Automatic online detection system for steel plate drilling based on machine vision

Abstract

The invention discloses a steel plate drilling automatic online detection system based on machine vision, and relates to the technical field of steel plate drilling machining systems. The system comprises a control system module, the input end of the control system module is electrically connected with an image processing module, the input end of the image processing module is electrically connected with an image acquisition module, and the output end and the input end of the control system module are electrically connected with a deep learning module. The output end and the input end of the control system module are electrically connected with a data processing module. According to the invention, the deep learning module uses a deep learning algorithm to train a large number of drilling images, automatically learns the characteristics and modes of drilling, improves the accuracy and robustness of detection, can continuously optimize the performance of itself, adapts to different types of steel plates and drilling requirements, and improves the accuracy and robustness of detection. And the intelligent diagnosis module is combined with an expert system and data analysis to perform intelligent diagnosis on a detection result.

Description

Technical Field

[0001] The invention relates to the technical field of steel plate drilling processing systems, and in particular to an automatic online detection system for steel plate drilling based on machine vision. Background Art

[0002] The automatic online detection system for steel plate drilling based on machine vision is an advanced detection technology, which is mainly used for real-time and accurate detection of drill holes on steel plates, thereby ensuring the precise processing of steel plate drilling holes.

[0003] However, the existing automatic online detection system for steel plate drilling has some defects in use. For example, it requires manual intervention, which greatly reduces the detection efficiency and production speed and increases labor costs. It also does not have the function of combining deep learning and intelligent diagnosis, and cannot perform intelligent analysis of the detection results, thereby failing to provide detailed defect analysis and improvement suggestions, and thus cannot meet the needs of today's market.

[0004] To this end, we provide an automatic online detection system for steel plate drilling based on machine vision to solve the above problems. Summary of the invention

[0005] The purpose of the present invention is to provide an automatic online detection system for steel plate drilling based on machine vision. Through the cooperation of a deep learning module, an intelligent diagnosis module and an adaptive adjustment module, the existing automatic online detection system for steel plate drilling in the prior art solves some defects in use, such as the need for manual intervention, which greatly reduces the detection efficiency and production speed, increases labor costs, and does not have the function of combining deep learning and intelligent diagnosis, and cannot perform intelligent analysis of the detection results, thereby failing to provide detailed defect analysis and improvement suggestions, and thus failing to meet the current market demand.

[0006] In order to solve the above technical problems, the present invention is achieved through the following technical solutions:

[0007] The present invention is an automatic online detection system for steel plate drilling based on machine vision, comprising a control system module, the input end of the control system module is electrically connected to an image processing module, the input end of the image processing module is electrically connected to an image acquisition module, the output end and input end of the control system module are electrically connected to a deep learning module, the output end and input end of the control system module are electrically connected to a data processing module, and the output end and input end of the control system module are electrically connected to an intelligent diagnosis module;

[0008] The output end and the input end of the control system module are both electrically connected to the adaptive adjustment module, and the output end and the input end of the control system module are both electrically connected to the human-machine interaction module.

[0009] By adopting the above technical solution, a large number of drilling images are trained through the deep learning module using the deep learning algorithm, and the characteristics and patterns of drilling are automatically learned to improve the accuracy and robustness of detection. The module can continuously optimize its own performance and adapt to different types of steel plates and drilling requirements. The intelligent diagnosis module combines the expert system and data analysis to perform intelligent diagnosis on the detection results, and provide detailed defect analysis and improvement suggestions. This module can help operators quickly locate problems and take corresponding measures to improve them. The adaptive adjustment module automatically adjusts the parameters and algorithms of the detection system according to changes in the material, thickness and drilling requirements of the steel plate to ensure the accuracy and stability of the detection.

[0010] The present invention is further configured such that the output end and the input end of the control system module are both electrically connected to a wireless transceiver module, and the output end and the input end of the wireless transceiver module are both electrically connected to a remote monitoring module.

