Method and system for evaluating magnesium oxide production based on machine vision

Abstract

The application discloses a magnesium oxide production evaluation method and system based on machine vision, relates to the technical field of magnesium oxide production evaluation, collects silicon steel images, analyzes coating fluctuation conditions and coating leakage conditions, analyzes magnesium oxide coating coating conditions, collects factory environment data, analyzes adverse environment conditions, analyzes initial damage risks based on the magnesium oxide coating coating conditions and the adverse environment conditions, simulates a factory environment erosion evolution process, predicts pore distribution conditions, analyzes comprehensive oxidation resistance conditions based on oxygen entry conditions and oxidation resistance conditions, and judges whether the oxidation resistance of the magnesium oxide coating coated on the silicon steel meets standards based on the comprehensive oxidation resistance conditions. The application takes the initial coating conditions of the silicon steel and the adverse environment conditions as input variables, predicts the distribution conditions of pores by simulating the erosion evolution process of the factory environment, and thus early warns of potential failure risks of the silicon steel, thereby improving the accuracy and timeliness of magnesium oxide production evaluation.

Description

Technical Field

[0001] This invention relates to the field of magnesium oxide production assessment technology, and in particular to a magnesium oxide production assessment method and system based on machine vision. Background Technology

[0002] Silicon steel, especially grain-oriented silicon steel, is a key soft magnetic material for manufacturing core components such as transformers and large motors. In its production process, in order to form an excellent glass film underlayer during the high-temperature annealing process and ultimately obtain good insulation performance and oxidation resistance, a magnesium oxide coating is usually applied to the surface of the silicon steel strip. The quality of the magnesium oxide coating, such as the uniformity and density of the coating and the final oxidation resistance, directly determines the final quality and service life of the silicon steel product. Therefore, accurate and efficient detection and evaluation of the magnesium oxide coating quality and its oxidation resistance is a crucial technical link in the silicon steel production field.

[0003] Existing technologies for quality control of magnesium oxide coatings typically suffer from limitations such as isolated and one-sided testing methods, delayed and static risk assessment, and insufficient evaluation of oxidation resistance. Traditional quality control largely relies on offline sampling inspection of coating surface quality and independent monitoring of environmental parameters. These data lack effective correlation analysis, making it impossible to systematically evaluate the comprehensive performance of the coating in a real and dynamic factory environment. Furthermore, the magnesium oxide coating quality assessment process is static and cannot simulate the performance evolution of the coating over time under specific environmental conditions. Moreover, the assessment of the coating's oxidation resistance fails to delve into the microscopic mechanisms of oxygen intrusion and its dynamic coupling relationship with environmental factors, resulting in inaccurate assessment results.

[0004] To address the aforementioned problems, this invention provides a machine vision-based method and system for evaluating magnesium oxide production. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a machine vision-based method and system for evaluating magnesium oxide production. This invention uses the initial coating condition of silicon steel and harsh environmental conditions as input variables. By simulating the erosion evolution process of the factory environment, it predicts the distribution of pores, thereby providing early warning of potential failure risks of silicon steel and improving the accuracy and timeliness of magnesium oxide production evaluation.

[0006] To achieve the above objectives, the present invention provides a machine vision-based method for evaluating magnesium oxide production, comprising the following specific steps:

[0007] Step 1: Acquire images of silicon steel, analyze coating fluctuations and coating omissions, and analyze the magnesium oxide coating application based on coating fluctuations and coating omissions.

[0008] Step 2: Collect factory environmental data and analyze the harsh environmental conditions;

[0009] Step 3: Analyze the initial damage risk based on the magnesium oxide coating application and harsh environmental conditions, simulate the erosion evolution process in the factory environment, and predict the porosity distribution.

[0010] Step 4: Analyze oxygen entry based on stomatal distribution and factory environment oxygen concentration; analyze antioxidant capacity based on oxidation weight gain per unit area; and analyze comprehensive antioxidant capacity based on oxygen entry and antioxidant capacity.

[0011] Step 5: Determine whether the oxidation resistance of the magnesium oxide coating applied to the silicon steel meets the standard based on the overall oxidation resistance.

[0012] Preferably, step 1 includes the following specific steps:

[0013] Industrial cameras are used to capture images of silicon steel, and these images are then preprocessed.

[0014] The coating volatility assessment value is obtained based on the ratio of the standard deviation of gray level to the mean gray level of all pixels in the coating area.

[0015] The coating failure rate is evaluated by the ratio of the total number of pixels in the uncoated area to the total number of pixels in the silicon steel image.

