On-site safety management and control system and method for drilling machine construction

Abstract

The invention discloses an on-site safety management and control system and method for drilling machine construction, and the system comprises an image collection module, a construction feature learning module and a safety management and control classification module, the image collection module is used for obtaining a pre-image of a drilling machine construction site through an image collection device according to an image collection instruction and a preset sampling frequency, based on a preset image verification algorithm, performing image processing on the pre-image of the drilling machine construction site to determine an initial image of the drilling machine construction site; the construction feature learning module is used for performing feature training on the initial image by using a convolutional neural network learning model to obtain target features of a drilling machine construction site; and the safety management and control classification module is used for identifying and classifying the images according to the target features of the construction site, and identifying unsafe construction behaviors in the construction area. Digitization of safety supervision work of the drilling machine construction site is achieved, safety management efficiency and accuracy of the drilling machine construction site are improved, and labor cost is reduced.

Description

Technical Field

[0001] The present invention belongs to the technical field of safety control, and particularly relates to a field safety control system and method for drilling rig construction. Background Art

[0002] In modern building construction, as an important construction equipment, drilling rigs are widely used in multiple fields such as foundation treatment, mine exploitation, and tunnel excavation. However, with the expansion of the construction scale and the complexity of the construction environment, the safety management of the drilling rig construction site faces severe challenges. The traditional manual safety supervision method is not only inefficient but also difficult to comprehensively cover various safety hazards in the construction process. Especially in a complex and changeable construction environment, manual supervision often has omissions, resulting in frequent safety accidents, which has become a technical problem that urgently needs to be solved. Summary of the Invention

[0003] The purpose of the present invention is to provide a field safety control system and method for drilling rig construction to solve the deficiencies in the prior art. It aims to realize the digitization of the safety supervision work at the drilling rig construction site through an intelligent field safety control system, improve the efficiency and accuracy of the safety management at the drilling rig construction site, and reduce the labor cost.

[0004] An embodiment of the present application provides a field safety control system for drilling rig construction, and the system includes: An image acquisition module, a construction feature learning module, and a safety control classification module that are communicatively connected. Among them, The image acquisition module is used to obtain a pre-image of the drilling rig construction site through an image acquisition device according to an image acquisition instruction at a preset sampling frequency, and perform image processing on the pre-image of the drilling rig construction site based on a preset image verification algorithm to determine an initial image of the drilling rig construction site; The construction feature learning module is used to perform feature training on the initial image by using a convolutional neural network learning model to obtain target features of the drilling rig construction site; among them, the convolutional neural network learning model includes two layers of multi-scale input preprocessing units, three layers of deep feature extraction units, one layer of scale adaptive pooling unit, and two layers of feature mapping and generalization units; The multi-scale input preprocessing unit is used to receive image inputs of different scales and perform preprocessing operations; The deep feature extraction unit is used to gradually extract deep features in the image through three layers of convolution and activation operations; The scale adaptive pooling unit is used to adaptively adjust the size of the pooling window according to the scale of the input image; The feature mapping and generalization unit is used to adjust the mapping of feature dimensions using a 2x2 convolutional layer, and integrate global features through a fully connected layer to generate target features of the drill rig construction site; The safety control and classification module is used to identify and classify images based on the target features of the construction site, and identify unsafe construction behaviors in the construction area.

[0005] Optionally, performing image processing on the pre-image of the drill rig construction site based on a preset image verification algorithm to determine the initial image of the drill rig construction site, including: Based on the preset image verification algorithm, perform threshold segmentation on the pre-image of the drill rig construction site, perform distance transformation on the threshold segmentation result to obtain a distance transformation result; Use the symmetric algorithm of hyperbolic operation to perform confusion calculation on the distance transformation result to obtain a confusion calculation result; Generate the initial image of the drill rig construction site according to the confusion calculation result.

[0006] Optionally, after generating the initial image of the drill rig construction site according to the confusion calculation result, further include: Obtain an image division matrix with the same size as the initial image; Perform scale division on the image division matrix to obtain a number of first image sub-division matrices; Perform re-division on the first image sub-division matrix to obtain multiple second image sub-division matrices; Perform local coding on the first image sub-division matrix and global coding on the second image sub-division matrix to achieve multi-scale division of the initial image.

