unit; Step 402: Calculate the frequency distribution of pixel grayscale values ​​in each small area using a histogram statistical method to generate a local histogram; Step 403: Calculate the brightness mean and variance of each small area based on the local histogram of each small area. If the grayscale value distribution in the small area is relatively concentrated, lower the preset threshold. If the grayscale value distribution in the small area is relatively dispersed, increase the preset threshold. Step 404: Based on the dynamically modified threshold, the grayscale histogram is divided into low-grayscale regions below the threshold and high-grayscale regions above the threshold. The low-grayscale regions below the threshold and the high-grayscale regions above the threshold are subjected to histogram equalization using a local histogram equalization method. The local histogram equalization method maps the local histogram of each small region through local contrast stretching and grayscale value redistribution to generate a new grayscale value distribution. Step 405: During the equalization process, if the grayscale value variation range in a certain area is small, grayscale merging is performed to merge some grayscale levels in the area into one grayscale level; if the grayscale value variation range in a certain area is large, grayscale subdivision is performed to further subdivide the grayscale levels in the area; Step 406: stitch together the equalization results of each small area through image stitching to generate a final enhanced image.

5. The intelligent conveyor belt damage monitoring system according to claim 1, characterized in that: The working steps of the pyramid pooling convolution structure include: Step 501: Receive the output image of the image enhancement module through the input layer; Step 502: Perform preliminary feature extraction on the input image through the depthwise separable convolution layer; the depthwise separable convolution independently processes the spatial information and channel information of the image through depthwise convolution and pointwise convolution operations respectively; Step 503: Perform a multi-scale pooling operation on the initially extracted feature map through the multi-scale pooling layer; the multi-scale pooling layer uses pooling windows of different sizes to downsample the feature map to obtain feature representations at different scales; Step 504: The multi-scale features output by the multi-scale pooling layer are fused through the feature fusion layer; the feature fusion layer integrates the feature maps of different scales into a unified feature representation by using an element-by-element addition method; Step 505: Output the feature map integrated by the feature fusion layer through the output layer as the final feature representation.

6. The intelligent conveyor belt damage monitoring system according to claim 5, characterized in that: The extraction method of the depthwise separable convolution layer for performing preliminary feature extraction on the input image is as follows: first, an independent convolution kernel is assigned to each channel of the input image through spatial convolution to capture the spatial features within the channel and generate an intermediate feature map; then, the intermediate feature map generated by the spatial convolution is subjected to inter-channel feature fusion through a point-by-point convolution mechanism, and the point-by-point convolution uses a 1x1 convolution kernel to perform weighted summation of the feature values ​​of all channels at each position through linear transformation, and adds a bias term; in the process of depthwise convolution and point-by-point convolution, each channel of the feature map is normalized through a batch normalization mechanism to normalize the distribution of the feature map and reduce the influence of internal covariate shift; on the feature map after point-by-point convolution, each element of the feature map is operated element-by-element through the nonlinear activation function ReLU, and if the element value is greater than 0, it remains unchanged; If the element value is less than or equal to 0, it is set to 0.