[0016] The evaluation value of magnesium oxide coating application anomaly is obtained by weighted summation of coating volatility evaluation value and coating omission rate evaluation value.

[0017] Preferably, step 2 includes the following specific steps:

[0018] Collect factory environmental data, including particulate matter concentration, temperature, humidity, and wind speed;

[0019] Pollution anomaly assessment values ​​are obtained based on the ratio of suspended particulate matter concentration to the safe particulate matter concentration threshold.

[0020] The temperature anomaly assessment value is obtained by dividing the absolute value of the difference between the temperature and the safe temperature threshold by the safe temperature threshold.

[0021] The humidity anomaly assessment value is obtained by dividing the absolute value of the difference between the humidity and the safe humidity threshold by the safe humidity threshold.

[0022] The wind speed anomaly assessment value is obtained based on the ratio of wind speed to the safe wind speed threshold.

[0023] The environmental severity assessment value is obtained by weighted summation of the pollution anomaly assessment value, temperature anomaly assessment value, humidity anomaly assessment value and wind speed anomaly assessment value.

[0024] Preferably, step 3 includes the following specific steps:

[0025] The initial comprehensive damage risk assessment value of the coating is obtained by weighted summation of the assessment values ​​of magnesium oxide coating application anomaly and environmental severity.

[0026] The grayscale value of each pixel in the silicon steel image is converted into a thickness value to obtain a silicon steel coating thickness distribution map. A mapping relationship between thickness influence and damage risk weight is established, and a risk weight function is defined as follows: ,in, This is a scaling constant used to adjust the sensitivity of the weighting function. For coating thickness, For the target coating thickness, As a fixed constant, this embodiment sets it to 0.001 to prevent the denominator from being zero, and to display the silicon steel coating thickness distribution map. Substitute each thickness value into the risk weight function Obtain risk weight values, and then generate a risk weight map for silicon steel coatings based on these risk weight values. The risk weight chart is obtained as follows: The sum of the risk weight map is normalized, and the initial damage risk assessment value is then distributed point-by-point to each pixel of the silicon steel coating thickness distribution map according to the normalized weights. The calculation formula for this distribution is as follows: ,in, The initial damage risk assessment value assigned to location (i,j) is... This is the initial comprehensive damage risk assessment value for the coating. This represents the sum of all risk weight values ​​in the risk weight diagram. Output the initial damage risk state map of the silicon steel coating, where the risk weight value is at location (i,j).

[0027] The prediction duration is preset, and the damage increment applied to each pixel is obtained based on the product of the environmental severity assessment value, the environmental damage conversion coefficient and the prediction duration. The damage increment is then superimposed on the initial damage risk assessment value of each pixel to obtain the superimposed damage risk assessment value.

[0028] The superimposed damage risk assessment value is compared with the preset coating peeling critical damage value. If the superimposed damage risk assessment value exceeds the preset coating peeling critical damage value, the corresponding pixel is marked as a pore.

[0029] The stomatal merging rule is executed to remove noisy stomata with too small an area, update the stomatal distribution, generate a new list of stomatal regions after merging, and output a stomatal distribution prediction map. The stomatal merging rule is the 8-neighborhood rule.

[0030] Preferably, step 4 includes the following specific steps:

[0031] The oxygen concentration in the factory is obtained using an oxygen concentration meter, and pore data is obtained based on a pore distribution prediction map, the pore data including the number of pores and the pore area.

[0032] The oxygen entry rate is obtained based on pore data and the oxygen concentration in the factory environment. The formula for calculating the oxygen entry rate is as follows: ,in, For the effective diffusion coefficient, The number of pores. Let n be the area of ​​the nth pore. The oxygen concentration in the factory environment. The average thickness of the coating;

[0033] An abnormal oxygen intake assessment value is obtained based on the ratio of the oxygen intake rate to a preset oxygen intake rate threshold.

[0034] Silicon steel samples were collected and simulated annealing was performed to obtain the oxidation weight gain per unit area. The oxidation anomaly assessment value was analyzed based on the ratio of the preset oxidation weight gain threshold per unit area to the oxidation weight gain per unit area.

[0035] The comprehensive antioxidant anomaly assessment value is obtained by weighted summation of the oxygen entry anomaly assessment value and the antioxidant anomaly assessment value.