[0007] Optionally, gradually extract deep features in the image through three-layer convolution and activation operations, which are represented by the following formula: Among them, represents the first-layer convolution and activation operation, represents the activation function, represents the preprocessed input image, represents the first convolution kernel, represents the first bias term; Among them, represents the second-layer convolution and activation operation, represents the activation function, represents the second convolution kernel, represents the second bias term; Among them, Represents the third - layer convolution and activation operations, Represents the activation function, Represents the third convolution kernel, Represents the third bias term.

[0008] Optionally, the mapping adjustment of the feature dimension is performed using a 2x2 convolutional layer, and the global features are integrated through a fully - connected layer to generate the target features of the drill rig construction site, which is represented by the following formula: Among them, Represents the target features of the drill rig construction site, Represents the weight matrix of the fully - connected layer, Represents the output after the operation of the 2x2 convolutional layer, satisfying , among which, Represents the output after passing through the scale - adaptive pooling unit, Represents the convolution kernel of the feature mapping and generalization unit, Represents the fourth bias term, Represents the operation of flattening into a one - dimensional vector, Represents the fifth bias term.

[0009] Another embodiment of the present application provides a method for on - site safety control of drill rig construction, and the method includes: According to the image acquisition instruction, at a preset sampling frequency, obtain the pre - image of the drill rig construction site through an image acquisition device; Based on a preset image verification algorithm, perform image processing on the pre - image of the drill rig construction site to determine the initial image of the drill rig construction site; Use a convolutional neural network learning model to perform feature training on the initial image to obtain the target features of the drill rig construction site; According to the target features, identify and classify the image to identify unsafe construction behaviors in the construction area.

[0010] Another embodiment of the present application provides a storage medium, in which a computer program is stored, and the computer program is set to implement the above - mentioned method when running.

[0011] Another embodiment of the present application provides an electronic device, including a memory and a processor, a computer program is stored in the memory, and the processor is set to run the computer program to implement the above - mentioned method.

[0012] Compared with the prior art, the present application includes an image acquisition module, a construction feature learning module, and a safety control classification module connected by communication. Among them, the image acquisition module is used to obtain a pre-image of the drill rig construction site through an image acquisition device according to an image acquisition instruction at a preset sampling frequency, and perform image processing on the pre-image of the drill rig construction site based on a preset image verification algorithm to determine an initial image of the drill rig construction site; the construction feature learning module is used to perform feature training on the initial image using a convolutional neural network learning model to obtain target features of the drill rig construction site; the safety control classification module is used to identify and classify images based on the target features of the construction site and identify unsafe construction behaviors in the construction area. It realizes the digitization of the safety supervision work at the drill rig construction site through an intelligent on-site safety control system, improves the efficiency and accuracy of safety management at the drill rig construction site, and reduces labor costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 FIG. is a hardware structure block diagram of a computer terminal for a method for on-site safety control of drill rig construction provided by an embodiment of the present invention; Figure 2 FIG. is a schematic structural diagram of a system for on-site safety control of drill rig construction provided by an embodiment of the present invention; Figure 3 FIG. is a schematic flow diagram of a method for on-site safety control of drill rig construction provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0014] The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention and should not be construed as limiting the present invention.

[0015] An embodiment of the present invention first provides a method for on-site safety control of drill rig construction, which can be applied to an electronic device, such as a computer terminal, specifically, a general computer, a tablet, etc.

[0016] The following will take running on a computer terminal as an example for a detailed description. Figure 1 FIG. is a hardware structure block diagram of a computer terminal for a method for on-site safety control of drill rig construction provided by an embodiment of the present invention. As Figure 1 shown, the computer device includes a processor, a memory, and a network interface connected by a system bus. Among them, the memory may include a non-volatile storage medium and an internal memory.

[0017] The non-volatile storage medium can store an operating system and a computer program. The computer program includes program instructions, and when the program instructions are executed, the processor can execute any method for on-site safety control of drill rig construction.

[0018] The processor is used to provide computing and control capabilities to support the operation of the entire computer device.

[0019] The internal memory provides an environment for the operation of the computer program in the non-volatile storage medium. When the computer program is executed by the processor, the processor can be caused to execute any one of the on-site safety control methods for drilling rig construction.

