Mica paper pulp quality control method based on image recognition, medium and system

Abstract

The invention discloses a mica paper pulp quality control method based on image recognition, a medium and a system, and relates to the field of image recognition, and the method comprises the following steps: obtaining a target area image of mica paper pulp through operation of a vibrator at the bottom of a mica paper pulp pool, and carrying out image recognition on the target area image to obtain vibration ripple characteristics of the target area image; determining a first concentration range of mica paper pulp according to the vibration ripple characteristics of the target area image; incident light is emitted to a target area; acquiring a current image of the target area; determining the number of scales on the surface of the mica paper pulp through the current image; determining a second concentration range of the mica paper pulp according to the scale number on the surface of the mica paper pulp; and determining a final concentration range of the mica paper pulp by combining the first concentration range and the second concentration range of the mica paper pulp so as to generate a quality control strategy according to the final concentration range. The mica paper pulp quality control method can effectively improve the accuracy of mica paper pulp quality control.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of image recognition, and in particular to a mica pulp quality control method, medium and system based on image recognition. Background Art

[0002] Mica paper has the characteristics of electrical insulation and thermal insulation, and is widely used in electrical, construction, aerospace, chemical, energy and other fields. The quality control of mica pulp is a crucial step in the production process of mica paper.

[0003] The existing technology mainly controls the quality of mica pulp through manual inspection. During the mica pulp production process, the concentration of mica pulp and impurities in mica pulp are observed by manual observation to ensure the quality of mica pulp. However, manual inspection with the naked eye will cause visual fatigue and the subjective influence of each person on the quality inspection of mica pulp. These factors will affect the effect of mica pulp quality inspection and lead to poor quality of mica paper products.

[0004] Therefore, there is an urgent need for a mica pulp quality control method, medium and system based on image recognition. Summary of the invention

[0005] The embodiments of the present application provide a mica pulp quality control method, medium and system based on image recognition, which are used to solve the problem of inaccurate mica pulp quality detection.

[0006] To achieve the above objectives, the embodiments of the present application adopt the following technical solutions: In a first aspect, a mica pulp quality control method based on image recognition is provided, the method comprising: issuing a first control instruction to control the operation of a vibrator disposed at the bottom of a mica pulp pool; Acquire a target area image of mica pulp, perform image recognition on the target area image, and obtain vibration ripple features of the target area image, wherein the vibration ripple features include the number of vibration ripples and the amplitude of vibration ripples; Determining a first concentration range of mica pulp according to the vibration ripple characteristics of the target area image, and judging whether the first concentration range is within a preset concentration range; If the first concentration range is not within the preset concentration range, issuing a warning instruction; If the first concentration range is within a preset concentration range, after the mica pulp is left to stand for a preset period of time, a second control instruction is issued to control the light source to emit incident light to a target area; Acquire a current image of the target area; Determining the number of scales on the surface of the mica pulp through the current image; Determining a second concentration range of the mica pulp according to the number of scales on the surface of the mica pulp; The final concentration range of the mica pulp is determined by combining the first concentration range and the second concentration range of the mica pulp to generate a quality control strategy according to the final concentration range.

[0007] In a possible implementation manner of the first aspect, performing image recognition on the target area image to obtain the vibration ripple feature of the target area includes: Dividing the target area image into a plurality of unit grid images, and preprocessing each of the unit grid images to obtain a corresponding grayscale unit grid image; Image recognition is performed on each of the grayscale unit grid images to obtain the number of vibration ripples and the amplitude of the vibration ripples.

[0008] In another possible implementation manner of the first aspect, after preprocessing each of the unit grid images to obtain a corresponding grayscale unit grid image, the method further includes: For any one of the grayscale unit grid images, performing denoising processing on the grayscale unit grid image; Performing feature extraction on the grayscale unit grid image after denoising to obtain target impurity edge features; Acquire a first pixel point of the target impurity edge feature and a second pixel point of the mica flake edge feature; Calculating the Euclidean distance and color contrast between the first pixel and the second pixel, and determining whether the target area is a polluted area according to the Euclidean distance and color contrast between the first pixel and the second pixel; If the Euclidean distance and the color contrast are both greater than a preset threshold, the target area is determined to be a polluted area, and an alarm message is issued; If the Euclidean distance and the color contrast are both less than or equal to a preset threshold, it is determined that the target area is not a polluted area, and the step of performing image recognition on each of the grayscale unit grid images to obtain the number of vibration ripples and the amplitude of the vibration ripples is performed.

[0009] In another possible implementation of the first aspect, determining the first concentration range of the mica pulp according to the vibration ripple characteristics of the target area includes: Obtaining the number of vibration ripples and the amplitude of vibration ripples according to the vibration ripple characteristics of the target area; Comparing the number of vibration ripples with the number of vibration ripples of a preset concentration to obtain a first concentration of mica pulp; Comparing the vibration ripple amplitude with the vibration ripple amplitude of a preset concentration to obtain a second concentration of the mica pulp; The first concentration range of the mica pulp is determined in combination with the first concentration of the mica pulp and the second concentration of the mica pulp.

[0010] In another possible implementation manner of the first aspect, if the first concentration range is not within a preset concentration range, issuing a warning instruction of mica pulp concentration includes: Comparing the first concentration range with a preset concentration range to determine a warning level of the first concentration range; Based on the warning level, a warning instruction for mica pulp concentration is issued.

[0011] In another possible implementation manner of the first aspect, determining the number of scales on the surface of the mica pulp by using the current image includes: Performing grayscale processing on the current image to obtain a grayscale image, wherein the current image includes a plurality of reflected light points obtained based on the incident light; Acquire a third pixel point of the grayscale image and a fourth pixel point of the background image of the mica pulp surface; acquire pixel points of the current image, and identify color features of all the pixel points of the current image; Determine the pixel point of the reflected light point according to the color feature; According to the pixel points of the reflected light spots, a preset edge detection algorithm is used to determine the edges of the reflected light spots, and according to the edges of the reflected light spots, the number of the reflected light spots is obtained, wherein the number of the reflected light spots is used to characterize the number of scales on the surface of the mica pulp.

[0012] In another possible implementation of the first aspect, determining the second concentration range of the mica pulp according to the number of scales on the surface of the mica pulp includes: The number of scales on the surface of the mica pulp is retrieved in a preset database to obtain a second concentration range of the mica pulp corresponding to the number of scales on the surface of the mica pulp.