[0011] By adopting the above technical solution, the wireless transceiver module can receive and send wireless signals, thereby facilitating remote control by users, and remote monitoring and management detection systems can be realized through the remote monitoring module, which facilitates enterprises to uniformly manage multiple production sites. This module can improve management efficiency and reduce maintenance costs.

[0012] The present invention is further configured such that the output end and input end of the control system module are electrically connected to a data reading and writing module, the output end and input end of the data reading and writing module are electrically connected to a data encryption module, and the output end and input end of the data encryption module are electrically connected to a data storage module.

[0013] By adopting the above technical solution, data can be read and written through the data reading and writing module, data can be encrypted and stored through the data encryption module to improve data security, and data can be stored through the data storage module.

[0014] The present invention is further configured such that the image acquisition module includes a camera selection submodule, a lens configuration submodule, a light source control submodule and a trigger sensing submodule, and the output ends of the camera selection submodule, the lens configuration submodule, the light source control submodule and the trigger sensing submodule are all electrically connected to the input end of the image processing module.

[0015] By adopting the above technical solution and accurately selecting the camera selection submodule, it can be ensured that the camera can adapt to different steel plate materials, drilling hole sizes and lighting conditions of the production environment, and provide high-quality original image data for subsequent image processing. The lens configuration submodule selects the optimal focal length, aperture and other parameters according to the specific requirements of the detection task, so that the drill hole presents a clear and accurate shape in the image, which is convenient for subsequent feature extraction and analysis. The light source control submodule adjusts the intensity, angle and type of the light source to ensure that the drill hole has sufficient contrast and clarity in the image, avoiding the influence of uneven lighting or too dark or too bright lighting on the detection results. The trigger sensing submodule monitors the position and status of the steel plate in real time through the sensor. When the steel plate enters the detection area, the camera is triggered to shoot in time to avoid missing the detection opportunity.

[0016] The present invention is further configured such that the image processing module includes an image preprocessing submodule, a feature extraction submodule and an image analysis submodule, the output end of the image acquisition module is electrically connected to the input ends of the image preprocessing submodule, the feature extraction submodule and the image analysis submodule respectively, and the output ends of the image preprocessing submodule, the feature extraction submodule and the image analysis submodule are all electrically connected to the input end of the control system module.

[0017] By adopting the above technical solution, the image preprocessing submodule removes noise interference in the image through operations such as denoising, enhancement, and binarization, improves the contrast and clarity of the image, and provides better quality image data for subsequent processing steps. The feature extraction submodule uses edge detection, contour extraction and other algorithms to extract the characteristic parameters of the drilling hole, such as aperture, position, shape, etc., from the preprocessed image to provide a basis for subsequent analysis and judgment. The image analysis submodule analyzes and calculates the characteristic parameters of the drilling hole according to preset standards and algorithms, determines whether the drilling hole is qualified, and identifies potential defects.

[0018] The present invention is further configured such that the deep learning module includes a data preparation submodule, a model training submodule, a model evaluation submodule and a model application submodule, and the data preparation submodule, the model training submodule, the model evaluation submodule and the model application submodule are all bidirectionally electrically connected to the control system module.

[0019] By adopting the above technical solution, the data preparation submodule is responsible for collecting and organizing a large amount of drilling image data, and annotating and preprocessing it to provide rich and accurate training samples for model training. The model training submodule uses the prepared data and adopts appropriate deep learning algorithms and frameworks to train the model, continuously optimize the model parameters, and improve the performance and accuracy of the model. The model evaluation submodule evaluates the trained model by using a validation set or a test set, calculates the model's accuracy, recall rate, F1 value and other indicators, and determines whether the model's performance meets the requirements. The model application submodule deploys the trained model to the detection system, detects and analyzes the real-time collected images, and quickly and accurately determines the quality and defects of the drilling.