[0036] Preferably, step 5 includes the following specific steps:

[0037] The oxidation resistance of the magnesium oxide coating on silicon steel is determined by comparing the comprehensive oxidation resistance anomaly assessment value with the preset comprehensive oxidation resistance anomaly threshold. If the comprehensive oxidation resistance anomaly assessment value is less than the preset comprehensive oxidation resistance anomaly threshold, the oxidation resistance of the magnesium oxide coating on silicon steel is determined to be up to standard. If the comprehensive oxidation resistance anomaly assessment value is greater than or equal to the preset comprehensive oxidation resistance anomaly threshold, the oxidation resistance of the magnesium oxide coating on silicon steel is determined to be down to standard.

[0038] This invention also provides a machine vision-based magnesium oxide production evaluation system, comprising:

[0039] The coating application analysis module is used to acquire images of silicon steel, analyze coating fluctuations and coating omissions, and analyze the application of magnesium oxide coating based on coating fluctuations and coating omissions.

[0040] The environmental severity analysis module is used to collect factory environmental data and analyze environmental severity conditions.

[0041] The porosity distribution prediction module is used to analyze the initial damage risk based on the magnesium oxide coating application and harsh environmental conditions, simulate the erosion evolution process in the factory environment, and predict the porosity distribution.

[0042] The oxygen entry analysis module is used to analyze oxygen entry based on stomatal distribution and factory environment oxygen concentration, analyze antioxidant capacity based on oxidation weight gain per unit area, and analyze comprehensive antioxidant capacity based on oxygen entry and antioxidant capacity.

[0043] The production compliance judgment module is used to determine whether the oxidation resistance of the magnesium oxide coating applied to silicon steel meets the standards based on the overall oxidation resistance.

[0044] The present invention also provides an electronic device, comprising: a processor and a memory, wherein the memory stores a computer program that can be called by the processor, and the processor executes the above-described machine vision-based magnesium oxide production evaluation method by calling the computer program stored in the memory.

[0045] The present invention also provides a computer-readable storage medium storing instructions that, when executed on a computer, cause the computer to perform the above-described machine vision-based magnesium oxide production evaluation method.

[0046] Compared with existing technologies, the beneficial effects of this invention are as follows: This invention acquires images of silicon steel, analyzes coating fluctuations and coating omissions, analyzes the application of magnesium oxide coatings based on coating fluctuations and coating omissions, collects factory environmental data, analyzes harsh environmental conditions, analyzes initial damage risks based on magnesium oxide coating application and harsh environmental conditions, simulates the erosion evolution process in the factory environment, predicts porosity distribution, analyzes oxygen entry based on porosity distribution and factory environment oxygen concentration, analyzes oxidation resistance based on oxidation weight gain per unit area, analyzes comprehensive oxidation resistance based on oxygen entry and oxidation resistance, and determines whether the oxidation resistance of the magnesium oxide coating applied to silicon steel meets the standards based on the comprehensive oxidation resistance. This invention uses the initial coating condition of silicon steel and harsh environmental conditions as input variables, and predicts the distribution of porosity by simulating the erosion evolution process in the factory environment, thereby providing early warning of potential failure risks of silicon steel and improving the accuracy and timeliness of magnesium oxide production assessment. Attached Figure Description

[0047] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0048] Figure 1 This is a schematic diagram of the magnesium oxide production evaluation method based on machine vision according to the present invention.

[0049] Figure 2This is a schematic diagram of step 1 of the machine vision-based magnesium oxide production evaluation method of the present invention.

[0050] Figure 3 This is a schematic diagram of step 3 of the machine vision-based magnesium oxide production evaluation method of the present invention.

[0051] Figure 4 This is a schematic diagram of the overall framework of the machine vision-based magnesium oxide production evaluation system of the present invention. Detailed Implementation

[0052] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention are within the scope of protection of the present invention.

[0053] Please see Figure 1 This invention provides a machine vision-based method for evaluating magnesium oxide production, comprising the following specific steps:

[0054] Step 1: Acquire images of silicon steel, analyze coating fluctuations and coating omissions, and analyze the magnesium oxide coating application based on coating fluctuations and coating omissions.

[0055] Please see Figure 2 In this embodiment, step 1 includes the following specific steps:

[0056] Industrial cameras are used to capture images of silicon steel, and these images are then preprocessed.

[0057] In this embodiment, the image preprocessing process includes cropping, grayscale conversion, noise reduction, and contrast enhancement. Cropping involves removing the edges and background of the silicon steel image while retaining the coating area to be analyzed.

[0058] The coating volatility assessment value is obtained based on the ratio of the standard deviation of gray level to the mean gray level of all pixels in the coating area.

[0059] The coating failure rate is evaluated by the ratio of the total number of pixels in the uncoated area to the total number of pixels in the silicon steel image.