[0013] In another possible implementation of the first aspect, determining a final concentration range of the mica pulp by combining the first concentration range and the second concentration range of the mica pulp includes: When the second concentration range of the mica pulp includes the first concentration range, the second concentration range is used as the final concentration range of the mica pulp; When the first concentration range and the second concentration range of the mica pulp overlap, the overlapping range of the first concentration range and the second concentration range is used as the final concentration range of the mica pulp; When the first concentration range and the second concentration range of the mica pulp are adjacent to each other, a range obtained by adding the first concentration range and the second concentration range is used as the final concentration range of the mica pulp.

[0014] In a second aspect, the present application provides a machine-readable storage medium having instructions stored thereon, the instructions being used to enable a machine to execute the above-mentioned mica pulp quality control method based on image recognition.

[0015] In a third aspect, the present application provides a mica pulp quality control system based on image recognition, comprising: a memory configured to store instructions; and The processor is configured to call the instructions from the memory and implement the above-mentioned mica pulp quality control method based on image recognition when executing the instructions.

[0016] Through the above technical solution, by using the vibrator at the bottom of the mica pulp pool to vibrate the mica pulp, the target area on the surface of the mica pulp is obtained, and based on the image recognition method, the vibration ripple characteristics of the target area image are determined, and the relationship between the mica pulp concentration and the vibration ripple characteristics can be obtained, so that the quality detection and control of the mica pulp concentration is more accurate. According to the vibration ripple characteristics, the first concentration range of the mica pulp is determined. By obtaining the first concentration range of the mica pulp, it is helpful to make the concentration analysis of the mica pulp more accurate, so that the quality control effect of the mica pulp is better. When the first concentration range is not within the preset concentration range, an early warning instruction is issued, which can warn the staff in advance and help to timely discover the potential risks of the mica pulp. By obtaining the number of scales on the surface of the mica pulp, the second concentration range of the mica pulp is determined, and the final concentration range of the mica pulp is determined. According to the final concentration range, a quality control strategy is generated, and the concentration of the mica pulp can be regulated, effectively ensuring the accuracy of the concentration of the mica pulp in the production process.

[0017] Other features and advantages of the embodiments of the present application will be described in detail in the subsequent specific implementation section. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 A schematic diagram of a process flow of a mica pulp quality control method based on image recognition provided in an embodiment of the present application; Figure 2 A schematic diagram of a process for detecting contaminated areas in a target area provided in an embodiment of the present application. DETAILED DESCRIPTION

[0019] In order to make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. It should be understood that the specific implementation methods described herein are only used to illustrate and explain the embodiments of the present application, and are not used to limit the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0020] It should be noted that if the embodiments of the present application involve directional indications (such as up, down, left, right, front, back...), such directional indications are only used to explain the relative position relationship, movement status, etc. between the components under a certain specific posture (as shown in the accompanying drawings). If the specific posture changes, the directional indication will also change accordingly.

[0021] In addition, if there are descriptions involving "first", "second", etc. in the embodiments of the present application, the descriptions of "first", "second", etc. are only used for descriptive purposes and cannot be understood as indicating or suggesting their relative importance or implicitly indicating the number of technical features indicated. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In addition, the technical solutions between the various embodiments can be combined with each other, but they must be based on the ability of ordinary technicians in the field to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be deemed that such combination of technical solutions does not exist and is not within the scope of protection required by this application.

[0022] Figure 1 The following schematically shows a flow chart of a mica pulp quality control method based on image recognition according to an embodiment of the present application. Figure 1 As shown, an embodiment of the present application provides a mica pulp quality control method based on image recognition, which may include the following steps.

[0023] S110, issuing a first control instruction to control the vibrator disposed at the bottom of the mica pulp pool to operate; S120, acquiring a target area image of mica pulp, performing image recognition on the target area image, and obtaining a vibration ripple feature of the target area image, wherein the vibration ripple feature includes the number of vibration ripples and the amplitude of the vibration ripples; S130, determining a first concentration range of mica pulp according to the vibration ripple characteristics of the target area image, and judging whether the first concentration range is within a preset concentration range; S140, if the first concentration range is not within the preset concentration range, issuing a warning instruction; S150, if the first concentration range is within the preset concentration range, after the mica pulp is left to stand for a preset period of time, a second control instruction is issued to control the light source to emit incident light to the target area; S160, obtaining a current image of the target area; S170, determining the number of scales on the surface of the mica pulp through the current image; S180, determining a second concentration range of the mica pulp according to the number of scales on the surface of the mica pulp; S190, determining a final concentration range of the mica pulp in combination with the first concentration range and the second concentration range, so as to generate a quality control strategy according to the final concentration range.

[0024] The mica pulp is vibrated by issuing a first control instruction to the vibrator disposed at the bottom of the mica pulp pool. In this embodiment, the first control instruction refers to a control instruction to operate the vibrator at the bottom of the mica pulp pool. After the mica pulp vibrates, an image of a target area of ​​the mica pulp is obtained. The image of the target area is subjected to image recognition technology to obtain vibration ripple features of the image of the target area. In this embodiment, the image of the target area is an image of the surface of the mica pulp after the vibrator is working. Image recognition refers to analyzing and understanding an image using a computer and artificial intelligence technology to automatically identify and classify information such as objects, scenes, and patterns in the image. The vibration ripple features of the target area image are obtained by obtaining the image of the target area of ​​the mica pulp and based on the image recognition technology. In this embodiment, the vibration ripple features include the number of vibration ripples and the amplitude of vibration ripples.

[0025] According to the vibration ripple characteristics of the target area image, the first concentration range of mica pulp is determined, and it is judged whether the first concentration range is within the preset concentration range. In this embodiment, the first concentration range of mica pulp refers to the concentration range of mica pulp obtained by analyzing the vibration ripple characteristics of the target area image of mica pulp. The concentration of mica pulp refers to the concentration of solid matter of mica pulp in water during pulping. The concentration of mica pulp is a relative concept, which is the ratio of mica and other fibers contained in every 100 ml of pulp. According to the vibration ripple characteristics of the target area image, the first concentration range of mica pulp is determined, that is, by comparing the number of vibration ripples and the amplitude of vibration ripples in the target area with the number of vibration ripples of the preset concentration, the first concentration of mica pulp is obtained. In this embodiment, the preset concentration can be determined according to the time situation. The number of vibration ripples of the preset concentration is the number of vibration ripples corresponding to each preset concentration stored in the database. For example, the number of vibration ripples in the target area is 35, and the concentration preset in the database is 2%, and the corresponding number of vibration ripples is 35. Therefore, the first concentration of mica pulp is determined to be 2%. Similarly, the vibration ripple amplitude is compared with the vibration ripple amplitude of the preset concentration to obtain the second concentration of the mica pulp. In this embodiment, the vibration ripple amplitude of the preset concentration is the vibration ripple amplitude corresponding to each preset concentration stored in the database. For example, the vibration ripple amplitude of the target area is 10 cm, and the preset concentration in the database is 3%, and the corresponding vibration ripple amplitude is 10 cm. Therefore, the second concentration of the mica pulp is determined to be 3%. Combined with the first concentration and the second concentration of the mica pulp, the first concentration range of the mica pulp is determined. For example, the first concentration of the mica pulp is 2%, and the second concentration of the mica pulp is 3%. The first concentration range of the mica pulp is determined to be 2%~3%. After determining the first concentration range of the mica pulp, it is determined whether the first concentration range is within the preset concentration range. In this embodiment, the preset concentration range can be determined according to actual conditions. For example, the preset concentration range is 2%~5%, and the first concentration range of the mica pulp is 2%~3%. It can be determined that the first concentration range is within the preset concentration range.