[0020] The network interface is used for network communication, such as sending the assigned tasks, etc. Those skilled in the art can understand that Figure 1 the structure shown in is only a block diagram of some structures related to the solution of this application, and does not constitute a limitation on the computer device to which the solution of this application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine some components, or have different component arrangements.

[0021] It should be understood that the processor may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field programmable gate arrays (FPGAs) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0022] See Figure 2 , Figure 2Schematic diagram of the structure of a on-site safety control system for drilling rig construction provided by an embodiment of the present invention. The on-site safety control system 200 for drilling rig construction may include an image acquisition module 201, a construction feature learning module 202, and a safety control classification module 203 that are communicatively connected. Among them, the image acquisition module 201 is configured to obtain a pre-image of the drilling rig construction site through an image acquisition device according to an image acquisition instruction at a preset sampling frequency, and perform image processing on the pre-image of the drilling rig construction site based on a preset image verification algorithm to determine an initial image of the drilling rig construction site; the construction feature learning module 202 is configured to perform feature training on the initial image by using a convolutional neural network learning model to obtain target features of the drilling rig construction site; among them, the convolutional neural network learning model includes two layers of multi-scale input preprocessing units, three layers of deep feature extraction units, one layer of scale adaptive pooling units, and two layers of feature mapping and generalization units; the multi-scale input preprocessing unit is configured to receive image inputs of different scales and perform preprocessing operations; the deep feature extraction unit is configured to gradually extract deep features in the image through three layers of convolution and activation operations; the scale adaptive pooling unit is configured to adaptively adjust the size of the pooling window according to the scale of the input image; the feature mapping and generalization unit is configured to use a 2x2 convolutional layer to perform mapping adjustment of the feature dimension, and integrate global features through a fully connected layer to generate target features of the drilling rig construction site; the safety control classification module is configured to identify and classify images according to the target features of the construction site, and identify unsafe construction behaviors in the construction area.

[0023] Specifically, the on-site safety control system 200 for this drilling rig construction aims to perform real-time monitoring on the drilling rig construction site through advanced image acquisition and analysis technologies, timely identify unsafe construction behaviors, and ensure construction safety. The system consists of an image acquisition module 201, a construction feature learning module 202, and a safety control classification module 203 that are communicatively connected. Among them, the image acquisition module 201 can obtain a pre-image of the drilling rig construction site through an image acquisition device according to an image acquisition instruction at a preset sampling frequency, and perform image processing on the pre-image based on a preset image verification algorithm to determine an initial image of the drilling rig construction site. Receive image acquisition instructions, which can be triggered by the system at a fixed time or manually issued. According to the preset sampling frequency, control the image acquisition device (such as a camera) to shoot the drilling rig construction site and obtain a pre-image. Input the pre-image into the preset image verification algorithm module, and the algorithm verifies the clarity and integrity of the image, and removes unqualified images such as blur and damage. Perform pre-processing operations such as enhancement and noise reduction on the qualified pre-image, finally determine the initial image, and transmit it to the construction feature learning module 202. The construction feature learning module 202 uses a convolutional neural network learning model to perform feature training on the initial image to obtain the target features of the drilling rig construction site. The convolutional neural network learning model includes two layers of multi-scale input pre-processing units, three layers of deep feature extraction units, one layer of scale adaptive pooling units, and two layers of feature mapping and generalization units. Among them, the multi-scale input pre-processing unit is used to receive the initial image from the image acquisition module 201, which is input at different scales, and then performs pre-processing operations such as normalization and standardization on the images of different scales to make the image data meet the subsequent processing requirements. The deep feature extraction unit is used to receive the image processed by the multi-scale input pre-processing unit. Through three layers of convolution and activation operations, the image is sequentially feature extracted. Each layer of convolution operation can extract features of different levels and complexities in the image. As the number of layers increases, deep features are gradually extracted. The scale adaptive pooling unit is used to automatically adjust the size of the pooling window according to the scale of the input image. The image processed by the deep feature extraction unit is pooled to reduce the amount of data while retaining the main features and improve processing efficiency. The feature mapping and generalization unit uses a 2x2 convolution layer to adjust the dimension mapping of the features processed by the scale adaptive pooling unit. The adjusted features are integrated through the fully connected layer to generate global features, that is, the target features of the drilling rig construction site, and transmitted to the safety control classification module 203. The safety control classification module 203 is used to identify and classify images according to the target features of the construction site, and identify unsafe construction behaviors in the construction area. The module first receives the target features from the construction feature learning module 202, and then compares and analyzes the target features with the predefined unsafe construction behavior feature library. Based on the comparison results, it is determined whether there are any unsafe construction behaviors in the current image. If so, they are classified and marked, such as "not wearing a safety helmet", "illegal operation of the drilling rig", etc. The identification and classification results are output, and the alarm system can be used to remind on-site personnel or recorded in the database for subsequent analysis. The system realizes real-time image acquisition and analysis of the drill rig construction site through the preset sampling frequency of the image acquisition module. By using an advanced convolutional neural network learning model, it can accurately extract the target features of the drill rig construction site and effectively identify unsafe construction behaviors. The scale adaptive pooling unit can adaptively adjust the pooling window size according to the image scale, improving the system's processing ability for images in different scenarios. This system is applicable to various drill rig construction sites, such as building construction sites, mine exploitation sites, etc., providing strong support for ensuring construction safety.

