A pellet size detection system and detection method based on image processing

Abstract

The present invention designs a pellet size detection system and detection method based on image processing. The pellet size detection system includes a feeding device, a test container device, and a detection device. When in use, the total weight m0 of the pellet sample to be tested is first weighed and fed into the pellet size detection system from the feed port. After preliminary particle size screening is performed by a screening and transmission mechanism, the pellet sample remaining on the screen falls into the test platform. After the gas purge device is purged and the total weight m1 of the pellet sample contained in the test platform is collected by the weighing sensor, an industrial camera is used to capture an image of the containing area of ​​the test platform and transmit it to a detection computer for image analysis to obtain the pellet size detection result. The present invention introduces a new pellet size analysis method based on image detection and analysis. It adopts a method of sample discretization and segmentation circle interpolation closure in image recognition to perform pellet size detection. The operation method is simple, labor-saving, and can more quickly, effectively, and accurately perform pellet size detection on pellet ore.

Description

Technical Field

[0001] The present invention relates to image processing technology and the technical field of ironmaking pellet intelligent detection, and in particular to a pellet particle size detection system and detection method based on image processing. Background Art

[0002] Particle size is a key parameter affecting pellet quality and is crucial to the calcination process. If the particle size is too large, the oxidation calcination process will be poor, and "green nuclei" will easily form. If the particle size is too small, the pellets will have a high bulk density and poor air permeability, which will also affect the calcination process. The industry generally accepts a reasonable particle size range of 8mm-16mm. Therefore, accurately measuring pellet size and its distribution is crucial for pellet production.

[0003] Commonly used modern particle size detection methods include sedimentation method, screening method, electric induction method, microscopy method, etc. According to different detection principles, the obtained particle size parameters include average diameter, equivalent volume diameter, equivalent area diameter, equivalent diameter, etc. The expression of particle size distribution is also different, such as the number of particles, percentage, mass percentage, etc. At present, the more popular pellet size detection method in steel mills is mechanical screening method. Usually, the operators of the steel mills use tools to obtain a small amount of pellet particles at the outlet of the disc pelletizing machine, and use a special pellet sieve to screen the sample. Finally, the particle size distribution of the pellets is obtained by manual screening and weighing. There are some problems with this method, including inaccurate particle size distribution data, the need for manual calculation of particle size distribution, and delayed results.

[0004] In order to make up for the shortcomings of manual screening methods, more and more studies have begun to use image processing technology. Bai Zhicheng et al. used the collected pellet particle images and used pre-processing schemes such as filtering, thresholding, enhancement, and optimization to solve the problems of high image noise and dark colors. They then proposed an optimized watershed segmentation algorithm based on adaptive labeling, which realized the image segmentation of pellets and obtained the statistical results of the particle size distribution of pellets, providing convenience for pellet production.

[0005] Although the above method can realize contactless pellet particle size detection, it lacks an effective solution for images with low signal-to-noise ratio, low contrast, uneven particle size and severe accumulation, and cannot effectively segment them correctly. In addition, the "center point + long axis" method is also used to calibrate the particle size during the image recognition process, which is insufficient in the recognition and differentiation of overlapping images.

[0006] Therefore, how to detect the pellet size of pellets more quickly, effectively and accurately has become a research topic. Summary of the Invention

[0007] In view of the above-mentioned deficiencies in the prior art, the present invention provides a pellet size detection system and method based on image processing, which has a simple operation method, saves manpower, and can detect the pellet size of pellet ore more quickly, effectively and accurately.

[0008] In order to solve the above technical problems, the present invention adopts the following technical solutions:

[0009] In the first aspect, the present invention provides a pellet size detection system based on image processing, such as Figure 1 As shown, it includes a feeding device, a test container device and a detection device;

[0010] The feeding device includes a material guide port 2 and a screening and transmission mechanism 3 for performing preliminary particle size screening on the pellet sample. The material guide port 2 is connected to the test container device through the screening and transmission mechanism 3.

[0011] The test container device includes a test platform 4, which is a platform-shaped container with a horizontal bottom plate and is used to hold the pellet sample introduced through the material guide port 2 and the screening and conveying mechanism 3. A gas purge device 6 is provided on the outer sides of the test platform 4 for blowing gas into the test platform 4. A weighing sensor 5 is installed at the bottom of the test platform 4 for weighing the total weight of the pellet sample held in the test platform 4.

[0012] The detection device includes a detection computer 8 and an industrial camera 7 mounted directly above the test platform 4; the industrial camera 7 is used to collect images of the storage area of ​​the test platform 4; the data receiving end of the detection computer 8 is connected to the image data output end of the industrial camera 7 and the weight data output end of the weighing sensor 5 for data transmission, and the detection computer 8 is used to perform image analysis on the storage area image collected by the industrial camera 7, and combine it with the total weight of the pellet sample collected by the weighing sensor 5 to obtain the pellet particle size detection result of the pellet sample contained in the test platform 4.