[0060] In this embodiment, the Otsu algorithm can be used to automatically find the optimal threshold. Pixels with gray values ​​higher than the threshold are considered as coated areas, and those lower than the threshold are considered as uncoated areas. A binary image is generated, in which white represents coated areas and black represents uncoated areas. Opening operation is performed on the binary image to remove small noise points, i.e., coated areas that are misjudged as uncoated areas. The mean and standard deviation of gray values ​​of all pixels in the coated area are calculated. The mean reflects the average thickness of the coating. The larger the standard deviation, the greater the fluctuation of the coating thickness, and the existence of problems such as uneven thickness, stripes or spots.

[0061] The evaluation value of magnesium oxide coating application anomaly is obtained by weighted summation of coating volatility evaluation value and coating omission rate evaluation value.

[0062] Step 2: Collect factory environmental data and analyze the harsh environmental conditions;

[0063] In this embodiment, step 2 includes the following specific steps:

[0064] Collect factory environmental data, including particulate matter concentration, temperature, humidity, and wind speed;

[0065] In this embodiment, the concentration of suspended particulate matter is collected by an optical particle counter to reflect the cleanliness of the factory environment.

[0066] The quality of the magnesium oxide isolation coating applied to the surface of the steel plate is crucial to the magnetic properties of the final product. However, this coating is in a very fragile state after application and before the final high-temperature annealing, and is easily affected by environmental factors. For example, dust, fibers and other particles in the air may fall onto the uncured wet coating and form bumps. After curing, these particles may fall off, leaving pinholes or pits, which will damage the continuity of the coating.

[0067] When humidity is too high, magnesium oxide reacts with water vapor in the air to form magnesium hydroxide. The loose structure of magnesium hydroxide leads to poor coating adhesion, making it easy to fall off and pulverize during subsequent winding and transportation. Excessive moisture in the coating reacts with the steel base at high temperatures to form an abnormal oxide layer, affecting magnetic properties. After winding, the undried coating will cause the steel strip layers to stick together and cannot be separated during annealing, resulting in scrap. When humidity is too low, the coating surface shrinks rapidly, forming tensile stress, which causes micro-cracks in the coating and destroys its integrity. Dust attracted by electrostatics will embed in the coating, forming contamination points.

[0068] When the temperature is too high, the solvent in the coating slurry evaporates too quickly, causing the coating surface to form a skin rapidly. The internal solvent cannot escape, resulting in orange peel or bubbles.

[0069] If there is a strong, uneven airflow, it will cause uneven solvent evaporation on the wet coating surface, resulting in streaks or uneven thickness. Turbulent airflow may also bring contaminants from the external area into the clean coating area.

[0070] The pollution anomaly assessment value is obtained based on the ratio of suspended particulate matter concentration to the safe particulate matter concentration threshold. In this embodiment, during specific implementation, experiments showed that when the concentration of particles greater than or equal to 0.5 μm is below 250,000 particles / m³... 3 At this time, the adhesion defect rate can be controlled below 0.05%, therefore, the safe particulate matter concentration threshold can be set at 250,000 particles / m³. 3 ;

[0071] The temperature anomaly assessment value is obtained by dividing the absolute value of the difference between the temperature and the safe temperature threshold by the safe temperature threshold. In this embodiment, the safe temperature threshold can be 26°C.

[0072] The humidity anomaly assessment value is obtained by dividing the absolute value of the difference between the humidity and the safe humidity threshold by the safe humidity threshold. In this embodiment, the safe humidity threshold can be 40%RH.

[0073] The wind speed anomaly assessment value is obtained based on the ratio of wind speed to a safe wind speed threshold. In this embodiment, the safe wind speed threshold can be 0.35 m / s.

[0074] The environmental severity assessment value is obtained by weighted summation of the pollution anomaly assessment value, temperature anomaly assessment value, humidity anomaly assessment value and wind speed anomaly assessment value.

[0075] Step 3: Analyze the initial damage risk based on the magnesium oxide coating application and harsh environmental conditions, simulate the erosion evolution process in the factory environment, and predict the porosity distribution.

[0076] Please see Figure 3 In this embodiment, step 3 includes the following specific steps:

[0077] The initial comprehensive damage risk assessment value of the coating is obtained by weighted summation of the assessment values ​​of magnesium oxide coating application anomaly and environmental severity.