[0024] In an alternative embodiment, the processing of the pre-image of the drill rig construction site based on a preset image verification algorithm to determine the initial image of the drill rig construction site may include: 1. Based on the preset image verification algorithm, perform threshold segmentation on the pre-image of the drill rig construction site, and perform distance transformation on the threshold segmentation result to obtain a distance transformation result; 2. Use the symmetric algorithm of hyperbolic operation to perform confusion calculation on the distance transformation result to obtain a confusion calculation result; 3. Generate the initial image of the drill rig construction site according to the confusion calculation result.

[0025] First, based on the preset image verification algorithm, perform threshold segmentation on the pre-image of the drill rig construction site. Threshold segmentation is an image segmentation method that sets one or more thresholds according to the gray value or color value of the image to divide the image into different regions. In this step, through threshold segmentation, the target regions (such as construction workers, construction equipment, etc.) in the image can be separated from the background region. Then, perform distance transformation on the result of threshold segmentation. Distance transformation is an image processing technique that calculates the distance from each pixel point in the image to the boundary of the nearest target region. Through distance transformation, a distance image can be obtained, where the value of each pixel represents the distance from that pixel point to the boundary of the nearest target region.

[0026] Then, use the symmetric algorithm of hyperbolic operation to perform confusion calculation on the distance transformation result. Hyperbolic operation is a mathematical operation that is usually used to process non-linear relationships. Here, the symmetric algorithm of hyperbolic operation is used to perform confusion processing on the distance transformation result to extract more meaningful feature information. The result of the confusion calculation will be used for subsequent marker generation.

[0027] Finally, generate the marker of the initial image according to the confusion calculation result. This step involves mapping the confusion calculation result back to the original image space and generating a marker image with the same size as the original image. In the marker image, the value of each pixel represents whether that pixel point belongs to the target region (such as construction workers, construction equipment, etc.), or represents a certain relationship (such as distance, direction, etc.) between that pixel point and the target region.

[0028] It should be noted that the selection of the symmetric algorithm for hyperbolic operations should be determined according to specific application scenarios and requirements. For example, the hyperbolic function can be: , where , 、 are constants. Different algorithms may extract different feature information, thus affecting the final marking result.

[0029] In an alternative embodiment, after converting and generating the initial image of the drilling rig construction site according to the confusion calculation result, the following steps may further be included: 1. Obtain an image partitioning matrix with the same size as the initial image; 2. Perform scale partitioning on the image partitioning matrix to obtain a number of first image sub-partitioning matrices; 3. Perform re-partitioning on the first image sub-partitioning matrices to obtain a number of second image sub-partitioning matrices; 4. Perform local encoding on the first image sub-partitioning matrices and global encoding on the second image sub-partitioning matrices to achieve multi-scale partitioning of the initial image.