[0013] In the above-mentioned pellet particle size detection system based on image processing, as a preferred solution, the screening transmission mechanism 3 includes a screening conveyor belt formed by a plurality of transversely arranged screening rollers arranged in an inclined manner, and a conveyor cover with enclosures arranged on both sides and above the screening conveyor belt. The interval distance between two adjacent screening rollers in the screening conveyor belt is 3mm-5.5mm.

[0014] In the above-mentioned pellet size detection system based on image processing, as a preferred solution, the color of the horizontal bottom plate of the test platform 4 has a color difference with the color of the pellet sample.

[0015] In the above-mentioned pellet particle size detection system based on image processing, as a preferred solution, the horizontal bottom plate of the test platform 4 is surrounded by side walls for preventing the pellet samples from falling, and a plurality of ventilation gaps are distributed in a grid shape on the side walls, and the gap spacing is 3mm-5.5mm. The gas purge device 6 can blow gas into the test platform 4 through the ventilation gaps on the side walls.

[0016] In the pellet size detection system based on image processing, as a preferred solution, the detection computer 8 processes the pellet size detection result of the pellet sample placed in the test platform 4 in the following process:

[0017] S1. Obtain the total weight m0 of the pellet sample to be tested, as well as the corresponding relationship between the actual space size and the pixel size corresponding to the holding area image, and save the data;

[0018] S2. Binarizing the collected image of the holding area using a pixel threshold method, wherein the pixel threshold is set between the color pixel value of the pellet sample and the color pixel value of the horizontal base plate, to obtain a binary image;

[0019] S3. Using a watershed segmentation algorithm to perform discretization segmentation processing on the obtained binary image into pellet sample regions, and performing edge closing processing on the pellet sample region edges of the segmented regions to obtain a pellet sample segmentation image containing the segmentation results of each pellet sample region;

[0020] S4. Counting the total number n0 of pellet sample regions contained in the pellet sample segmentation image, and calculating the geometric center pixel position and pixel radius of each pellet sample region respectively, and calculating and determining the particle size of each pellet sample region based on the correspondence between the actual spatial size and the pixel size corresponding to the holding area image, and counting the number n1 of pellet sample regions that meet the particle size requirements;

[0021] S5. Combined with the total weight m1 of the pellet sample placed in the test platform 4 collected by the weighing sensor 5, the qualified pellet ratio β1 and the unqualified pellet ratio β2 are calculated according to the following formulas, and output as the pellet size test results:

[0022]

[0023] In the above-mentioned pellet particle size detection system based on image processing, as a preferred solution, in step S3, the edge closure processing of the edge of the segmented area is specifically carried out as follows: using Lagrange interpolation method, Spline interpolation method, Cubic interpolation method or polynomial interpolation method, the edge of each pellet sample area in the segmented area is subjected to difference processing to form the closed edge segmentation result of each pellet sample area.

[0024] In the pellet size detection system based on image processing, as a preferred solution, step S4 is specifically as follows:

[0025] S401, counting the total number n0 of pellet sample regions contained in the pellet sample segmentation image;

[0026] S402, for any i-th pellet sample region in the pellet sample segmentation image, calculate the pixel area A of the pellet sample region i and the contour perimeter P i , and determine the shape coefficient C of the i-th pellet sample area i =4πA i / P i 2 ; If the shape coefficient C i <0.5, then ignore the pellet sample area; if the shape coefficient C i ≥0.5, then calculate and determine the pixel position coordinates of the geometric center of the i-th pellet sample area. Go to step S403;

[0027] S403: Count the pixel coordinates of the outline of the i-th pellet sample area relative to the geometric center pixel. The pixel distance of the i-th pellet sample area is determined by Mean i , combined with the correspondence between the actual space size and pixel size corresponding to the image of the holding area, the particle size G of the i-th pellet sample area is calculated i :

[0028] G i =2×Mean i ×d max / n max ;

[0029] Among them, d max / n max Indicates the correspondence ratio between the actual space size and pixel size corresponding to the holding area image, d max is the actual horizontal spatial size of the image of the storage area captured by the industrial camera, n max The total number of horizontal pixels of the storage area image captured by the industrial camera;

[0030] S404 , using the processing of steps S402 to S403 , respectively calculate and determine the particle size of each pellet sample area, and then count the number n1 of pellet sample areas that meet the particle size requirements.