[0026] If the first concentration range is not within the preset concentration range, an early warning instruction is issued. The mica pulp concentration refers to the concentration of the solid matter of the mica pulp in the water during pulping. The mica pulp concentration is a relative concept, which is the ratio of mica and other fibers contained in every 100 ml of pulp. Specifically, the obtained first concentration range of the mica pulp is compared with the preset concentration range to determine the early warning level of the first concentration range. For example, the first concentration range of the mica pulp is 1%~3%, and the preset concentration range is 5%~7%. The first concentration range is not within the preset concentration range. The average value of the first concentration range is 2%, and the average value of the preset concentration range is 6%. The preset concentration range is 3 times the first concentration range. The early warning level can be a primary early warning level, an intermediate early warning level, and a senior early warning level. When the preset concentration range is 3 times the first concentration range, it is a primary early warning level, and a primary early warning instruction of the mica pulp concentration is issued. The instruction of issuing the primary early warning level can be notified to the staff at the production site by broadcasting to give an early warning prompt.

[0027] When it is determined that the first concentration range is not within the preset concentration range and an early warning instruction is issued, the staff will adjust the first concentration range of the mica pulp. First, the concentration of the first concentration range is determined to determine whether the first concentration range is greater than or less than the preset concentration range. When the first concentration range is greater than the preset concentration range, it indicates that the concentration of the mica pulp solid matter in the water in the first concentration range is too high, and it is necessary to increase the amount of water appropriately to reduce the pulp concentration. Specifically, first determine the total water supply required for dilution according to the amount of pulp in the stirring tank, and set the water supply rate. Stir the pulp and start water supply at the same time. Determine the current water supply rate based on the current water consumption. This method is divided into two stages: the first stage is carried out at a lower water supply rate. When the water consumption reaches 40%-70% of the total amount, switch to the high water supply rate of the second stage to ensure uniform mixing of the pulp and water, and determine whether the first concentration range is within the preset concentration range, until the first concentration range of the mica pulp is within the preset concentration range, stop adjusting the concentration of the mica pulp.

[0028] When the first concentration range is smaller than the preset concentration range, it indicates that the concentration of mica pulp solid matter in water in the first concentration range is too low, and the proportion of mica and other fibers needs to be increased appropriately to increase the pulp concentration. In addition, the concentration of mica pulp can also be adjusted by mechanical pressing, which is a commonly used pulp concentration adjustment method. In this method, mechanical pressing equipment can be used to press the pulp to gradually discharge the water therein, thereby adjusting the pulp concentration. After the mechanical pressing equipment is used to squeeze out the water therein, it is determined whether the first concentration range is within the preset concentration range, and the concentration of the mica pulp is stopped until the first concentration range of the mica pulp is within the preset concentration range. ‌‌ If the first concentration range is within the preset concentration range, after the mica pulp is allowed to stand for a preset time period, a second control instruction is issued to control the light source to emit incident light to the target area. In this embodiment, the second control instruction refers to an instruction to control the light source to emit incident light to the target area. The preset time period refers to a time period when the vibration ripples on the surface of the mica pulp disappear. That is to say, when the first concentration range is within the preset concentration range, after the mica pulp is allowed to stand for a preset time period, when the vibration ripples on the surface of the mica pulp disappear, the light source emits incident light to the target area to reduce the impact of the vibration ripples of the mica pulp on the incident light. In this embodiment, the light source can be an artificial light source, such as a fluorescent lamp and a laser, and a current image of the target area of ​​the mica pulp is acquired.

[0029] The number of scales on the surface of mica pulp is determined by the current image. Specifically, first, grayscale processing is performed on the current image to obtain a grayscale image, wherein the current image includes a plurality of reflected light points obtained based on the incident light. The third pixel point of the grayscale image and the fourth pixel point of the background image of the mica pulp surface are obtained. In this embodiment, the third pixel point refers to the pixel point of the light source reflection point formed by controlling the light source to emit incident light to the mica pulp surface in the grayscale image, and the fourth pixel point refers to the pixel point in the background of the mica pulp surface image. The third pixel point and the fourth pixel point can be obtained by image recognition. Secondly, the pixel points of the current image are obtained, and the color characteristics of all the pixel points of the current image are identified, for example, the color characteristics of the pixel points of the light source reflection point are black, and the color characteristics of the pixel points in the background of the mica pulp surface image are white, and the pixel points of the reflected light point are determined according to the color characteristics. The edge of the pixel point of the reflected light point is obtained by using a preset edge detection algorithm, and the number of the reflected light points is obtained according to the edge of the reflected light point, wherein the number of the reflected light points is used to characterize the number of scales on the surface of the mica pulp.

[0030] According to the number of scales on the surface of mica pulp, the second concentration range of mica pulp is determined. Specifically, the number of scales on the surface of mica pulp is retrieved in a preset database to obtain the mica pulp concentration corresponding to the number of scales on the surface of mica pulp, and the retrieved concentration of mica pulp is used as the second concentration range. In this embodiment, the second concentration range is the concentration range of mica pulp obtained by retrieving the concentration of mica pulp corresponding to the number of scales on the surface of mica pulp in a database. For example, it is determined that the number of scales on the surface of mica pulp is 530, and the mica pulp concentration corresponding to the number of scales on the surface of mica pulp 530 in the preset database is 3%~5%, so the mica pulp concentration is determined to be 3%~5%.