[0030] Specifically, first create a matrix with the same size as the marked initial image. This matrix will be used as the basis for subsequent scale partitioning. Usually, each element of this matrix corresponds to a pixel or pixel block in the original image. Then, the entire image partitioning matrix is divided into several sub-matrices according to a certain rule (such as uniform partitioning, content-based partitioning, etc.). These sub-matrices are called first image sub-partitioning matrices, and each sub-matrix represents a part or feature region of the original image.

[0031] Furthermore, each first image sub-partitioning matrix is divided to obtain smaller sub-matrices, which are called second image sub-partitioning matrices. Finally, local encoding is performed on the first image sub-partitioning matrices and global encoding is performed on the second image sub-partitioning matrices. Among them, the local encoding method can capture the features inside each sub-matrix. For the second image sub-partitioning matrices, the global encoding method is used to capture the position and relationship of the sub-matrices in the entire image.

[0032] Multi-scale partitioning can analyze the image from multiple scales. By capturing the features of the image at different scales, the content of the image can be understood more comprehensively, and more meaningful feature information can be extracted.

[0033] In an alternative embodiment, a convolutional neural network learning model is used to perform feature training on the initial image to obtain the target features of the drill rig construction site; wherein, the convolutional neural network learning model includes two layers of multi-scale input preprocessing units, three layers of deep feature extraction units, one layer of scale adaptive pooling unit, and two layers of feature mapping and generalization units; the multi-scale input preprocessing unit is used to receive image inputs of different scales and perform preprocessing operations; the deep feature extraction unit is used to gradually extract deep features in the image through three layers of convolution and activation operations; the scale adaptive pooling unit is used to adaptively adjust the size of the pooling window according to the scale of the input image; the feature mapping and generalization unit is used to perform mapping adjustment of the feature dimension using a 2x2 convolutional layer, and integrate global features through a fully connected layer to generate the target features of the drill rig construction site.

[0034] Among them, the two layers of multi-scale input preprocessing units are used to receive image inputs from different scales and perform necessary preprocessing operations, such as image size adjustment, normalization, etc., to ensure the consistency and stability of the input image in subsequent processing. The preprocessing of multi-scale input can be achieved through methods such as image scaling, cropping, or padding.

[0035] The three layers of deep feature extraction units can gradually extract deep features in the image through three layers of convolution operations and activation functions. Each layer of convolution captures different levels of feature information in the image. The convolution kernel slides on the image, calculates the convolution result, and performs a non-linear transformation through the activation function to extract the features in the image.

[0036] Exemplarily, the gradual extraction of deep features in the image through three layers of convolution and activation operations is represented by the following formula: Among them, represents the first layer of convolution and activation operation, represents the activation function, represents the preprocessed input image, represents the first convolution kernel, represents the first bias term; Among them, represents the second layer of convolution and activation operation, represents the activation function, represents the second convolution kernel, represents the second bias term; Among them, represents the third layer of convolution and activation operation, represents the activation function, represents the third convolutional kernel, represents the third bias term.

[0037] A layer of scale adaptive pooling unit is used to adaptively adjust the size of the pooling window according to the scale of the input image, so as to extract scale-independent feature information. For example, methods such as max pooling or average pooling can be used, but the size of the pooling window will be dynamically adjusted according to the scale of the input image.

[0038] The two-layer feature mapping and generalization unit first uses a 2x2 convolutional layer to adjust the mapping of the feature dimension, and then integrates the global features through a fully connected layer to generate a pixelated general expression. This general expression can be used for subsequent tasks such as image classification, recognition, or segmentation.

[0039] Specifically, the 2x2 convolutional layer is used to reduce the dimension of the feature map, while the fully connected layer is used to integrate the global features and generate the final output.

[0040] Exemplarily, the use of the 2x2 convolutional layer to adjust the mapping of the feature dimension, integrating the global features through the fully connected layer to generate a pixelated general expression, is represented by the following formula: where, represents the target feature of the drill rig construction site, represents the weight matrix of the fully connected layer, represents the output after operating on the 2x2 convolutional layer, satisfying , where, represents the output after passing through the scale adaptive pooling unit, represents the convolutional kernel of the feature mapping and generalization unit, represents the fourth bias term, represents flattened into a one-dimensional vector, represents the fifth bias term.