[0031] In the pellet size detection system based on image processing, as a preferred solution, in step S402, the geometric center pixel position coordinates of the i-th pellet sample area are calculated and determined. The specific method is:

[0032] Count the pixel coordinates of each contour of the i-th pellet sample area K represents the number of pixels contained in the outline of the i-th pellet sample area, and then the coordinates of the geometric center pixel position of the i-th pellet sample area are calculated as follows:

[0033]

[0034] In the pellet size detection system based on image processing, as a preferred solution, in step S403, the pixel radius Mean of the i-th pellet sample area is determined. i The specific method is:

[0035] First, count the pixel coordinates of the contour of the i-th pellet sample area relative to the geometric center pixel The pixel distance L1, L2, ..., L k ,…,L K , K represents the number of pixels contained in the contour of the i-th pellet sample area, where:

[0036]

[0037] Then, determine the shape coefficient C of the i-th pellet sample area i The value size of ;

[0038] If the shape factor 0.8≤C i ≤1, calculate the pixel radius of the i-th pellet sample area as follows:

[0039]

[0040] Among them, L K is the mean pixel distance of each contour pixel point relative to the geometric center pixel coordinate of the i-th pellet sample area;

[0041] If the shape factor 0.5≤C i <0.8, then the coordinates of the outline pixels of the i-th pellet sample area relative to the geometric center pixel are counted. The pixel distance is greater than or equal to The pixel distance values ​​are marked as L1′, L2′,…, L k ′,…,L′ N , N represents the statistical value greater than or equal to The number of pixel distance values ​​is then calculated as follows: i :

[0042]

[0043] In a second aspect, the present invention further provides a pellet size detection method based on image processing, which uses the above-mentioned pellet size detection system to perform detection, and specifically includes the following steps:

[0044] Step A: First, weigh the total weight m0 of the pellet sample to be tested and input it into the detection computer 8 to save the data; then, feed all the pellet samples to be tested into the pellet size detection system from the material guide port 2. After preliminary particle size screening by the screening and transmission mechanism 3, the remaining pellet samples fall into the test platform 4, and the gas purge device 6 is started to spray gas into the test platform 4 to ensure that the pellet samples contained in the test platform 4 are evenly distributed. Then, the gas purge device 6 is stopped.

[0045] Step B: The total weight m1 of the pellet sample placed on the test platform 4 is collected by the weighing sensor 5 and input into the detection computer 8 for storage. Then, an image of the placement area of ​​the test platform 4 is captured by the industrial camera 7 and transmitted to the detection computer 8.

[0046] In Step C, the detection computer 8 performs image analysis on the image of the storage area captured by the industrial camera 7 and combines it with the total weight m1 of the pellet sample collected by the weighing sensor 5 to obtain the pellet size detection result of the pellet sample stored in the test platform 4.

[0047] Compared with the prior art, the present invention has the following beneficial effects:

[0048] The pellet size detection system and detection method of the present invention overcome the shortcomings of traditional detection methods that damage the performance of raw materials, and circumvent the problem that existing image analysis methods cannot effectively solve the problem of measuring the particle size of overlapping parts of pellets. A new pellet size analysis method based on image detection and analysis is introduced, and pellet size detection is performed by using a method of sample discretization and segmentation circle interpolation closure in image recognition. The operation method is simple, saves manpower, and can perform pellet size detection on pellet ore more quickly, effectively and accurately. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] In order to make the purpose, technical solutions and advantages of the invention more clear, the present invention will be further described in detail below with reference to the accompanying drawings, in which:

[0050] Figure 1 It is a schematic diagram of the framework of the pellet size detection system based on image processing of the present invention.

[0051] Figure 2 It is a schematic diagram of a top view of the test platform in the pellet size detection system of the present invention.

[0052] Figure 3 This is a process diagram of Example 1 of the present invention.

[0053] Figure 4 This is a result diagram of Example 1 of the present invention.

[0054] Figure 5 This is a process diagram of Example 2 of the present invention.

[0055] Figure 6 This is a result diagram of Example 2 of the present invention.

[0056] Figure 7 This is a process diagram of Example 3 of the present invention.

[0057] Figure 8 3 is a schematic diagram of the process of performing edge closing processing on the edge of the pellet sample area No. 587 in Example 3 of the present invention.

[0058] In the figure: 1- pellet sample; 2- material guide port; 3- screening and transmission mechanism; 4- testing platform; 5- weighing sensor; 6- gas purge device; 7- industrial camera; 8- detection computer. DETAILED DESCRIPTION

[0059] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. The components of the embodiments of the present invention generally described and shown in the drawings herein can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the drawings is not intended to limit the scope of the invention claimed for protection, but only represents selected embodiments of the present invention. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0060] In the description of the present invention, it should be understood that the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "up", "down", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", "axial", "radial", "circumferential" and the like to indicate orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as limiting the present invention.

[0061] In the description of the present invention, it should be noted that, unless otherwise specified or limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a direct connection, an indirect connection through an intermediate medium, or it can be internal communication between two components.

[0062] Those skilled in the art can understand the specific meanings of the above terms in the present invention according to specific circumstances.

[0063] In the description of the present invention, it should be noted that the terms used herein are only for the purpose of describing specific embodiments and are not intended to limit the exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is intended to include the plural form. In addition, it should be understood that when the terms "comprise" and / or "include" are used in this specification, they indicate the presence of features, steps, operations, devices, components and / or combinations thereof.