7. The intelligent conveyor belt damage monitoring system according to claim 1, characterized in that: The working method of the target detection algorithm based on the attention mechanism is: Step 701: pre-process the real-time conveyor belt image data collected by the data acquisition module through image enhancement, denoising and normalization processing; Step 702: Perform multi-layer convolution operations on the pre-processed input image through a convolutional neural network to extract multi-level features of the image; Step 703: Calculate the weight of each region in the image through the multi-scale attention mechanism to generate an attention map. The calculation formula is: （1） In formula (1), The feature map representing the output of the feature extraction network; and Represent global average pooling and global maximum pooling operations, respectively, to capture global information; and Represent local average pooling and local maximum pooling operations respectively; Represents different pooling window sizes to capture multi-scale features; Represents the inter-channel relationship vector, which is used to capture the correlation between channels; Represents the Sigmoid activation function, which is used to limit the value of the attention map to between 0 and 1; Represents the splicing operation, which is used to integrate information between different scales and channels; Represents the convolution operation, which is used to further extract features and generate attention maps; Step 704: Based on the attention map output in step 703, multiply the attention map by the eigenvalues ​​at the corresponding positions of the feature map through an element-by-element multiplication operation to generate a weighted feature map; Step 705: Based on the weighted feature map, the extracted feature vector is classified through a multi-classification Softmax layer to determine the damage type. The multi-classification Softmax layer maps the feature vector of the weighted feature map to a fixed-length vector space through a fully connected layer to generate a classification feature vector. The classification feature vector is normalized using an exponential function and a normalization operation to generate a probability distribution for each damage category. Step 706: During the training process of the algorithm, the difference between the predicted result and the true label is calculated through the cross entropy loss function, and the back propagation algorithm is applied to pass the error back to each parameter in the network layer by layer.

8. The intelligent conveyor belt damage monitoring system according to claim 1, characterized in that: The attention mechanism is also used to guide the region proposal network to generate candidate regions, thereby improving the accuracy of damage localization. The specific working method is as follows: first, the weight of each region in the image is calculated through the attention mechanism, and the image features are weighted. Based on the feature map after weighted processing by the attention mechanism, a series of candidate regions are generated on the feature map by sliding several predefined anchor boxes on the feature map. In the process of sliding the anchor box, the foreground / background classification and bounding box regression are predicted for each anchor box through the convolution layer. In the process of generating candidate regions, a multi-scale feature map is also constructed through the feature pyramid network to strengthen the key feature association across scales.

9. The intelligent conveyor belt damage monitoring system according to claim 1, characterized in that: The working method of the multi-dimensional data fusion and analysis module is as follows: using graph neural networks to process damage location distribution information, constructing the topological structure of conveyor belt damage, capturing the spatial correlation between damages, and processing the damage development rate through long short-term memory networks to capture the change trend of damage over time. Finally, the outputs of the graph neural network and the long short-term memory network are fused through a fully connected layer to generate a damage risk assessment report, including the damage type, degree, and development trend.

10. A multi-dimensional data analysis method for a conveyor belt damage intelligent monitoring system, characterized by: An intelligent conveyor belt damage monitoring system according to any one of claims 1 to 9 comprises the following steps: S1. Process damage location distribution information through a graph neural network. The graph neural network constructs a topological structure of conveyor belt damage through nodes and edges, where each node represents a damage point and the edges represent the spatial relationship between damage points. The graph neural network also aggregates the features of each node through a graph convolution operation, so that the feature representation of each node not only includes its own attributes but also integrates the influence of adjacent nodes to form a topological structure diagram of the conveyor belt damage. The formula for constructing the topological structure diagram of the conveyor belt damage through the graph neural network is: （2） In formula (2), Indicates the The node in Feature representation of the layer; Represents the ReLU nonlinear activation function; Representation node The set of neighbors of Represents the normalization coefficient, which is used to balance the contributions of different neighbor nodes; Indicates the The weight matrix of the layer is used to learn the relationship between node features; Indicates the The node in Feature representation of the layer; For the The bias term of the layer is used to adjust the baseline of the feature representation; S2. Processing the damage progression rate through a long short-term memory network to capture the changing trend of damage over time. The long short-term memory network captures the changing trend of damage over time through input gates, forget gates, output gates, and cell states, and dynamically adjusts the degree of memory of historical information. S3. After completing spatial correlation capture and temporal trend analysis, the spatial features output by the graph neural network and the temporal features output by the long short-term memory network are integrated through the weight matrix and nonlinear activation function of the fully connected layer to generate a comprehensive feature vector , S4, comprehensive feature vector based on the output of the fully connected layer ,The comprehensive feature vector is normalized by the Softmax function to generate the probability distribution of each damage category.

Images