[0078] The grayscale value of each pixel in the silicon steel image is converted into a thickness value to obtain a thickness distribution map of the silicon steel coating. A mapping relationship between thickness influence and damage risk weight is established, meaning that areas with darker grayscale colors (higher values) in the silicon steel image are assigned a greater risk weight. A risk weight function is defined as follows: ,in, The scaling constant is used to adjust the sensitivity of the weighting function. It is obtained by collecting thickness distribution maps and corresponding defect distribution maps of a batch of silicon steel samples. An objective function is defined, which can be to maximize the average value of the risk map in the actual defect area while minimizing the average value in the defect-free area. A grid search is used to find the optimal value for the objective function. value, For coating thickness, The target coating thickness is obtained from the coating type's process specification document. As a fixed constant, this embodiment sets it to 0.001 to prevent the denominator from being zero and to ensure the stability of the numerical calculation. The silicon steel coating thickness distribution map is then used. Substitute each thickness value into the risk weight function Obtain risk weight values, and then generate a risk weight map for silicon steel coatings based on these risk weight values. The risk weight chart is obtained as follows: The sum of the risk weight map is normalized, and the initial damage risk assessment value is then distributed point-by-point to each pixel of the silicon steel coating thickness distribution map according to the normalized weights. The calculation formula for the distribution is as follows: ,in, The initial damage risk assessment value assigned to location (i,j) is... This is the initial comprehensive damage risk assessment value for the coating. This represents the sum of all risk weight values ​​in the risk weight diagram. Given the risk weight value at position (i,j), output the initial damage risk status map of the silicon steel coating. The color or value of each pixel in the map represents the level of initial damage risk at that position. The brighter the color (or the higher the value), the greater the initial risk at that position.

[0079] In this embodiment, a silicon steel sample with a known coating thickness is prepared, and an image of the silicon steel is acquired in the same manner. Its average grayscale value is measured, and the thickness value is fitted with the corresponding grayscale value to obtain the functional relationship between the grayscale value and the thickness value.

[0080] The prediction duration is preset, and the damage increment applied to each pixel is obtained based on the product of the environmental severity assessment value, the environmental damage conversion coefficient and the prediction duration. The damage increment is then superimposed on the initial damage risk assessment value of each pixel to obtain the superimposed damage risk assessment value.

[0081] The superimposed damage risk assessment value is compared with the preset coating peeling critical damage value. If the superimposed damage risk assessment value exceeds the preset coating peeling critical damage value, the corresponding pixel is marked as a pore.

[0082] The stomatal merging rule is executed to remove noisy stomata with too small an area, update the stomatal distribution, generate a new list of stomatal regions after merging, and output a stomatal distribution prediction map. The stomatal merging rule is the 8-neighborhood rule.

[0083] In this embodiment, the connectedComponents function is used to mark adjacent vent pixels as the same region by setting the connectivity parameter to 8.

[0084] Step 4: Analyze oxygen entry based on stomatal distribution and factory environment oxygen concentration; analyze antioxidant capacity based on oxidation weight gain per unit area; and analyze comprehensive antioxidant capacity based on oxygen entry and antioxidant capacity.

[0085] In this embodiment, step 4 includes the following specific steps:

[0086] Oxygen concentration in the factory is obtained using an oxygen concentration meter, and stoma data is obtained based on a stoma distribution prediction map. The stoma data includes the number of pores and the area of ​​pores.

[0087] The oxygen entry rate is obtained based on pore data and the oxygen concentration in the factory environment. The formula for calculating the oxygen entry rate is as follows: ,in, The effective diffusion coefficient reflects the rate of oxygen diffusion in the pores. It is obtained through the effective diffusion coefficient estimation formula, which can be expressed as: ,in, Porosity is the percentage of pore area to the total coating area. Tortuosity describes the degree of curvature of the gas path; a value greater than 1. For disordered porous media, an empirical relationship can be used: , The diffusion coefficient of oxygen in air can be taken as 2.0 × 10⁻⁶. -5 m 2 / s, The number of pores. Let n be the area of ​​the nth pore. The oxygen concentration in the factory environment. The average thickness of the coating;

[0088] An abnormal oxygen intake assessment value is obtained based on the ratio of the oxygen intake rate to a preset oxygen intake rate threshold.

[0089] Silicon steel samples were collected and simulated annealing was performed to obtain the oxidation weight gain per unit area. The oxidation anomaly assessment value was analyzed based on the ratio of the preset oxidation weight gain threshold per unit area to the oxidation weight gain per unit area.