[0020] The present invention is further configured such that the data processing module includes a data analysis submodule and a data transmission submodule, and both the data analysis submodule and the data transmission submodule are bidirectionally electrically connected to the control system module.

[0021] By adopting the above technical solution, the data analysis submodule performs statistical analysis on the data, generates reports and charts, and provides a decision-making basis for production management and quality control. The data transmission submodule is responsible for transmitting the data in the data processing module to other modules, such as the control system module and the remote monitoring module, to realize data sharing and collaborative processing.

[0022] The present invention is further configured such that the intelligent diagnosis module includes a defect analysis submodule, a cause inference submodule, an improvement suggestion submodule and a knowledge management submodule, and the defect analysis submodule, cause inference submodule, improvement suggestion submodule and knowledge management submodule are all bidirectionally electrically connected to the control system module.

[0023] By adopting the above technical solution, the defect analysis submodule performs a detailed defect analysis on the detected unqualified drilling holes, determines the defect type, location and severity, and provides a basis for subsequent cause inference and improvement suggestions. The cause inference submodule infers the possible causes of defects, such as tool wear, improper processing parameters, etc., based on the results of defect analysis, combined with expert systems and data analysis. The improvement suggestion submodule provides specific improvement suggestions based on the results of cause inference, such as replacing tools, adjusting processing parameters, optimizing process flow, etc., to help production personnel solve problems in a timely manner and improve product quality. The knowledge management submodule establishes and manages a defect knowledge base, organizes and stores common defect types, causes and solutions, which is convenient for operators to query and refer to, and improves the efficiency and accuracy of diagnosis.

[0024] The present invention is further configured such that the adaptive adjustment module includes a material identification submodule, a parameter adjustment submodule, an algorithm optimization submodule and a feedback control submodule, and the material identification submodule, the parameter adjustment submodule, the algorithm optimization submodule and the feedback control submodule are all bidirectionally electrically connected to the control system module.

[0025] By adopting the above technical solution, the material identification submodule automatically identifies the material and characteristics of the steel plate, providing a basis for subsequent parameter adjustment and algorithm optimization. The parameter adjustment submodule automatically adjusts the parameters of the detection system, such as camera parameters, light source intensity, image processing algorithm parameters, etc., according to factors such as the material and thickness of the steel plate and the feedback of the detection results, to ensure the accuracy and stability of the detection. The algorithm optimization submodule automatically optimizes the detection algorithm according to different drilling requirements and production environments to improve the accuracy and efficiency of the detection. The feedback control submodule adjusts the system parameters and algorithms in real time according to the feedback of the detection results, so that the detection system can continuously adapt to changes in the production process and improve the robustness and adaptability of the system.

[0026] The present invention has the following beneficial effects:

[0027] High-precision detection: Using high-resolution cameras and advanced image processing algorithms, it is possible to accurately detect parameters such as the size, position and shape of the drilled holes. The detection accuracy can reach the micron level, and it can detect tiny defects such as hole diameter deviation and hole wall roughness, ensuring that product quality meets high standards.

[0028] High degree of automation: It realizes online automatic detection without manual intervention, greatly improves detection efficiency and production speed, reduces labor costs, can be seamlessly integrated with the production line, and provides real-time feedback on detection results, which facilitates timely adjustment of production processes and improves the level of intelligent production;

[0029] Strong reliability: It uses stable and reliable hardware equipment and software systems, has strong anti-interference ability, can operate stably for a long time in complex industrial environments, and has undergone rigorous testing and verification. The detection results are accurate and reliable, reducing false detection and missed detection caused by human factors;

[0030] Good flexibility: It can be flexibly adjusted according to different steel plate materials, thicknesses and drilling requirements to adapt to a variety of production scenarios. By adding new algorithms and modules, the system's functions can be continuously expanded to meet the needs of future production development;