[0031] The final concentration range of the mica pulp is determined by the first concentration range and the second concentration range of the mica pulp, and a quality control strategy is generated according to the final concentration range, that is, when the second concentration range of the mica pulp is included in the first concentration range, the second concentration range is the final concentration range of the mica pulp. When the first concentration range and the second concentration range of the mica pulp overlap, the overlapping range of the first concentration range and the second concentration range is used as the final concentration range of the mica pulp. When the first concentration range and the second concentration range of the mica pulp are adjacent, the range obtained by adding the first concentration range and the second concentration range is used as the final concentration range of the mica pulp, and the concentration of the mica pulp is alarmed, and the staff is notified to re-test the first concentration range and the second concentration range of the mica pulp. Generating a quality control strategy according to the final concentration range specifically means that when the final concentration range of the mica pulp is high, it indicates that the content of solid matter in the mica pulp is greater than the content of liquid in the mica pulp. At this time, it is necessary to add liquid to the mica pulp by generating a quality control strategy to reduce the final concentration range of the mica pulp. Similarly, when the final concentration range of mica pulp is low, it indicates that the content of solid matter in the mica pulp is less than the content of liquid in the mica pulp. At this time, solid matter needs to be added to the mica pulp according to the generated quality control strategy to increase the final concentration range of the mica pulp.

[0032] Through the above technical solution, by using the vibrator at the bottom of the mica pulp pool to vibrate the mica pulp, the target area on the surface of the mica pulp is obtained, and based on the image recognition method, the vibration ripple characteristics of the target area image are determined, and the relationship between the mica pulp concentration and the vibration ripple characteristics can be obtained, so that the quality detection and control of the mica pulp concentration is more accurate. According to the vibration ripple characteristics, the first concentration range of the mica pulp is determined. By obtaining the first concentration range of the mica pulp, it is helpful to make the concentration analysis of the mica pulp more accurate, so that the quality control effect of the mica pulp is better. When the first concentration range is not within the preset concentration range, an early warning instruction is issued, which can warn the staff in advance and help to timely discover the potential risks of the mica pulp. By obtaining the number of scales on the surface of the mica pulp, the second concentration range of the mica pulp is determined, and the final concentration range of the mica pulp is determined. According to the final concentration range, a quality control strategy is generated, and the concentration of the mica pulp can be regulated, effectively ensuring the accuracy of the concentration of the mica pulp in the production process.

[0033] In one implementation of this embodiment, image recognition is performed on the target area image to obtain the vibration ripple features of the target area, including: S210, dividing the target area image into a plurality of unit grid images, and preprocessing each unit grid image to obtain a corresponding grayscale unit grid image; S220, performing image recognition on each grayscale unit grid image to obtain the number of vibration ripples and the amplitude of the vibration ripples.

[0034] The target area image is divided into a plurality of unit grid images, and each unit grid image is preprocessed to obtain a corresponding grayscale unit grid image. That is to say, the target area image of the mica pulp surface is divided into a plurality of unit grid images. In this embodiment, the target area image refers to an area image of the mica pulp surface without contamination area. The image without contamination area is divided to obtain a plurality of unit grid images, and each unit grid image is preprocessed to obtain a plurality of grayscale unit grid images.

[0035] Image recognition refers to the use of computers and artificial intelligence technology to analyze and understand images to automatically identify and classify objects, scenes, patterns and other information in the image. Each grayscale unit grid image can be image recognized through a preset image recognition algorithm to obtain the number and amplitude of vibration ripples in each grayscale unit grid image.

[0036] By dividing the target area image into multiple unit grid images and preprocessing each unit grid image to obtain the corresponding grayscale unit grid image, the target area image can be made clearer. By identifying each grayscale unit grid image through image recognition, the number of vibration ripples and the vibration ripple amplitude can be obtained, which can make the analysis of the target area image more accurate and ensure the accuracy of the obtained vibration ripple number and vibration ripple amplitude data.

[0037] In one implementation manner of this embodiment, after preprocessing each unit grid image to obtain a corresponding grayscale unit grid image, the method further includes: S310, for any grayscale unit grid image, performing denoising processing on the grayscale unit grid image; S320, extracting features from the grayscale unit grid image after denoising to obtain target impurity edge features; S330, obtaining a first pixel point of a target impurity edge feature and a second pixel point of a mica flake edge feature; S340, calculating the Euclidean distance and color contrast between the first pixel and the second pixel, and determining whether the target area is a polluted area according to the Euclidean distance and color contrast between the first pixel and the second pixel; S350: If the Euclidean distance and the color contrast are both greater than a preset threshold, the target area is determined to be a polluted area, and an alarm message is issued; S360: If the Euclidean distance and the color contrast are both less than or equal to the preset threshold, it is determined that the target area is not a polluted area, and image recognition is performed on each grayscale unit grid image to obtain the number of vibration ripples and the amplitude of the vibration ripples.

[0038] Figure 2 A schematic diagram of a process for detecting a contaminated area in a target area according to an embodiment of the present application is shown schematically. Figure 2 As shown, after issuing the first control instruction to make the vibrator work, the target area on the surface of the mica pulp is obtained and it is determined whether there is a contaminated area in the target area. If there is a contaminated area in the target area, the contaminated area will be identified and a warning message will be issued. After removing the contaminated area and confirming that there is no contaminated area in the target area, the instruction to identify the vibration ripples is executed. If there is no contaminated area in the target area, the instruction to identify the vibration ripples is directly executed.

[0039] For any grayscale unit grid image, the grayscale unit grid image is denoised. In this embodiment, the grayscale unit grid image is denoised mainly through a pre-built deep convolutional neural network model, and the pre-built model is trained with a large amount of image data to achieve efficient image denoising.

[0040] After the grayscale unit grid image is denoised, the feature extraction of the denoised grayscale unit grid image is performed to obtain the target impurity edge feature. In this embodiment, the target impurity edge feature can be the edge feature of suspended matter in the water during the production process, such as the edge feature of plastic, wood chips, bark, weeds and other materials. The image gradient of the target impurity is obtained through image recognition technology. The image gradient refers to the vector obtained by treating the image as a two-dimensional discrete function and calculating the rate of change of each pixel point in the image in the X-axis and Y-axis directions. This vector is used to represent the edge information of the image at that point and is often used in edge detection, feature extraction and image enhancement tasks. The target impurity edge feature is determined by the image gradient of the target impurity.

[0041] The first pixel point of the edge feature of the target impurity and the second pixel point of the edge feature of the mica flake are obtained. In this embodiment, the first pixel point refers to the pixel point of the edge feature of the impurity on the surface of the mica pulp, and the second pixel point is the pixel point of the edge feature of the mica flake on the surface of the mica pulp. The first pixel point of the edge feature of the target impurity and the second pixel point of the edge feature of the mica flake can be obtained by image recognition technology.