[0064] Unless otherwise specifically stated, the relative arrangement of the parts and steps, numerical expressions and numerical values ​​described in these embodiments do not limit the scope of the present invention. At the same time, it should be clear that, for ease of description, the sizes of the various parts shown in the drawings are not drawn according to the actual proportional relationship. The techniques, methods and equipment known to ordinary technicians in the relevant fields may not be discussed in detail, but where appropriate, the techniques, methods and equipment should be considered as part of the authorization specification. In all examples shown and discussed here, any specific value should be interpreted as merely exemplary and not as a limitation. Therefore, other examples of the exemplary embodiments may have different values. It should be noted that similar numbers and letters represent similar items in the following figures, so once an item is defined in one figure, it does not need to be further discussed in subsequent figures.

[0065] In addition, it should be noted that the use of terms such as "first" and "second" to limit components is only for the convenience of distinguishing the corresponding components. Unless otherwise stated, the above terms have no special meaning and therefore cannot be understood as limiting the scope of protection of the present invention.

[0066] like Figure 1 As shown, the present invention provides a pellet size detection system based on image processing, which is characterized by comprising a feeding device, a test container device and a detection device.

[0067] The feeding device includes a material guide port 2 and a screening and transmission mechanism 3 for performing preliminary particle size screening on the pellet sample. The material guide port 2 is connected to the test container device through the screening and transmission mechanism 3.

[0068] The test container device includes a test platform 4, which is a platform-shaped container with a horizontal bottom plate, used to hold the pellet sample introduced through the material guide port 2 and the screening and transmission mechanism 3; a gas purge device 6 capable of blowing gas into the test platform 4 is provided on the outer side of the test platform 4, and a weighing sensor 5 is installed at the bottom of the test platform 4 for weighing the total weight of the pellet sample contained in the test platform 4.

[0069] The detection device includes a detection computer 8 and an industrial camera 7 mounted directly above the test platform 4; the industrial camera 7 is used to collect images of the storage area of ​​the test platform 4; the data receiving end of the detection computer 8 is connected to the image data output end of the industrial camera 7 and the weight data output end of the weighing sensor 5 for data transmission, and the detection computer 8 is used to perform image analysis on the storage area image collected by the industrial camera 7, and combine it with the total weight of the pellet sample collected by the weighing sensor 5 to obtain the pellet size detection result of the pellet sample placed in the test platform 4.

[0070] The operation flow of pellet size detection using the pellet size detection system of the present invention is as follows:

[0071] Step A: First, weigh the total weight m0 of the pellet sample to be tested and input it into the detection computer 8 to save the data; then, feed all the pellet samples to be tested into the pellet size detection system from the material guide port 2. After preliminary particle size screening by the screening and transmission mechanism 3, the remaining pellet samples fall into the test platform 4, and the gas purge device 6 is started to spray gas into the test platform 4 to ensure that the pellet samples contained in the test platform 4 are evenly distributed. Then, the gas purge device 6 is stopped.

[0072] Step B: The total weight m1 of the pellet sample placed on the test platform 4 is collected by the weighing sensor 5 and input into the detection computer 8 for storage. Then, an image of the placement area of ​​the test platform 4 is captured by the industrial camera 7 and transmitted to the detection computer 8.

[0073] In Step C, the detection computer 8 performs image analysis on the image of the storage area captured by the industrial camera 7 and combines it with the total weight m1 of the pellet sample collected by the weighing sensor 5 to obtain the pellet size detection result of the pellet sample stored in the test platform 4.

[0074] The pellet size detection system of the present invention overcomes the disadvantage of traditional detection methods that they damage the properties of raw materials, and circumvents the problem that existing image analysis methods cannot effectively solve the problem of measuring the particle size of overlapping parts of pellets. It introduces a new pellet size analysis method based on image detection and analysis, and adopts the method of sample discretization and segmentation circle interpolation closure in image recognition to perform pellet size detection. The operation method is simple, saves manpower, and can perform pellet size detection on pellet ore more quickly, effectively and accurately.

[0075] In practice, the screening and conveying mechanism 3 comprises a screening conveyor belt formed by a plurality of horizontally arranged screening rollers arranged in an oblique arrangement, and a conveyor hood arranged on both sides and above the screening conveyor belt. The screening rollers in the screening conveyor belt are arranged obliquely and spaced apart. The spacing between two adjacent screening rollers is the particle size screening size of the pellet sample. For example, the spacing between the screening rollers is designed to be 3mm-5.5mm. This serves to initially screen out some unqualified pellet samples with too small particle size, thereby reducing unnecessary interference with the image detection of pellet size caused by these pellets entering the testing platform 4.

[0076] In a specific embodiment, the horizontal base of the test platform 4 has a color that differs from the color of the pellet sample. For example, if the pellet sample 1 is typically dark gray-brown, black, or other dark colors, the horizontal base of the test platform 4 can be a color with a significant color difference from the pellet sample, such as white, yellow, green, blue, or red. This improves image detection and recognition of the pellet sample and enhances the accuracy of pellet sample area identification. The horizontal base of the test platform 4 is surrounded by side walls to prevent the pellet sample from falling out, ensuring that pellet samples entering the test platform 4 do not fall out of the storage area of ​​the test platform 4. The side walls of the horizontal base are arranged in a grid-like pattern with ventilation slits spaced 3 mm to 5.5 mm apart. The gas purge device 6 can inject gas into the test platform 4 through these ventilation slits. This ensures that the gas purge device 6 injects gas into the test platform 4 with a better purge effect.