[0041] It should be noted that the training and optimization process of the convolutional neural network model for drill rig construction site feature extraction is crucial for its performance. During the training process, a large amount of drill rig construction site image data can be used to train the model, and its performance can be optimized by adjusting the parameters and structure of the model. At the same time, a validation set and a test set can also be used to evaluate the performance and generalization ability of the model.

[0042] Compared with the prior art, the present application includes an image acquisition module, a construction feature learning module, and a safety control classification module connected by communication. Among them, the image acquisition module is used to obtain a pre-image of the drill rig construction site through an image acquisition device according to an image acquisition instruction at a preset sampling frequency, and perform image processing on the pre-image of the drill rig construction site based on a preset image verification algorithm to determine an initial image of the drill rig construction site; the construction feature learning module is used to perform feature training on the initial image by using a convolutional neural network learning model to obtain target features of the drill rig construction site; the safety control classification module is used to identify and classify images according to the target features of the construction site and identify unsafe construction behaviors in the construction area. It realizes the digitization of the safety supervision work at the drill rig construction site through an intelligent on-site safety control system, improves the efficiency and accuracy of safety management at the drill rig construction site, and reduces labor costs.

[0043] See Figure 3 , Figure 3 which is a schematic flow chart of a method for on-site safety control for drill rig construction provided by an embodiment of the present invention, including: S301: Obtain a pre-image of the drill rig construction site through an image acquisition device according to an image acquisition instruction at a preset sampling frequency; S302: Perform image processing on the pre-image of the drill rig construction site based on a preset image verification algorithm to determine an initial image of the drill rig construction site; S303: Perform feature training on the initial image by using a convolutional neural network learning model to obtain target features of the drill rig construction site; S304: Identify and classify the image according to the target feature to identify unsafe construction behaviors in the construction area.

[0044] Specifically, a pre-image of the drill rig construction site is obtained through an image acquisition device according to an image acquisition instruction at a preset sampling frequency. The image acquisition device can be a high-definition camera installed at different positions of the drill rig construction site, etc., and the preset sampling frequency can be flexibly set according to the actual construction situation and monitoring requirements to ensure that sufficient and effective image information can be obtained. Perform image processing on the pre-image of the drill rig construction site based on a preset image verification algorithm to determine an initial image of the drill rig construction site. The preset image verification algorithm is used to remove invalid information such as noise and interference in the pre-image, optimize and correct the image, improve the image quality, and provide a reliable basis for subsequent analysis and processing. The convolutional neural network learning model is used to train the features of the initial image to obtain the target features of the drilling rig construction site. The convolutional neural network learning model has a strong ability to extract image features. Through learning and training the initial image, it can accurately extract key features related to drilling rig construction safety, such as the posture of the workers and the operating status of the equipment. The image is identified and classified according to the target features to identify unsafe construction behaviors in the construction area. The extracted target features are compared and analyzed with the pre-set safe and unsafe behavior patterns to determine whether there are unsafe construction behaviors in the construction area, such as workers not wearing safety helmets, operating equipment in violation of regulations, etc.

[0045] The above method combines image processing and machine learning technology to achieve safety control of drilling rig construction sites. It can not only capture and analyze image data of the construction site in real time, but also accurately identify unsafe construction behaviors, providing strong support for safety management of the construction site.

[0046] Compared with the prior art, the present application first obtains a pre-image of the drilling rig construction site through an image acquisition device according to an image acquisition instruction and a preset sampling frequency; performs image processing on the pre-image of the drilling rig construction site based on a preset image verification algorithm to determine the initial image of the drilling rig construction site; performs feature training on the initial image using a convolutional neural network learning model to obtain target features of the drilling rig construction site; and recognizes and classifies the image according to the target features to identify unsafe construction behaviors in the construction area. It can not only capture and analyze image data of the construction site in real time, but also accurately identify unsafe construction behaviors, providing strong support for safety management of the construction site.

[0047] An embodiment of the present invention further provides a storage medium, in which a computer program is stored, wherein the computer program is configured to implement the steps in the above method embodiment when running.