[0031] Convenient data management: The test data can be stored and managed to facilitate subsequent statistical analysis and quality traceability, providing strong data support for production management. Detailed reports and charts can be generated to intuitively display the quality status of the production process, helping enterprises to identify problems in a timely manner and take improvement measures;

[0032] Intelligent diagnosis and optimization: Combining deep learning and intelligent diagnosis modules, it can intelligently analyze the test results, provide detailed defect analysis and improvement suggestions, and help enterprises improve the production process level. The adaptive adjustment module can automatically adjust the parameters and algorithms of the detection system according to changes in the production process to ensure the accuracy and stability of the detection;

[0033] Remote monitoring and management: Through the remote monitoring module, enterprise managers can view the operating status and test results of the detection system anytime and anywhere, realize remote monitoring and management of the production process, facilitate enterprises to uniformly manage multiple production sites, improve management efficiency and reduce management costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for describing the embodiments are briefly introduced below.

[0035] Figure 1 This is the system schematic diagram of the automatic online detection system for steel plate drilling based on machine vision;

[0036] Figure 2 This is the system diagram of the image acquisition module in the automatic online detection system for steel plate drilling based on machine vision;

[0037] Figure 3 This is the system diagram of the image processing module in the automatic online detection system for steel plate drilling based on machine vision;

[0038] Figure 4 This is the system diagram of the deep learning module in the automatic online detection system for steel plate drilling based on machine vision;

[0039] Figure 5 This is the system diagram of the data processing module in the automatic online detection system for steel plate drilling based on machine vision.

[0040] Figure 6 This is the system diagram of the intelligent diagnosis module in the automatic online detection system for steel plate drilling based on machine vision.

[0041] Figure 7 This is the system diagram of the adaptive adjustment module in the automatic online detection system for steel plate drilling based on machine vision. DETAILED DESCRIPTION

[0042] The technical solutions in the embodiments of the present invention will be described below in conjunction with the drawings in the embodiments of the present invention. The described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. Specific embodiment 1

[0044] See also Figure 1-Figure 7The present invention is an automatic online detection system for steel plate drilling based on machine vision, including a control system module, an input end of the control system module is electrically connected to an image processing module, an input end of the image processing module is electrically connected to an image acquisition module, an output end and an input end of the control system module are electrically connected to a deep learning module, an output end and an input end of the control system module are electrically connected to a data processing module, and an output end and an input end of the control system module are electrically connected to an intelligent diagnosis module;

[0045] The output end and the input end of the control system module are both electrically connected to the adaptive adjustment module, and the output end and the input end of the control system module are both electrically connected to the human-machine interaction module.

[0046] Specifically: Through the deep learning module, a large number of drilling images are trained using deep learning algorithms to automatically learn the characteristics and patterns of drilling, thereby improving the accuracy and robustness of detection. The module can continuously optimize its own performance to adapt to different types of steel plates and drilling requirements. Through the intelligent diagnosis module, combined with the expert system and data analysis, the detection results are intelligently diagnosed, and detailed defect analysis and improvement suggestions are provided. This module can help operators quickly locate problems and take corresponding measures to improve them. Through the adaptive adjustment module, the parameters and algorithms of the detection system are automatically adjusted according to changes in the material, thickness and drilling requirements of the steel plate to ensure the accuracy and stability of detection. Specific embodiment 2