[0090] In this embodiment, the steps of the simulated annealing operation are as follows: a silicon steel sample of a set size covered with a magnesium oxide coating is cut from the production line, the sample is placed in a controlled atmosphere tube furnace or box furnace, a temperature (1150°C-1200°C) and a gas atmosphere of high-purity nitrogen and trace hydrogen are set to simulate the factory annealing environment, and a controllable amount of oxygen or water vapor is introduced as an oxidation source. After maintaining the set predicted time, the sample is taken out and cooled, and the mass change of the sample before and after the experiment is measured. The increase in mass per unit area is the amount of oxidation. Under standard experimental conditions, the lower the mass increase per unit area and per unit time, the higher the antioxidant function value.

[0091] The comprehensive antioxidant anomaly assessment value is obtained by weighted summation of the oxygen entry anomaly assessment value and the antioxidant anomaly assessment value.

[0092] The oxygen ingress anomaly assessment value quantifies the risk of oxygen intrusion under the combined effects of environmental factors and microscopic defects in the coating. The oxidation resistance anomaly assessment value quantifies the coating material's inherent resistance to oxidation, determined by its composition, density, and adhesion to the substrate. By comprehensively assessing the oxygen ingress anomaly assessment value and the oxidation resistance anomaly assessment value, missed and false judgments can be reduced. For example, a missed judgment may occur if the coating itself is of good quality, but the current batch has increased porosity due to extremely harsh environmental conditions or accidental process fluctuations. Or, a false judgment may occur if some minor coating fluctuations are detected, but the current factory environment is ideal and there are very few pores.

[0093] Step 5: Determine whether the oxidation resistance of the magnesium oxide coating applied to the silicon steel meets the standard based on the overall oxidation resistance.

[0094] In this embodiment, step 5 includes the following specific steps:

[0095] The oxidation resistance of the magnesium oxide coating on silicon steel is determined by comparing the comprehensive oxidation resistance anomaly assessment value with the preset comprehensive oxidation resistance anomaly threshold. If the comprehensive oxidation resistance anomaly assessment value is less than the preset comprehensive oxidation resistance anomaly threshold, the oxidation resistance of the magnesium oxide coating on silicon steel is determined to be up to standard. If the comprehensive oxidation resistance anomaly assessment value is greater than or equal to the preset comprehensive oxidation resistance anomaly threshold, the oxidation resistance of the magnesium oxide coating on silicon steel is determined to be down to standard.

[0096] In this embodiment, the weights, coefficients, and preset thresholds can be obtained as follows: acquire historical silicon steel images and historical factory environment data, and simultaneously acquire historical oxygen entry rates to determine whether the oxidation resistance of the magnesium oxide coating meets production standards. Import the acquired historical silicon steel images and historical factory environment data into each step of this embodiment to obtain the judgment results on whether the oxidation resistance of the magnesium oxide coating meets production standards. Import the two judgment results into MATLAB fitting software for fitting to obtain the values ​​of weights, coefficients, and preset thresholds that meet the maximum judgment accuracy.

[0097] Please see Figure 4 This invention also provides a machine vision-based magnesium oxide production evaluation system, including:

[0098] The coating application analysis module is used to acquire images of silicon steel, analyze coating fluctuations and coating omissions, and analyze the application of magnesium oxide coating based on coating fluctuations and coating omissions.

[0099] The environmental severity analysis module is used to collect factory environmental data and analyze environmental severity conditions.

[0100] The porosity distribution prediction module is used to analyze the initial damage risk based on the magnesium oxide coating application and harsh environmental conditions, simulate the erosion evolution process in the factory environment, and predict the porosity distribution.

[0101] The oxygen entry analysis module is used to analyze oxygen entry based on stomatal distribution and factory environment oxygen concentration, analyze antioxidant capacity based on oxidation weight gain per unit area, and analyze comprehensive antioxidant capacity based on oxygen entry and antioxidant capacity.

[0102] The production compliance judgment module is used to determine whether the oxidation resistance of the magnesium oxide coating applied to silicon steel meets the standards based on the overall oxidation resistance.

[0103] This invention also provides an electronic device, including a processor and a memory, wherein the memory stores a computer program that can be called by the processor, and the processor executes the above-described machine vision-based magnesium oxide production evaluation method by calling the computer program stored in the memory.

[0104] The electronic device can vary considerably depending on its configuration or performance. It may include one or more Central Processing Units (CPUs) and one or more memories, wherein the memory stores at least one computer program, which is loaded and executed by the processor to implement the machine vision-based magnesium oxide production evaluation method provided in the above-described method embodiments. The electronic device may also include other components for implementing its functions. For example, it may have wired or wireless network interfaces and input / output interfaces for data input and output, which will not be elaborated upon in this embodiment.