[0048] Specifically, in this embodiment, the above storage medium may be configured to store a computer program for performing the following steps: S301: According to the image acquisition instruction and the preset sampling frequency, a preliminary image of the drilling rig construction site is obtained through an image acquisition device; S302: performing image processing on the preliminary image of the drilling rig construction site based on a preset image verification algorithm to determine an initial image of the drilling rig construction site; S303: Perform feature training on the initial image using a convolutional neural network learning model to obtain target features of the drilling rig construction site; S304: Identify and classify the image according to the target features to identify unsafe construction behaviors in the construction area.

[0049] Specifically, in this embodiment, the above storage medium may include but is not limited to: various media such as USB flash drives, read-only memories (ROM for short), random access memories (RAM for short), external hard drives, magnetic disks, or optical discs that can store computer programs.

[0050] Compared with the prior art, this application first obtains a pre-image of the drill rig construction site through an image acquisition device according to an image acquisition instruction at a preset sampling frequency; performs image processing on the pre-image of the drill rig construction site based on a preset image verification algorithm to determine an initial image of the drill rig construction site; uses a convolutional neural network learning model to perform feature training on the initial image to obtain target features of the drill rig construction site; and identifies and classifies the image according to the target features to identify unsafe construction behaviors in the construction area. It can not only capture and analyze image data on the construction site in real time, but also accurately identify unsafe construction behaviors, providing strong support for safety management on the construction site.

[0051] An embodiment of the present invention also provides an electronic device, including a memory and a processor, where a computer program is stored in the memory, and the processor is configured to run the computer program to execute the steps in the above method embodiment.

[0052] Specifically, the above electronic device may further include a transmission device and an input / output device, where the transmission device is connected to the above processor, and the input / output device is connected to the above processor.

[0053] Specifically, in this embodiment, the above processor may be configured to execute the following steps through a computer program: S301: Obtain a pre-image of the drill rig construction site through an image acquisition device according to an image acquisition instruction at a preset sampling frequency; S302: Perform image processing on the pre-image of the drill rig construction site based on a preset image verification algorithm to determine an initial image of the drill rig construction site; S303: Use a convolutional neural network learning model to perform feature training on the initial image to obtain target features of the drill rig construction site; S304: Identify and classify the image according to the target features to identify unsafe construction behaviors in the construction area.

[0054] Compared with the prior art, the present application first obtains a pre-image of the drill rig construction site through an image acquisition device according to an image acquisition instruction at a preset sampling frequency; performs image processing on the pre-image of the drill rig construction site based on a preset image verification algorithm to determine an initial image of the drill rig construction site; uses a convolutional neural network learning model to perform feature training on the initial image to obtain target features of the drill rig construction site; and identifies and classifies the image according to the target features to identify unsafe construction behaviors in the construction area. It can not only capture and analyze image data of the construction site in real time, but also accurately identify unsafe construction behaviors, providing strong support for safety management of the construction site.

[0055] It should be noted that for the foregoing method embodiments, for the sake of simple description, they are all expressed as a series of action combinations. However, those skilled in the art should know that the present invention is not limited by the described action sequence, because according to the present invention, certain steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily essential to the present invention.

[0056] In the above embodiments, the descriptions of the respective embodiments have their own emphases. For parts not detailed in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0057] In several embodiments provided by the present invention, it should be understood that the disclosed device can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the above division of units is only a logical function division, and there can be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed coupling or direct coupling or communication connection to each other can be through some interfaces, and the indirect coupling or communication connection of the device or unit can be in an electrical or other form.

[0058] The units described as separate components above may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they can be located in one place, or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0059] In addition, the functional units in various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above integrated unit can be implemented in the form of hardware or in the form of a software functional unit.

[0060] If the above integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable memory. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a memory and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the above methods in various embodiments of the present invention. The aforementioned memory includes various media that can store program codes, such as USB flash drives, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), mobile hard disks, magnetic disks, or optical discs.

[0061] The above has introduced the embodiments of the present invention in detail. Specific examples are used in this article to elaborate on the principle and implementation manner of the present invention. The description of the above embodiments is only used to help understand the method and its core idea of the present invention; at the same time, for those of ordinary skill in the art, according to the idea of the present invention, there will be changes in the specific implementation manner and application scope. In summary, the content of this specification should not be construed as a limitation to the present invention.