[0048] See also Figure 1-Figure 7On the basis of the specific embodiment 1, the output end and input end of the control system module are electrically connected to the wireless transceiver module, the output end and input end of the wireless transceiver module are electrically connected to the remote monitoring module, the output end and input end of the control system module are electrically connected to the data reading and writing module, the output end and input end of the data reading and writing module are electrically connected to the data encryption module, the output end and input end of the data encryption module are electrically connected to the data storage module, the image acquisition module includes a camera selection submodule, a lens configuration submodule, a light source control submodule and a trigger sensing submodule, the output ends of the camera selection submodule, the lens configuration submodule, the light source control submodule and the trigger sensing submodule are electrically connected to the input end of the image processing module, the image processing module includes an image preprocessing submodule, a feature extraction submodule and an image analysis submodule, the output end of the image acquisition module is electrically connected to the input ends of the image preprocessing submodule, the feature extraction submodule and the image analysis submodule respectively, the image preprocessing submodule, the feature extraction submodule and the image The output ends of the analysis submodules are all electrically connected to the input ends of the control system module. The deep learning module includes a data preparation submodule, a model training submodule, a model evaluation submodule and a model application submodule. The data preparation submodule, the model training submodule, the model evaluation submodule and the model application submodule are all electrically connected to the control system module in a bidirectional manner. The data processing module includes a data analysis submodule and a data transmission submodule. The data analysis submodule and the data transmission submodule are all electrically connected to the control system module in a bidirectional manner. The intelligent diagnosis module includes a defect analysis submodule, a cause inference submodule, an improvement suggestion submodule and a knowledge management submodule. The defect analysis submodule, the cause inference submodule, the improvement suggestion submodule and the knowledge management submodule are all electrically connected to the control system module in a bidirectional manner. The adaptive adjustment module includes a material recognition submodule, a parameter adjustment submodule, an algorithm optimization submodule and a feedback control submodule. The material recognition submodule, the parameter adjustment submodule, the algorithm optimization submodule and the feedback control submodule are all electrically connected to the control system module in a bidirectional manner.

[0049] Specifically: The wireless transceiver module can receive and send wireless signals, which is convenient for users to perform remote control. The remote monitoring module can realize remote monitoring and management of the detection system, which is convenient for enterprises to uniformly manage multiple production sites. This module can improve management efficiency and reduce maintenance costs. The data reading and writing module can read and write data. The data encryption module can encrypt and store data to improve data security. The data storage module can store data. Through the precise selection of the camera selection submodule, it can ensure that the camera can adapt to different steel plate materials, drilling hole sizes and lighting conditions of the production environment, providing high-quality original image data for subsequent image processing. The lens configuration submodule can be used according to the specific conditions of the detection task. The optimal focal length, aperture and other parameters are selected according to the requirements of the body, so that the drilled holes present a clear and accurate shape in the image, which is convenient for subsequent feature extraction and analysis. The light source control submodule ensures that the drilled holes have sufficient contrast and clarity in the image by adjusting the intensity, angle and type of the light source to avoid affecting the detection results due to uneven lighting or too dark or too bright. The trigger sensing submodule monitors the position and status of the steel plate in real time through the sensor. When the steel plate enters the detection area, the camera is triggered to shoot in time to avoid missing the detection opportunity. The image preprocessing submodule removes noise interference in the image through denoising, enhancement, binarization and other operations, improves the contrast and clarity of the image, and provides better quality image data for subsequent processing steps. The feature extraction submodule uses edge detection , contour extraction and other algorithms are used to extract the characteristic parameters of the drilling hole, such as aperture, position, shape, etc., from the preprocessed image to provide a basis for subsequent analysis and judgment. The image analysis submodule analyzes and calculates the characteristic parameters of the drilling hole according to the preset standards and algorithms, determines whether the drilling hole is qualified, and identifies potential defects. The data preparation submodule is responsible for collecting and organizing a large amount of drilling image data, and annotating and preprocessing it to provide rich and accurate training samples for model training. The model training submodule uses the prepared data, adopts appropriate deep learning algorithms and frameworks, trains the model, continuously optimizes the model parameters, and improves the performance and accuracy of the model. The model evaluation submodule evaluates the trained model by using the validation set or test set. , calculate the model's accuracy, recall rate, F1 value and other indicators to determine whether the model's performance meets the requirements. The model application submodule deploys the trained model to the detection system, detects and analyzes the real-time collected images, and quickly and accurately determines the quality and defects of the drilling. The data analysis submodule performs statistical analysis on the data, generates reports and charts, and provides a decision-making basis for production management and quality control. The data transmission submodule is responsible for transmitting the data in the data processing module to other modules, such as the control system module and the remote monitoring module, to achieve data sharing and collaborative processing. The defect analysis submodule performs a detailed defect analysis on the detected unqualified drilling holes, determines the defect type, location and severity, and provides a basis for subsequent cause inference and improvement suggestions.The cause inference submodule infers the possible causes of defects, such as tool wear and improper processing parameters, based on the results of defect analysis, combined with expert systems and data analysis. The improvement suggestion submodule provides specific improvement suggestions based on the results of cause inference, such as replacing tools, adjusting processing parameters, optimizing process flow, etc., to help production personnel solve problems in a timely manner and improve product quality. The knowledge control submodule establishes and manages a defect knowledge base, organizes and stores common defect types, causes and solutions, which is convenient for operators to query and refer to, and improves the efficiency and accuracy of diagnosis. The material identification submodule automatically identifies the material and characteristics of the steel plate, providing a basis for subsequent parameter adjustment and algorithm optimization. The parameter adjustment submodule automatically adjusts the parameters of the detection system, such as camera parameters, light source intensity, image processing algorithm parameters, etc., according to factors such as the material and thickness of the steel plate and the feedback of the detection results, to ensure the accuracy and stability of the detection. The algorithm optimization submodule automatically optimizes the detection algorithm for different drilling requirements and production environments to improve the accuracy and efficiency of the detection. The feedback control submodule adjusts the system parameters and algorithms in real time based on the feedback of the detection results, so that the detection system can continuously adapt to changes in the production process and improve the robustness and adaptability of the system.