[0042] After obtaining the first pixel point of the target impurity edge feature and the second pixel point of the mica scale edge feature, the Euclidean distance and color contrast between the first pixel point and the second pixel point are calculated, wherein the Euclidean distance is used to calculate the true distance between two points in the m-dimensional space. In this embodiment, the Euclidean distance between the first pixel point and the second pixel point can be calculated by the coordinates of the first pixel point and the second pixel point. In the two-dimensional space, the calculation formula of the Euclidean distance is: ‌ Among them, x1 and y1 represent the horizontal coordinate and vertical coordinate of the first pixel point in the two-dimensional space, and x2 and y2 represent the horizontal coordinate and vertical coordinate of the second pixel point in the two-dimensional space.

[0043] Calculating color contrast refers to measuring the different brightness levels between the brightest white and the darkest black in the light and dark areas of an image. In this embodiment, calculating the color contrast between the first pixel and the second pixel refers to calculating the color contrast between the first pixel of the target impurity edge feature and the second pixel of the mica scale edge feature. For example, the color of the first pixel of the target impurity edge feature is black, and the second pixel of the mica scale edge feature is white. By calculating the color contrast between the first pixel and the second pixel, the mica scale and the target impurities can be effectively distinguished, which facilitates the accurate detection of the quality of the mica pulp concentration.

[0044] According to the Euclidean distance and color contrast between the first pixel and the second pixel, determine whether the target area is a contaminated area. In this embodiment, the contaminated area refers to the area where impurities in the mica pulp are located. That is to say, by calculating the Euclidean distance and color contrast between the first pixel and the second pixel, the obtained Euclidean distance and color contrast are compared with the preset threshold. If both the Euclidean distance and the color contrast are greater than the preset threshold, the target area is determined to be a contaminated area. The preset threshold can be determined according to production requirements. For example, the Euclidean distance between the first pixel and the second pixel is calculated to be 5, and the preset Euclidean distance threshold is 10, indicating that the Euclidean distance between the first pixel and the second pixel is too large, and there are impurities in the current image. The color contrast between the first pixel and the second pixel is calculated to be 120:1, and the preset color contrast threshold is 50:1, indicating that the current color contrast between the first pixel and the second pixel exceeds the preset threshold. When both the Euclidean distance and the color contrast are greater than the preset threshold, the target area is determined to be a contaminated area. ‌ After confirming the existence of the contaminated area, an alarm message is issued, that is, when the mica pulp has a contaminated area, an early warning message is issued to the staff, and the mobile phone number reserved by the staff is used as the first emergency contact through a mobile phone text message to issue the corresponding early warning message. After confirming that there is no contaminated area in the target area, the image recognition is performed on each grayscale unit grid image to obtain the number of vibration ripples and the amplitude of vibration ripples.

[0045] By detecting whether there is a contaminated area in the target area, the quality detection of the mica pulp concentration can be further monitored. When there is no contaminated area in the target area of ​​the mica pulp, the step of performing image recognition on the grayscale unit grid image to obtain the number and amplitude of vibration ripples can make the number and amplitude of vibration ripples more accurate, thereby ensuring the accuracy of mica pulp concentration detection.

[0046] In one implementation of this embodiment, determining the first concentration range of mica pulp according to the vibration ripple characteristics of the target area includes: S410, obtaining the number of vibration ripples and the amplitude of vibration ripples according to the vibration ripple characteristics of the target area; S420, comparing the number of vibration ripples with the number of vibration ripples of a preset concentration to obtain a first concentration of mica pulp; S430, comparing the vibration ripple amplitude with the vibration ripple amplitude of a preset concentration to obtain a second concentration of the mica pulp; S440, determining a first concentration range of the mica pulp based on the first concentration of the mica pulp and the second concentration of the mica pulp.

[0047] After obtaining the vibration ripple characteristics of the target area, the number of vibration ripples and the amplitude of the vibration ripples are determined. Specifically, the vibration ripple characteristics are obtained through image recognition technology, and the number of vibration ripples and the amplitude of vibration ripples on the surface of the mica pulp are determined. The number of vibration ripples is compared with the number of vibration ripples of the preset concentration to obtain the first concentration of mica pulp. The concentration of mica pulp refers to the concentration of solid matter in mica pulp in water during pulping. The division of mica pulp concentration is a relative concept. The proportion of mica and other fibers contained in every 100 ml of pulp, for example, the concentration of mica pulp is 2%, which means that in every 100 ml of pulp, the proportion of mica powder and other papermaking fibers is 2%. In this embodiment, the preset concentration can be determined according to actual conditions. The number of vibration ripples of the preset concentration is the number of vibration ripples corresponding to each preset concentration stored in the database, for example, the number of vibration ripples corresponding to the preset concentration of 1% in the database is 20, and the number of vibration ripples corresponding to the preset concentration of 2% is 35. The number of vibration ripples is compared with the number of vibration ripples of the preset concentration to obtain the first concentration of mica pulp. For example, the number of vibration ripples in the target area is 35, and the preset concentration range corresponding to the number of vibration ripples of 35 in the database is 3%. Therefore, the first concentration of mica pulp is determined to be 3%.

[0048] The vibration ripple amplitude is compared with the vibration ripple amplitude of the preset concentration to obtain the second concentration of the mica pulp. Specifically, the obtained vibration ripple amplitude is compared with the vibration ripple amplitude of the preset concentration in the database. In this embodiment, the preset concentration can be determined according to the actual situation. The vibration ripple amplitude of the preset concentration is the vibration ripple amplitude corresponding to each preset concentration stored in the database. For example, the vibration ripple amplitude corresponding to the preset concentration of 1% in the database is 10, and the vibration ripple amplitude corresponding to the preset concentration of 2% is 30. The number of vibration ripples is compared with the number of vibration ripples of the preset concentration to obtain the first concentration of the mica pulp. For example, the vibration ripple amplitude of the target area is 10 cm, and the concentration of the preset concentration range of 1% in the database corresponds to a vibration ripple amplitude of 10 cm. Therefore, the second concentration of the mica pulp is determined to be 1%.

[0049] The first concentration range of the mica pulp is determined by combining the first concentration of the mica pulp and the second concentration of the mica pulp. In this embodiment, the first concentration range is the mica pulp concentration range determined by the first concentration and the second concentration of the mica pulp. The first concentration range of the mica pulp is determined by combining the first concentration and the second concentration of the mica pulp. For example, the first concentration of the mica pulp is 3%, the second concentration of the mica pulp is 1%, and the first concentration range of the mica pulp is determined to be 1% to 3%.