[0077] The gas purge device 6 is used to purge the test platform 4 with airflow, so that the pellet samples 1 falling into the test platform 4 are distributed as evenly as possible and overlap is reduced. This is more conducive to accurately identifying the size and particle size of each pellet sample through image processing and reducing the recognition error caused by the overlap of pellet samples. In specific implementation, Figure 2 As shown, the gas purging device 6 can be implemented by using multiple hair dryers distributed around the test platform 4. The gas purging device 6 facing the feeding side of the test platform 4 can be started and opened when the unloading begins, and its gas flow rate is not less than 3m / s. After the pellet sample is unloaded, the gas nozzle of the gas purging device in the feeding side is closed, and then the gas purging devices 6 in other directions are started, and their nozzles are purged in sequence along the circumferential direction of the test platform 4. Alternatively, when there is a special need, the airflow of each gas purging device 6 can be manually controlled to make the pellet sample 1 in the test platform 4 as evenly distributed as possible.

[0078] In a specific implementation, the number of image acquisition pixels of the industrial camera 7 is not less than 2 million to ensure the accuracy of image acquisition.

[0079] In a specific implementation, the detection computer 8 can be implemented by a desktop computer, a mobile terminal computer, or other equipment with data processing capabilities. The process of the detection computer 8 processing the pellet size test results of the pellet sample placed in the test platform 4 is as follows:

[0080] S1. Obtain the total weight m0 of the pellet sample to be tested, as well as the corresponding relationship between the actual space size and the pixel size corresponding to the holding area image, and save the data;

[0081] S2. Binarizing the collected image of the holding area using a pixel threshold method, wherein the pixel threshold is set between the color pixel value of the pellet sample and the color pixel value of the horizontal base plate, to obtain a binary image;

[0082] S3. Using a watershed segmentation algorithm to perform discretization segmentation processing on the obtained binary image into pellet sample regions, and performing edge closing processing on the pellet sample region edges of the segmented regions to obtain a pellet sample segmentation image containing the segmentation results of each pellet sample region;

[0083] S4. Counting the total number n0 of pellet sample regions contained in the pellet sample segmentation image, and calculating the geometric center pixel position and pixel radius of each pellet sample region respectively, and calculating and determining the particle size of each pellet sample region based on the correspondence between the actual spatial size and the pixel size corresponding to the holding area image, and counting the number n1 of pellet sample regions that meet the particle size requirements;

[0084] S5. Combined with the total weight m1 of the pellet sample placed in the test platform 4 collected by the weighing sensor 5, the qualified pellet ratio β1 and the unqualified pellet ratio β2 are calculated according to the following formulas, and output as the pellet size test results:

[0085]

[0086] As can be seen, the pellet size detection process of the present invention, starting from step S2, gradually utilizes image binarization and discretization segmentation to address the difficulty of particle size detection. Furthermore, to address the problem of rough particle size detection after segmentation, the present invention uses edge closure processing to improve the adverse effects caused by the incomplete closure of the edges of the pellet sample areas during the discretization segmentation process. Finally, based on the calculated particle size of each pellet sample area, the number of pellet sample areas that meet the particle size requirements is counted, and finally the proportion of qualified and unqualified pellets is calculated. In a specific implementation, the pellet size range that meets the particle size requirements can be set to 8-16mm.

[0087] In specific implementation, if the horizontal bottom plate color of the test platform 4 is white or light yellow, in step S2, when the collected image of the holding area is binarized using the pixel threshold method, the pixel threshold can be set to 80~120, which can achieve a better binarization effect.

[0088] In a specific implementation, in step S3, during edge closure processing of the segmented regions, Lagrange interpolation, Spline interpolation, Cubic interpolation, or polynomial interpolation can be used to perform interpolation processing on the edges of each pellet sample region in the segmented region to form closed edge segmentation results for each pellet sample region. Using these interpolation methods for edge interpolation processing has a good edge closure enhancement effect.

[0089] In a specific implementation, step S4 is specifically as follows:

[0090] S401, counting the total number n0 of pellet sample regions contained in the pellet sample segmentation image;

[0091] S402, for any i-th pellet sample region in the pellet sample segmentation image, calculate the pixel area A of the pellet sample region i and the contour perimeter P i , and determine the shape coefficient C of the i-th pellet sample area i =4πA i / P i 2 ; If the shape coefficient C i <0.5, then ignore the pellet sample area; if the shape coefficient C i ≥0.5, then calculate and determine the pixel position coordinates of the geometric center of the i-th pellet sample area. Go to step S403;

[0092] S403: Count the pixel coordinates of the outline of the i-th pellet sample area relative to the geometric center pixel. The pixel distance of the i-th pellet sample area is determined by Mean i , combined with the correspondence between the actual space size and pixel size corresponding to the image of the holding area, the particle size G of the i-th pellet sample area is calculated i :

[0093] G i =2×Mean i ×d max / n max ;

[0094] Among them, d max / n maxIndicates the correspondence ratio between the actual space size and pixel size corresponding to the holding area image, d max is the actual horizontal spatial size of the image of the storage area captured by the industrial camera, n max The total number of horizontal pixels of the storage area image captured by the industrial camera;

[0095] S404 , using the processing of steps S402 to S403 , respectively calculate and determine the particle size of each pellet sample area, and then count the number n1 of pellet sample areas that meet the particle size requirements.