[0105] This invention also provides a computer-readable storage medium storing instructions that, when a computer program is run on a computer device, cause the computer device to execute the aforementioned machine vision-based magnesium oxide production evaluation method.

[0106] For example, computer-readable storage media can be read-only memory (ROM), random access memory (RAM), compact disc read-only memory (CD-ROM), magnetic tape, floppy disk, and optical data storage devices.

[0107] It should be understood that, in various embodiments of the present invention, the order of the above-mentioned process numbers does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0108] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any other combination. When implemented by software, the above embodiments can be implemented in whole or in part as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, they generate in whole or in part the flow or function according to the embodiments of the present invention. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via a wired network and / or wireless network. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that contains one or more sets of available media. The available media can be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., DVDs), or semiconductor media. The semiconductor media can be a solid-state drive.

[0109] In the description of this specification, references to terms such as "an embodiment," "example," "specific example," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example, and the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0110] The preferred embodiments of the present invention disclosed above are only for the purpose of illustrating the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to specific implementation methods. Obviously, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present invention, so that those skilled in the art can better understand and utilize the present invention. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. A method for the evaluation of magnesium oxide production based on machine vision, characterized in that, The method comprises the following specific steps: Step 1: Collecting a silicon steel image, analyzing coating fluctuation and coating missing conditions, and analyzing magnesium oxide coating coating conditions based on the coating fluctuation and coating missing conditions; Step 2: Collecting factory environment data and analyzing severe environment conditions; Step 3: Analyzing initial damage risks based on the magnesium oxide coating coating conditions and the severe environment conditions, simulating a factory environment erosion evolution process, and predicting pore distribution conditions; Step 4: Analyzing oxygen entry conditions based on the pore distribution conditions and the factory environment oxygen concentration, analyzing oxidation resistance conditions based on unit area oxidation weight gain, and analyzing comprehensive oxidation resistance conditions based on the oxygen entry conditions and the oxidation resistance conditions; Step 5: Judging whether the oxidation resistance ability of the magnesium oxide coating coated on the silicon steel meets the standard based on the comprehensive oxidation resistance conditions; The step 3 comprises the following specific steps: A coating initial comprehensive damage risk evaluation value is obtained by weighted summation based on a magnesium oxide coating coating abnormality evaluation value and an environment severity evaluation value; The gray value of each pixel of the silicon steel image is converted into a thickness value to obtain a silicon steel coating thickness distribution map, a mapping relationship between thickness influence and damage risk weight is established, and a risk weight function is defined, and the risk weight function is expressed as: wherein, is a scaling constant, is a coating thickness, is a target coating thickness, is a fixed constant, each thickness value of the silicon steel coating thickness distribution map is substituted into the risk weight function to obtain a risk weight value, a risk weight map of the silicon steel coating is obtained based on the risk weight value, a sum of the risk weight map is normalized, and an initial damage risk assessment value is point by point distributed to each pixel of the silicon steel coating thickness distribution map according to the normalized weight, and a calculation formula of the distribution is: wherein, is an initial damage risk assessment value distributed to a position (i, j), is a coating initial comprehensive damage risk assessment value, is a sum of all risk weight values of the risk weight map, is a risk weight value at the position (i, j), and an initial damage risk state map of the silicon steel coating is output. A preset prediction time length is obtained, a damage increment acting on each pixel point is obtained based on a product of the environment severity evaluation value, an environment damage conversion coefficient and the prediction time length, the damage increment is added to an initial damage risk evaluation value of each pixel, and a superimposed damage risk evaluation value is obtained; The superimposed damage risk evaluation value is compared with a preset coating peeling critical damage value, if the superimposed damage risk evaluation value exceeds the preset coating peeling critical damage value, the corresponding pixel point is marked as a pore; A pore merging rule is executed, a new pore region list is generated after merging, and a pore distribution prediction map is output, wherein the pore merging rule is an 8-neighborhood rule.

2. The machine vision-based magnesium oxide production evaluation method according to claim 1, characterized by, The step 1 comprises the following specific steps: An industrial camera is used to collect a silicon steel image and the silicon steel image is preprocessed; A coating fluctuation evaluation value is obtained based on a ratio of a gray scale standard deviation to a gray scale mean value of all pixels in a coating region; A coating missing rate evaluation value is obtained based on a ratio of a total number of pixels in a coating missing region to a total number of pixels in the silicon steel image; A magnesium oxide coating coating abnormality evaluation value is obtained by weighted summation based on the coating fluctuation evaluation value and the coating missing rate evaluation value.