[0050] By obtaining the vibration ripple characteristics of the target area, comparing the number of vibration ripples with the number of vibration ripples in a preset concentration range, the first concentration of mica pulp is obtained, and the vibration ripple amplitude is compared with the vibration ripple amplitude in a preset concentration range to obtain the second concentration of mica pulp. By combining the first concentration of mica pulp and the second concentration of mica pulp, the first concentration range of mica pulp is determined, and the concentration of mica pulp can be obtained more accurately, so that the quality detection effect of mica pulp is better and the production quality of mica paper is guaranteed.

[0051] In one implementation of this embodiment, if the first concentration range is not within the preset concentration range, issuing a warning instruction of the mica pulp concentration includes: S510, comparing the first concentration range with the preset concentration range to determine the warning level of the first concentration range; S520. Based on the warning level, a warning instruction for mica pulp concentration is issued.

[0052] When it is determined that the first concentration range is not within the preset concentration range, an early warning instruction of the mica pulp concentration is issued. The preset concentration range can be determined according to the actual situation. Specifically, the first concentration range is compared with the preset concentration range to determine the early warning level of the first concentration range. That is to say, different early warning levels are generated for mica pulp according to different preset concentration ranges. In this embodiment, the early warning level can be divided into primary early warning level, intermediate early warning level and advanced early warning level, and different early warning levels correspond to different early warning concentration ranges. For example, when the preset concentration range is exceeded by ±3 times, it is the primary early warning level. The primary early warning level can be used to notify the staff at the production site by broadcasting that the current first concentration range is not within the preset concentration range and exceeds the preset concentration range by ±3 times. When the preset concentration range is exceeded by ±6 times, it is the intermediate early warning level. The intermediate early warning level can be used to notify the staff with the reserved mobile phone number and the staff at the production site by calling and broadcasting, and the staff is reminded by calling the reserved mobile phone number of the staff that the current first concentration range of mica pulp is not within the preset concentration range and exceeds the preset concentration range by ±6 times. When the concentration exceeds the preset range by ±9 times, it is a high-level warning level. The high-level warning level can be notified to the staff through broadcasting, calling the staff with the reserved mobile phone number, or sending text messages that the current mica pulp first concentration range is not within the preset concentration range and exceeds the preset concentration range by ±9 times.

[0053] Based on the warning level, a warning instruction of mica pulp concentration is issued, that is, by judging that the first concentration range of mica pulp is within the corresponding warning level, different warning level notifications are issued in different ways. For example, the first concentration range of mica pulp is 1%~3%, the preset concentration range is 5%~7%, the first concentration range is not within the preset concentration range, the average value of the first concentration range is 2%, the average value of the preset concentration range is 6%, and the preset concentration range is 3 times the first concentration range. The warning level can be a primary warning level, an intermediate warning level, and a senior warning level. The preset concentration range is 3 times the first concentration range as the primary warning level. Therefore, a primary warning instruction of mica pulp concentration is issued, and the instruction of issuing the primary warning level can be notified to the staff at the production site by broadcasting to give a warning prompt.

[0054] By comparing the first concentration range with the preset concentration range, determining the warning level of the first concentration range, and issuing a warning instruction for the mica pulp concentration, the risks of mica pulp quality control can be discovered in time, the inaccurate concentration of mica pulp can be prevented, and relevant staff can be notified in time to deal with the concentration of mica pulp, thereby ensuring the accuracy of the concentration of mica pulp during the production process.

[0055] In one implementation of this embodiment, determining the number of scales on the surface of mica pulp through the current image includes: S610, grayscale processing is performed on the current image to obtain a grayscale image, wherein the current image includes a plurality of reflected light points obtained based on incident light; S620, obtaining a third pixel point of the grayscale image and a fourth pixel point of the background image of the mica pulp surface; obtaining pixel points of the current image, and identifying color features of all pixel points of the current image; S630, determining the pixel point of the reflected light point according to the color feature; S640. According to the pixel points of the reflected light spots, a preset edge detection algorithm is used to determine the edges of the reflected light spots, and according to the edges of the reflected light spots, the number of reflected light spots is obtained, wherein the number of reflected light spots is used to characterize the number of scales on the surface of the mica pulp.

[0056] The number of scales on the surface of mica pulp is determined by the current mica pulp surface image. Specifically, first, the current image is gray-processed to obtain a gray-scale image. In this embodiment, the gray-scale image is a gray-scale image of the surface of mica pulp. The current image is gray-processed, and the image can be processed by the average method. The average method is to perform arithmetic averaging on the values ​​of the pixels in the original image, and the average result is the gray value of the pixel. The gray-scale image includes a plurality of reflected light points based on the incident light. That is, the gray-scale image contains a plurality of reflected light points. The reflected light points are incident light emitted by the light source to the target area, and the incident light is reflected by the mica scales on the surface of the mica pulp. By gray-processing the current image to obtain a gray-scale image, it is convenient to analyze the mica pulp surface image more clearly.

[0057] The third pixel of the grayscale image and the fourth pixel of the background image of the mica pulp surface are obtained. In this embodiment, the third pixel of the grayscale image refers to the pixel of the grayscale image of the mica scale on the mica pulp surface, and the fourth pixel refers to the pixel in the background image of the mica pulp surface. The third pixel of the grayscale image and the fourth pixel of the background image of the mica pulp surface are obtained, and the color features of all the pixels of the current image are identified according to the pixels of the current image. That is, the pixels of the current image are analyzed by the color features of the third and fourth pixels, and the color features of all the pixels of the current image are determined. For example, the color feature of the third pixel of the grayscale image is obtained to be black, and the color feature of the fourth pixel of the background image of the mica pulp surface is white. The color features of the pixels of the current image are identified by using image recognition technology. Image recognition refers to the use of computers and artificial intelligence technology to analyze and understand images to automatically identify and classify objects, scenes, patterns and other information in the image.

[0058] The color features obtained from the current image are compared with the color features of the third pixel point of the grayscale image and the color features of the fourth pixel point of the background image of the mica pulp surface, and the pixel points of the reflected light points in the current image are determined based on the color features. For example, the color features of the reflected light point pixels in the current image are black, and the color features of the fourth pixel point of the background image of the mica pulp surface are white. By determining all areas in the current image where the color features are black, these areas are used as the locations of the reflected light points.

[0059] According to the pixel points of the reflected light spot, a preset edge detection algorithm is used to determine the edge of the reflected light spot. Specifically, the position of the reflected light spot can be determined by color features, and a preset edge detection algorithm is used. In this embodiment, the preset edge detection algorithm can be a Sobel operator, and the Sobel operator is mainly used for edge detection. The edge of the reflected light spot is determined by the Sobel operator. For example, if the edge of the reflected light spot is an irregular elliptical edge shape, the position of the edge with the irregular elliptical edge shape is determined in the current image, thereby determining the position of the reflected light spot in the current image.