[0096] In the above process, the statistical shape coefficient C is used. i In this way, the shape factor C i The pellet sample area with a value <0.5 is considered to be an initially identified shape with an irregular shape or a closed area where most of the area has been blocked. These areas are excluded and only the shape coefficient C is considered to be i The particle size detection and calculation are performed on the pellet sample area with a particle size of ≥0.5. This processing method also reduces the interference of misidentification and helps improve the accuracy of the final pellet particle size detection.

[0097] In the specific implementation, in step S402, the geometric center pixel position coordinates of the i-th pellet sample area are calculated and determined. The specific method is:

[0098] Count the pixel coordinates of each contour of the i-th pellet sample area K represents the number of pixels contained in the outline of the i-th pellet sample area, and then the coordinates of the geometric center pixel position of the i-th pellet sample area are calculated as follows:

[0099]

[0100] In step S403, the pixel radius Mean of the i-th pellet sample area is determined. i The specific method is:

[0101] First, count the pixel coordinates of the contour of the i-th pellet sample area relative to the geometric center pixel The pixel distance L1, L2, ..., L k ,…,L K , K represents the number of pixels contained in the contour of the i-th pellet sample area, where:

[0102]

[0103] Then, determine the shape coefficient C of the i-th pellet sample area i The value size of ;

[0104] If the shape factor 0.8≤C i ≤1, calculate the pixel radius of the i-th pellet sample area as follows:

[0105]

[0106] Among them, L K is the mean pixel distance of each contour pixel point relative to the geometric center pixel coordinate of the i-th pellet sample area;

[0107] If the shape factor 0.5≤C i <0.8, then the coordinates of the outline pixels of the i-th pellet sample area relative to the geometric center pixel are counted. The pixel distance is greater than or equal to The pixel distance values ​​are marked as L1′, L2′,…, L k ′,…,L′ N , N represents the statistical value greater than or equal to The number of pixel distance values ​​is then calculated as follows: i :

[0108]

[0109] In the above process, according to the shape coefficient C i The value of the value is divided into different radius calculation methods, taking into account: if the shape coefficient 0.8≤C i If the shape coefficient is 0.5≤C, the pellet sample area is considered to be a relatively complete and overall circular area, and the radius is directly calculated by the mean of all pixel distances to determine the pellet size. i <0.8, which means that the pellet sample area may be partially blocked and not very complete, or the area shape is elliptical, so the radius is calculated based on the average distance of the pixels with larger distance values ​​to determine the pellet size; this method can more realistically reflect the particle size of each pellet area.

[0110] Below, the present invention is further described by examples.

[0111] Example:

[0112] First, a pellet sample with a total weight of m0 is placed into the pellet size detection system through the material guide port and enters the test platform through the screening transmission mechanism. Then, after being weighed by the weighing sensor at the bottom of the test platform and purged by the gas purge device around it (gas flow rate 5-8m / s), the industrial camera above it (2 million pixels) is used to collect images of the storage area of ​​the test platform. Finally, the collected images are processed and analyzed by the detection computer to obtain the pellet size test results, giving the proportion of qualified pellets and the proportion of unqualified pellets.

[0113] [Example 1]

[0114] In this embodiment, the total weight of the pellet sample of dispersed sample 1 is m0 = 2000 g, the total weight of the pellet sample contained in the test platform collected by the weighing sensor is m1 = 1985 g, the pixel threshold in the image analysis is set to 100, the total number of identified pellet sample areas n0 = 42, and after detection, the number of pellet sample areas that meet the particle size requirements is n1 = 32.

[0115] The image detection process of this embodiment is as follows Figure 3 As shown in the figure, the pellet size test results are as follows Figure 4 As shown, the output is:

[0116]

[0117] [Example 2]

[0118] In this embodiment, the total mass of the pellets of dispersed sample 2 is m o =2000g, test sample mass m1 = 1993g, the threshold value in image analysis is set to 120;

[0119] In this embodiment, the total weight of the pellet sample of dispersed sample 2 is m0 = 2000 g, the total weight of the pellet sample contained in the test platform collected by the weighing sensor is m1 = 1993 g, the pixel threshold in the image analysis is set to 100, the total number of identified pellet sample areas n0 = 24, and after detection, the number of pellet sample areas that meet the particle size requirements is n1 = 22.