3. The machine vision-based magnesium oxide production evaluation method according to claim 2, characterized by, The step 2 comprises the following specific steps: Factory environment data is collected, and the factory environment data comprises a suspended particulate matter concentration, a temperature, a humidity and a wind speed; A pollution abnormality evaluation value is obtained based on a ratio of the suspended particulate matter concentration to a safe particulate matter concentration threshold value; A temperature abnormality evaluation value is obtained based on an absolute value of a difference between the temperature and a safe temperature threshold value divided by the safe temperature threshold value; A humidity abnormality evaluation value is obtained based on an absolute value of a difference between the humidity and a safe humidity threshold value divided by the safe humidity threshold value; A wind speed abnormality evaluation value is obtained based on a ratio of the wind speed to a safe wind speed threshold value; An environment severity evaluation value is obtained by weighted summation based on the pollution abnormality evaluation value, the temperature abnormality evaluation value, the humidity abnormality evaluation value and the wind speed abnormality evaluation value.

4. The machine vision-based magnesium oxide production evaluation method according to claim 3, characterized by, The step 4 comprises the following specific steps: A factory oxygen concentration is obtained, pore data is obtained based on the pore distribution prediction map, and the pore data comprises a pore number and a pore area; The oxygen ingress rate is obtained based on the pore data and the oxygen concentration of the factory environment, and the oxygen ingress rate is calculated by the following formula: wherein, is the effective diffusion coefficient, is the number of pores, is the area of the nth pore, is the oxygen concentration of the factory environment, is the average thickness of the coating. An oxygen entry abnormality evaluation value is obtained based on a ratio of an oxygen entry rate to a preset oxygen entry rate threshold value; The silicon steel sample is collected for simulation annealing operation to obtain the unit area oxidation weight gain, and the anti-oxidation abnormality evaluation value is analyzed based on the ratio of the preset unit area oxidation weight gain threshold value to the unit area oxidation weight gain; The comprehensive anti-oxidation abnormality evaluation value is obtained by weighted summation based on the oxygen entry abnormality evaluation value and the anti-oxidation abnormality evaluation value.

5. The machine vision-based magnesium oxide production evaluation method according to claim 4, characterized by, The step 5 comprises the following specific steps: The anti-oxidation ability of the magnesium oxide coating coated on the silicon steel is judged based on the comparison result of the comprehensive anti-oxidation abnormality evaluation value and the preset comprehensive anti-oxidation abnormality threshold value, if the comprehensive anti-oxidation abnormality evaluation value is less than the preset comprehensive anti-oxidation abnormality threshold value, it is judged that the anti-oxidation ability of the magnesium oxide coating coated on the silicon steel meets the standard, if the comprehensive anti-oxidation abnormality evaluation value is greater than or equal to the preset comprehensive anti-oxidation abnormality threshold value, it is judged that the anti-oxidation ability of the magnesium oxide coating coated on the silicon steel does not meet the standard.

6. A machine vision-based magnesium oxide production evaluation system for implementing the machine vision-based magnesium oxide production evaluation method according to any one of claims 1 to 5, characterized by Comprise: The coating coating analysis module is used for collecting silicon steel image, analyzing coating fluctuation and coating missing situation, and analyzing magnesium oxide coating coating situation based on coating fluctuation and coating missing situation; The environment analysis module is used for collecting factory environment data and analyzing environment situation; The pore distribution prediction module is used for analyzing initial damage risk based on magnesium oxide coating coating situation and environment situation, simulating factory environment erosion evolution process, and predicting pore distribution situation; The oxygen entry analysis module is used for analyzing oxygen entry situation based on pore distribution situation and factory environment oxygen concentration, analyzing anti-oxidation situation based on unit area oxidation weight gain, and analyzing comprehensive anti-oxidation situation based on oxygen entry situation and anti-oxidation situation; The production standard judgment module is used for judging whether the anti-oxidation ability of the magnesium oxide coating coated on the silicon steel meets the standard based on the comprehensive anti-oxidation situation.

7. An electronic device, comprising: Comprise: A memory and a processor, wherein the memory stores a computer program that can be called by the processor, and the processor executes the magnesium oxide production evaluation method based on machine vision in any one of claims 1-5 by calling the computer program stored in the memory.

8. A computer-readable storage medium, characterized in that, The instructions are stored, and when the instructions run on the computer, the computer executes the magnesium oxide production evaluation method based on machine vision in any one of claims 1-5.

Images