Claims

1. A on-site safety control system for drilling rig construction, the system comprising: An image acquisition module, a construction feature learning module, and a safety control classification module that are communicatively connected, wherein, The image acquisition module is configured to obtain a pre-image of the drilling rig construction site through an image acquisition device according to an image acquisition instruction at a preset sampling frequency, and perform image processing on the pre-image of the drilling rig construction site based on a preset image verification algorithm to determine an initial image of the drilling rig construction site; The construction feature learning module is configured to perform feature training on the initial image using a convolutional neural network learning model to obtain target features of the drilling rig construction site; wherein, the convolutional neural network learning model includes two layers of multi-scale input preprocessing units, three layers of deep feature extraction units, one layer of scale adaptive pooling units, and two layers of feature mapping and generalization units; The multi-scale input preprocessing unit is configured to receive image inputs of different scales and perform preprocessing operations; The deep feature extraction unit is configured to gradually extract deep features in the image through three layers of convolution and activation operations; The scale adaptive pooling unit is configured to adaptively adjust the size of the pooling window according to the scale of the input image; The feature mapping and generalization unit is configured to perform mapping adjustment of the feature dimension using a 2x2 convolutional layer, and integrate global features through a fully connected layer to generate target features of the drilling rig construction site; The safety control classification module is configured to identify and classify the image based on the target features of the construction site, and identify unsafe construction behaviors in the construction area.

2. The system according to claim 1, wherein The performing image processing on the pre-image of the drilling rig construction site based on a preset image verification algorithm to determine an initial image of the drilling rig construction site includes: Performing threshold segmentation on the pre-image of the drilling rig construction site based on a preset image verification algorithm, performing distance transformation on the threshold segmentation result to obtain a distance transformation result; Performing confusion calculation on the distance transformation result using a symmetric algorithm of hyperbolic operation to obtain a confusion calculation result; Converting and generating an initial image of the drilling rig construction site according to the confusion calculation result.

3. The system according to claim 2, wherein After converting and generating an initial image of the drilling rig construction site according to the confusion calculation result, it further includes: Obtaining an image division matrix having the same size as the initial image; Performing scale division on the image division matrix to obtain a plurality of first image division sub-matrices; Performing re-division on the first image division sub-matrices to obtain a plurality of second image division sub-matrices; Performing local coding on the first image division sub-matrices and performing global coding on the second image division sub-matrices to implement multi-scale division of the initial image.

4. The system according to claim 3, wherein The gradually extracting deep features in the image through three layers of convolution and activation operations is represented by the following formula: Among them, represents the first layer of convolution and activation operations, represents the activation function, represents the preprocessed input image, represents the first convolution kernel, represents the first bias term; Among them, represents the second layer of convolution and activation operations, represents the activation function, represents the second convolutional kernel, represents the second bias term; Among them, represents the third-layer convolution and activation operation, represents the activation function, represents the third convolution kernel, represents the third bias term.

5. The system according to claim 4, wherein The performing mapping adjustment of the feature dimension using a 2x2 convolutional layer and integrating global features through a fully connected layer to generate target features of the drilling rig construction site is represented by the following formula: Among them, represents the target feature at the drill rig construction site, represents the weight matrix of the fully connected layer, represents the output after operating with a 2x2 convolutional layer, satisfying , where, represents the output after passing through the scale adaptive pooling unit, represents the convolution kernel of the feature mapping and generalization unit, represents the fourth bias term, represents that is flattened into a one-dimensional vector, represents the fifth bias term.

6. A method for on-site safety control of a drilling rig construction that executes the system according to any one of claims 1-5, characterized in that, The method includes: Obtaining a pre-image of the drilling rig construction site through an image acquisition device according to an image acquisition instruction at a preset sampling frequency; Based on a preset image verification algorithm, perform image processing on the pre-image of the drilling rig construction site to determine the initial image of the drilling rig construction site; Use a convolutional neural network learning model to perform feature training on the initial image to obtain the target features of the drilling rig construction site; Identify and classify the image according to the target features to identify unsafe construction behaviors in the construction area.

7. A storage medium, characterized in that, A computer program is stored in the storage medium, wherein the computer program is set to implement the method described in claim 6 when running.

8. An electronic device, comprising a memory and a processor, characterized in that, A computer program is stored in the memory, and the processor is set to run the computer program to implement the method described in claim 6.

Images