[0120] The image detection process of this embodiment is as follows Figure 5 As shown in the figure, the pellet size test results are as follows Figure 6 As shown, the output is:

[0121]

[0122] [Example 3]

[0123] (Overlapping sample) Total mass of pellets m o=10000g, the mass of the test sample m1 =9888g, and the threshold value in the image analysis is set to 120; the total number of edge lines is 2470, of which, taking edge line No. 587 as an example, the number of recognition points is 1992, the number of valid points is 1246, and the number after interpolation is 3601. The interpolation method is "cubic". The original data is interpolated after radius / angle conversion. The property coefficient of the closed segmentation figure is 0.632. After calculation, the radius of the segmented image without interpolation is 234, and the radius of the segmented image after interpolation is 186.

[0124] In this embodiment, the total weight of the overlapping pellet samples is m0 = 10000 g, the total weight of the pellet samples placed in the test platform collected by the weighing sensor is m1 = 8534 g, the pixel threshold in the image analysis is set to 120, and the total number of pellet sample areas identified is n0 = 127; among them, taking the edge identification of pellet sample area No. 587 as an example, Figure 7 As shown in the figure, the original effective edge pixel number is 1246, and the number of edge pixel points after interpolation is 3601. The interpolation method uses the Cubic interpolation method for difference processing. The process of edge closure processing by difference is as follows: Figure 8 As shown, the shape coefficient C of the pellet sample area after closed edge segmentation is calculated. i is 0.632; finally, the number of pellet sample areas n1=111 that meet the particle size requirements is detected.

[0125] The output results of the pellet size detection in this embodiment are:

[0126]

[0127] It can be seen that the pellet size detection system and detection method of the present invention overcome the shortcomings of traditional detection methods that destroy the performance of raw materials, and circumvent the problem that existing image analysis methods cannot effectively solve the problem of measuring the particle size of overlapping parts. The operation method is simple, saves manpower, and can perform pellet size detection on pellet ore more quickly, effectively and accurately.

[0128] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit the technical solutions. Those skilled in the art should understand that modifications or equivalent replacements of the technical solutions of the present invention that do not depart from the purpose and scope of the technical solutions of the present invention should be included in the scope of the claims of the present invention.

Claims

1. A pellet size detection system based on image processing, characterized in that: It includes a feeding device, a test container device and a detection device; The feeding device comprises a material guide port (2) and a screening transmission mechanism (3) for performing preliminary particle size screening on the pellet sample, and the material guide port (2) is connected to the test container device through the screening transmission mechanism (3); The test container device comprises a test platform (4), which is a platform-shaped container with a horizontal bottom plate and is used to hold pellet samples introduced through a material guide port (2) and a screening and transmission mechanism (3); a gas purge device (6) capable of blowing gas into the test platform (4) is provided on the outer sides of the test platform (4); a weighing sensor (5) is installed at the bottom of the test platform (4) for weighing the total weight of the pellet samples held in the test platform (4); The detection device comprises a detection computer (8) and an industrial camera (7) mounted directly above the test platform (4); the industrial camera (7) is used to collect images of the holding area of ​​the test platform (4); a data receiving end of the detection computer (8) is connected to an image data output end of the industrial camera (7) and a weight data output end of a weighing sensor (5) for data transmission, and the detection computer (8) is used to perform image analysis on the holding area image collected by the industrial camera (7), and to obtain a pellet size detection result of the pellet sample contained in the test platform (4) by combining the total weight of the pellet sample collected by the weighing sensor (5); The process of the detection computer (8) processing the pellet size detection result of the pellet sample placed in the test platform (4) is as follows: S1. Obtain the total weight m0 of the pellet sample to be tested, as well as the corresponding relationship between the actual space size and the pixel size corresponding to the holding area image, and save the data; S2. Binarizing the collected image of the holding area using a pixel threshold method, wherein the pixel threshold is set between the color pixel value of the pellet sample and the color pixel value of the horizontal base plate, to obtain a binary image; S3. Using a watershed segmentation algorithm to perform discretization segmentation processing on the obtained binary image into pellet sample regions, and performing edge closing processing on the pellet sample region edges of the segmented regions to obtain a pellet sample segmentation image containing the segmentation results of each pellet sample region; S4. Counting the total number n0 of pellet sample regions contained in the pellet sample segmentation image, and calculating the geometric center pixel position and pixel radius of each pellet sample region respectively, and calculating and determining the particle size of each pellet sample region based on the correspondence between the actual spatial size and the pixel size corresponding to the holding area image, and counting the number n1 of pellet sample regions that meet the particle size requirements; S5. Combined with the total weight m1 of the pellet sample placed in the test platform (4) collected by the weighing sensor (5), the qualified pellet ratio β1 and the unqualified pellet ratio β2 are calculated according to the following formulas, and output as the pellet size test results:

2. The pellet size detection system based on image processing according to claim 1, characterized in that: The screening transmission mechanism (3) comprises a screening conveyor belt formed by a plurality of transversely arranged screening rollers arranged in an inclined manner, and a conveyor cover arranged on both sides and above the screening conveyor belt, wherein the spacing between two adjacent screening rollers in the screening conveyor belt is 3 mm to 5.5 mm.