[0050] The working principle of the present invention is as follows: when the steel plate enters the detection area, the sensor triggers the image acquisition module to start working, and the industrial camera shoots the drilled holes on the steel plate at a suitable position and angle. At the same time, the camera parameters (such as exposure time, focal length, etc.) and the light source intensity are adjusted according to factors such as the material and thickness of the steel plate to ensure that a clear and high-quality image is obtained. The captured image is transmitted to the image processing module for subsequent processing. The image processing module pre-processes the collected image, including denoising, enhancement, binarization and other operations to improve the image quality and contrast. Edge detection, contour extraction, feature point matching and other algorithms are used to extract the characteristic parameters of the drilled hole, such as aperture, position, shape, etc. , the processed images and feature parameters are transmitted to the data processing module and other related modules for further analysis. The images and feature parameters output by the image processing module are analyzed by the deep learning module using the pre-trained deep learning model. The model automatically identifies the characteristics and patterns of the drilling holes, determines whether the drilling holes are qualified, and further identifies potential complex defects. The detection results and more detailed defect information are transmitted to the intelligent diagnosis module and other modules. The intelligent diagnosis module receives the detection results and defect information provided by the deep learning module and the data processing module, analyzes the defects in detail, determines the defect type, location and severity, and infers the possible causes by combining the expert system and data analysis. The cause of the defect is determined by the cause analysis, and specific improvement suggestions are provided based on the cause analysis, such as replacing tools, adjusting processing parameters, optimizing process flow, etc. The diagnosis results and improvement suggestions are transmitted to the human-computer interaction module and the remote monitoring module so that operators and managers can take corresponding measures. The information of the image acquisition module, image processing module, deep learning module, data processing module, intelligent diagnosis module and control system module is received in real time through the network connection. Authorized users can remotely access the detection system to view real-time detection results, equipment status and historical data. When the detection system has an abnormal situation or detects unqualified products, an alarm notification is sent to relevant personnel in a timely manner, and other production sites can be Experience and knowledge are fed back to the intelligent diagnosis module to improve the performance of the entire system. The adaptive adjustment module automatically adjusts the parameters and algorithms of the detection system according to changes in the material, thickness and drilling requirements of the steel plate, as well as information provided by the deep learning module and the intelligent diagnosis module. The adjustment process and results are fed back to other modules so that the entire system can better adapt to changes in the production process. The human-computer interaction module displays the detection results, images and information provided by the intelligent diagnosis module, which is convenient for operators to monitor the detection process in real time, and to perform parameter setting, equipment control, data query and other operations. When an alarm occurs, the operator is reminded to handle it through sound and light.