[0060] According to the edges of the reflected light spots, the number of reflected light spots is obtained, wherein the number of reflected light spots is used to characterize the number of scales on the surface of the mica pulp. That is, by determining the characteristics of the edges of the reflected light spots, the positions of the reflected light spots are determined in the current image according to the characteristics of the edges of the reflected light spots, and the number of positions of the reflected light spots is counted, thereby determining the number of reflected light spots.

[0061] By identifying the color features in the current image and using image recognition technology to obtain the number of reflected light spots on the surface of mica pulp, the number of mica flakes on the surface of mica pulp can be accurately obtained, which helps to determine the accuracy of the concentration of mica pulp through the number of mica flakes and ensure the accuracy of mica pulp quality control.

[0062] In one implementation of this embodiment, determining the second concentration range of the mica pulp according to the number of scales on the surface of the mica pulp includes: S710, searching for the number of scales on the surface of mica pulp in a preset database, and obtaining a second concentration range of mica pulp corresponding to the number of scales on the surface of mica pulp.

[0063] The second concentration range of the mica pulp is determined by the number of scales on the surface of the mica pulp obtained previously. The concentration of the mica pulp refers to the concentration of the solid matter of the mica pulp in the water during pulping. The concentration of the mica pulp is a relative concept, which is the ratio of mica to other fibers contained in every 100 ml of pulp. For example, the concentration of the mica pulp is 2%, which means that in every 100 ml of pulp, the proportion of mica powder and other papermaking fibers is 2%. In this embodiment, the second concentration range is the concentration range of the mica pulp corresponding to the number of scales on the surface of the mica pulp retrieved in the preset database. For example, the number of scales on the surface of the mica pulp determined by the image recognition technology is 200, and the number of scales on the surface of the mica pulp retrieved in the database is 200, and the concentration of the mica pulp corresponding to the number of scales of 200 is 4%~5%. Therefore, it can be determined that when the number of mica scales on the surface of the mica pulp is 200, the concentration of the mica pulp is 4%~5%. The concentration of mica pulp can be determined by the number of mica flakes on the surface of the mica pulp. The concentration of the liquid can be determined by utilizing the Fresnel reflection law. The Fresnel reflection law determines the concentration of the liquid by measuring the reflection intensity of light. In this embodiment, the concentration of the mica pulp is determined by the reflectivity of the mica flakes.

[0064] By searching the concentration of mica pulp based on mica flakes in the database, the concentration of mica pulp can be effectively determined based on the number of scales on the surface of the mica pulp. This not only achieves accurate identification of the number of mica flakes and improves the accuracy of identification, but also reduces the inaccurate detection of mica pulp concentration, which helps to improve the safety and reliability of mica pulp quality control.

[0065] In one implementation of this embodiment, the final concentration range of the mica pulp is determined by combining the first concentration range and the second concentration range of the mica pulp, including: S810, when the second concentration range of the mica pulp includes the first concentration range, setting the second concentration range as the final concentration range of the mica pulp; S820, when the first concentration range and the second concentration range of the mica pulp overlap, taking the overlapping range of the first concentration range and the second concentration range as the final concentration range of the mica pulp; S830. When the first concentration range and the second concentration range of the mica pulp are adjacent to each other, a range obtained by adding the first concentration range and the second concentration range is used as a final concentration range of the mica pulp.

[0066] A first concentration range and a second concentration range of mica pulp are obtained, and a final concentration range of mica pulp is determined. In this embodiment, the first concentration range of mica pulp refers to the concentration range of mica pulp determined by the first concentration and the second concentration range of mica pulp, and the second concentration range is the concentration range of mica pulp obtained by searching a database for the concentration of mica pulp corresponding to the number of scales on the surface of mica pulp. The final concentration range refers to the range of mica pulp concentration finally determined by the first concentration range and the second concentration range of mica pulp.

[0067] In combination with the first concentration range and the second concentration range of the mica pulp, the final concentration range of the mica pulp is determined. Specifically, when the second concentration range of the mica pulp is included in the first concentration range, the second concentration range is the final concentration range of the mica pulp. That is to say, when the first concentration range of the mica pulp is greater than the second concentration range of the mica pulp, the second concentration range is the final concentration range of the mica pulp. For example, when the first concentration range of the mica pulp is [1%, 5%] and the second concentration range of the mica pulp is [2%, 3%], the second concentration range of the mica pulp is within the first concentration range of the mica pulp. Therefore, the second concentration range of the mica pulp is [2%, 3%] as the final concentration range of the mica pulp, and the concentration range of the mica pulp is narrowed, so that the concentration of the mica pulp is more accurate.

[0068] When the first concentration range and the second concentration range of mica pulp overlap, the overlapping range of the first concentration range and the second concentration range is used as the final concentration range of the mica pulp. Specifically, when the first concentration range and the second concentration range of mica pulp overlap, the overlapping part of the first concentration range and the second concentration range is used as the final concentration range of the mica pulp, so that the concentration range of the mica pulp is reduced and the concentration of the mica pulp is more accurate. For example, the first concentration range of mica pulp is [1%, 5%], the second concentration range of mica pulp is [2%, 7%], and the overlapping concentration range of the first concentration range and the second concentration range of mica pulp is [2%, 5%]. Therefore, the overlapping concentration range of the mica pulp is [2%, 5%], which is used as the final concentration range of the mica pulp.

[0069] When the first concentration range and the second concentration range of mica pulp are adjacent to each other, the range obtained by adding the first concentration range and the second concentration range is used as the final concentration range of mica pulp. That is to say, when the first concentration range of mica pulp is [1%, 5%], the second concentration range of mica pulp is [6%, 7%]. The range obtained by adding the first concentration range and the second concentration range of mica pulp is used as the final concentration range of mica pulp. Therefore, the final concentration range of mica pulp is [1%, 7%]. However, after combining the first concentration range and the second concentration range of mica pulp, the final concentration range of mica pulp becomes larger, resulting in the final concentration range of mica pulp becoming inaccurate. Therefore, when the first concentration range and the second concentration range of mica pulp are adjacent to each other, it is an abnormal situation of the concentration range of mica pulp, indicating that the first concentration range or the second concentration range of mica pulp is inaccurate, resulting in the enlargement of the final concentration range of mica pulp. When the first concentration range and the second concentration range of mica pulp are adjacent to each other, an alarm of concentration detection error is issued, and the staff re-tests the first concentration range and the second concentration range of the mica pulp to determine the latest first concentration range and the second concentration range, and determines the final concentration range of the mica pulp in combination with the latest first concentration range and the second concentration range.