3. The pellet size detection system based on image processing according to claim 1, characterized in that: The color of the horizontal bottom plate of the test platform (4) is different from the color of the pellet sample.

4. The pellet size detection system based on image processing according to claim 1, characterized in that: The horizontal bottom plate of the test platform (4) is surrounded by side walls for preventing pellet samples from falling, and a plurality of ventilation slits are arranged in a grid pattern on the side walls, with the slit spacing being 3 mm to 5.5 mm. The gas purge device (6) is capable of spraying gas into the test platform (4) through the ventilation slits on the side walls.

5. The pellet size detection system based on image processing according to claim 1, characterized in that: In step S3, the edge closing process is performed on the edge of the segmented area by using Lagrange interpolation, Spline interpolation, Cubic interpolation or polynomial interpolation to perform difference processing on the edge of each pellet sample area in the segmented area to form a closed edge segmentation result of each pellet sample area.

6. The pellet size detection system based on image processing according to claim 1, characterized in that: The step S4 is specifically as follows: S401, counting the total number n0 of pellet sample regions contained in the pellet sample segmentation image; S402, for any i-th pellet sample region in the pellet sample segmentation image, calculate the pixel area A of the pellet sample region i and the contour perimeter P i , and determine the shape coefficient C of the i-th pellet sample area i =4πA i / P i 2 ; If the shape coefficient C i <0.5, then ignore the pellet sample area; if the shape coefficient C i ≥0.5, then calculate and determine the pixel position coordinates of the geometric center of the i-th pellet sample area. Go to step S403; S403: Count the pixel coordinates of the outline of the i-th pellet sample area relative to the geometric center pixel. The pixel distance of the i-th pellet sample area is determined by Mean i , combined with the correspondence between the actual space size and pixel size corresponding to the image of the holding area, the particle size G of the i-th pellet sample area is calculated i : G i =2×Mean i ×d max / n max ; Among them, d max / n max Indicates the correspondence ratio between the actual space size and pixel size corresponding to the holding area image, d max is the actual horizontal spatial size of the image of the storage area captured by the industrial camera, n max The total number of horizontal pixels of the storage area image captured by the industrial camera; S404 , using the processing of steps S402 to S403 , respectively calculate and determine the particle size of each pellet sample area, and then count the number n1 of pellet sample areas that meet the particle size requirements.

7. The pellet size detection system based on image processing according to claim 6, characterized in that: In step S402, the geometric center pixel position coordinates of the i-th pellet sample area are calculated and determined. The specific method is: Count the pixel coordinates of each contour of the i-th pellet sample area K represents the number of pixels contained in the outline of the i-th pellet sample area, and then the coordinates of the geometric center pixel position of the i-th pellet sample area are calculated as follows:

8. The pellet size detection system based on image processing according to claim 7, characterized in that: In step S403, the pixel radius Mean of the i-th pellet sample area is determined. i The specific method is: First, count the pixel coordinates of the contour of the i-th pellet sample area relative to the geometric center pixel The pixel distance L1, L2, ..., L k ,…,L K , K represents the number of pixels contained in the contour of the i-th pellet sample area, where: Then, determine the shape coefficient C of the i-th pellet sample area i The value size of ; If the shape factor 0.8≤C i ≤1, calculate the pixel radius of the i-th pellet sample area as follows: in, is the mean pixel distance of each contour pixel point relative to the geometric center pixel coordinate of the i-th pellet sample area; If the shape factor 0.5≤C i <0.8, then the coordinates of the outline pixels of the i-th pellet sample area relative to the geometric center pixel are counted. The pixel distance is greater than or equal to The pixel distance values ​​are marked as L′1, L′2, …, L′ k ,…,L′ N , N represents the statistical value greater than or equal to The number of pixel distance values ​​is then calculated as follows: i :

9. A pellet size detection method based on image processing, characterized in that: The pellet size detection system according to any one of claims 1 to 8 is used to perform the detection, specifically comprising the following steps: Step A, first weigh the total weight m0 of the pellet sample to be tested, and input the weight into the detection computer (8) to save the data; send all the pellet samples to be tested from the material guide port (2) into the pellet size detection system, and after preliminary particle size screening by the screening transmission mechanism (3), the pellet samples remaining after screening fall into the test platform (4), and start the gas purge device (6) to spray gas into the test platform (4), so that the pellet samples contained in the test platform (4) are evenly distributed, and then stop the operation of the gas purge device (6); Step B: The total weight m1 of the pellet sample placed in the test platform (4) is collected by the weighing sensor (5) and input into the detection computer (8) for data storage; then, an industrial camera (7) is used to collect an image of the placement area of ​​the test platform (4) and transmit it to the detection computer (8); Step C: The detection computer (8) performs image analysis on the image of the holding area captured by the industrial camera (7), and combines the total weight m1 of the pellet sample captured by the weighing sensor (5) to obtain the pellet size detection result of the pellet sample contained in the test platform (4).

Images