[0051] The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to the specific implementation methods described. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present invention, so that technicians in the relevant technical field can better understand and utilize the present invention.

Claims

1. The automatic online detection system for steel plate drilling based on machine vision includes a control system module, which is characterized by: The input end of the control system module is electrically connected to the image processing module, the input end of the image processing module is electrically connected to the image acquisition module, the output end and input end of the control system module are electrically connected to the deep learning module, the output end and input end of the control system module are electrically connected to the data processing module, and the output end and input end of the control system module are electrically connected to the intelligent diagnosis module; The output end and the input end of the control system module are both electrically connected to the adaptive adjustment module, and the output end and the input end of the control system module are both electrically connected to the human-machine interaction module.

2. The automatic online detection system for steel plate drilling based on machine vision according to claim 1 is characterized in that: The output end and the input end of the control system module are both electrically connected to the wireless transceiver module, and the output end and the input end of the wireless transceiver module are both electrically connected to the remote monitoring module.

3. The automatic online detection system for steel plate drilling based on machine vision according to claim 1 is characterized in that: The output end and input end of the control system module are electrically connected to a data reading and writing module, the output end and input end of the data reading and writing module are electrically connected to a data encryption module, and the output end and input end of the data encryption module are electrically connected to a data storage module.

4. The automatic online detection system for steel plate drilling based on machine vision according to claim 1 is characterized in that: The image acquisition module includes a camera selection submodule, a lens configuration submodule, a light source control submodule and a trigger sensing submodule, and the output ends of the camera selection submodule, the lens configuration submodule, the light source control submodule and the trigger sensing submodule are all electrically connected to the input end of the image processing module.

5. The automatic online detection system for steel plate drilling based on machine vision according to claim 1 is characterized in that: The image processing module includes an image preprocessing submodule, a feature extraction submodule and an image analysis submodule. The output end of the image acquisition module is electrically connected to the input ends of the image preprocessing submodule, the feature extraction submodule and the image analysis submodule respectively. The output ends of the image preprocessing submodule, the feature extraction submodule and the image analysis submodule are all electrically connected to the input end of the control system module.

6. The automatic online detection system for steel plate drilling based on machine vision according to claim 1 is characterized in that: The deep learning module includes a data preparation submodule, a model training submodule, a model evaluation submodule and a model application submodule, and the data preparation submodule, the model training submodule, the model evaluation submodule and the model application submodule are all bidirectionally electrically connected to the control system module.

7. The automatic online detection system for steel plate drilling based on machine vision according to claim 1 is characterized in that: The data processing module includes a data analysis submodule and a data transmission submodule, and both the data analysis submodule and the data transmission submodule are bidirectionally electrically connected to the control system module.

8. The automatic online detection system for steel plate drilling based on machine vision according to claim 1 is characterized in that: The intelligent diagnosis module includes a defect analysis submodule, a cause inference submodule, an improvement suggestion submodule and a knowledge management submodule, and the defect analysis submodule, the cause inference submodule, the improvement suggestion submodule and the knowledge management submodule are all bidirectionally electrically connected to the control system module.

9. The automatic online detection system for steel plate drilling based on machine vision according to claim 1 is characterized in that: The adaptive adjustment module includes a material identification submodule, a parameter adjustment submodule, an algorithm optimization submodule and a feedback control submodule, and the material identification submodule, the parameter adjustment submodule, the algorithm optimization submodule and the feedback control submodule are all bidirectionally electrically connected to the control system module.

Images