[0070] By combining the first concentration range and the second concentration range of mica pulp to determine the final concentration range of mica pulp, the error in the detection of mica pulp concentration can be reduced, the detection of mica pulp can be made more accurate, the accuracy of mica pulp concentration in the mica paper production process can be guaranteed, and the quality of the produced mica paper can be guaranteed.

[0071] An embodiment of the present application further provides a machine-readable storage medium having instructions stored thereon, the instructions being used to enable a machine to execute the above-mentioned mica pulp quality control method based on image recognition.

[0072] The embodiment of the present application also provides a mica pulp quality control system based on image recognition, comprising: a memory configured to store instructions; and The processor is configured to call instructions from the memory and implement the above-mentioned mica pulp quality control method based on image recognition when executing the instructions.

[0073] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems or computer program products. Therefore, the present application may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0074] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of the processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0075] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0076] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0077] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0078] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0079] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0080] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0081] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included within the scope of the claims of the present application.

Claims

1. A mica pulp quality control method based on image recognition, characterized in that: include: issuing a first control instruction to control the operation of a vibrator disposed at the bottom of a mica pulp pool; Acquire a target area image of mica pulp, perform image recognition on the target area image, and obtain vibration ripple features of the target area image, wherein the vibration ripple features include the number of vibration ripples and the amplitude of vibration ripples; Determining a first concentration range of mica pulp according to the vibration ripple characteristics of the target area image, and judging whether the first concentration range is within a preset concentration range; If the first concentration range is not within the preset concentration range, issuing a warning instruction; If the first concentration range is within a preset concentration range, after the mica pulp is left to stand for a preset period of time, a second control instruction is issued to control the light source to emit incident light to a target area; Acquire a current image of the target area; Determining the number of scales on the surface of mica pulp through the current image; Determining a second concentration range of the mica pulp according to the number of scales on the surface of the mica pulp; The final concentration range of the mica pulp is determined by combining the first concentration range and the second concentration range of the mica pulp to generate a quality control strategy according to the final concentration range.

2. The method according to claim 1, characterized in that The performing image recognition on the target area image to obtain the vibration ripple features of the target area includes: Dividing the target area image into a plurality of unit grid images, and preprocessing each of the unit grid images to obtain a corresponding grayscale unit grid image; Image recognition is performed on each of the grayscale unit grid images to obtain the number of vibration ripples and the amplitude of the vibration ripples.

3. The method according to claim 2, characterized in that After preprocessing each of the unit grid images to obtain a corresponding grayscale unit grid image, the method further includes: For any one of the grayscale unit grid images, performing denoising processing on the grayscale unit grid image; Performing feature extraction on the grayscale unit grid image after denoising to obtain target impurity edge features; Acquire a first pixel point of the target impurity edge feature and a second pixel point of the mica flake edge feature; Calculating the Euclidean distance and color contrast between the first pixel and the second pixel, and determining whether the target area is a polluted area according to the Euclidean distance and color contrast between the first pixel and the second pixel; If the Euclidean distance and the color contrast are both greater than a preset threshold, the target area is determined to be a polluted area, and an alarm message is issued; If the Euclidean distance and the color contrast are both less than or equal to a preset threshold, it is determined that the target area is not a polluted area, and the step of performing image recognition on each of the grayscale unit grid images to obtain the number of vibration ripples and the amplitude of the vibration ripples is performed.

4. The method according to claim 1, characterized in that: Determining the first concentration range of mica pulp according to the vibration ripple characteristics of the target area includes: Obtaining the number of vibration ripples and the amplitude of vibration ripples according to the vibration ripple characteristics of the target area; Comparing the number of vibration ripples with the number of vibration ripples of a preset concentration to obtain a first concentration of mica pulp; Comparing the vibration ripple amplitude with the vibration ripple amplitude of a preset concentration to obtain a second concentration of the mica pulp; The first concentration range of the mica pulp is determined in combination with the first concentration of the mica pulp and the second concentration of the mica pulp.

5. The method according to claim 1, characterized in that If the first concentration range is not within the preset concentration range, issuing a warning instruction of the mica pulp concentration includes: Comparing the first concentration range with a preset concentration range to determine a warning level of the first concentration range; Based on the warning level, a warning instruction for mica pulp concentration is issued.

6. The method according to claim 1, characterized in that Determining the number of scales on the surface of mica pulp through the current image includes: Performing grayscale processing on the current image to obtain a grayscale image, wherein the current image includes a plurality of reflected light points obtained based on the incident light; Acquire a third pixel point of the grayscale image and a fourth pixel point of the background image of the mica pulp surface; acquire pixel points of the current image, and identify color features of all the pixel points of the current image; Determine the pixel point of the reflected light point according to the color feature; According to the pixel points of the reflected light spots, a preset edge detection algorithm is used to determine the edges of the reflected light spots, and according to the edges of the reflected light spots, the number of the reflected light spots is obtained, wherein the number of the reflected light spots is used to characterize the number of scales on the surface of the mica pulp.

7. The method according to claim 6, characterized in that Determining the second concentration range of the mica pulp according to the number of scales on the surface of the mica pulp comprises: The number of scales on the surface of the mica pulp is retrieved in a preset database to obtain a second concentration range of the mica pulp corresponding to the number of scales on the surface of the mica pulp.

8. The method according to claim 7, characterized in that The step of combining the first concentration range and the second concentration range of the mica pulp to determine a final concentration range of the mica pulp comprises: When the second concentration range of the mica pulp includes the first concentration range, the second concentration range is used as the final concentration range of the mica pulp; When the first concentration range and the second concentration range of the mica pulp overlap, the overlapping range of the first concentration range and the second concentration range is used as the final concentration range of the mica pulp; When the first concentration range and the second concentration range of the mica pulp are adjacent to each other, a range obtained by adding the first concentration range and the second concentration range is used as the final concentration range of the mica pulp.

9. A machine-readable storage medium, characterized in that: The machine-readable storage medium stores instructions for causing a machine to execute the mica pulp quality control method based on image recognition according to any one of claims 1 to 8.

10. A mica pulp quality control system based on image recognition, characterized in that: include: a memory configured to store instructions; as well as A processor is configured to call the instructions from the memory and implement the mica pulp quality control method based on image recognition according to any one of claims 1 to 8 when executing the instructions.

Images