Steel slag volume stability detection method based on image recognition technology

Abstract

The invention discloses a steel slag volume stability detection method based on an image recognition technology, and the method comprises the steps: obtaining a to-be-recognized target image, and extracting the color features of the target image; according to the color features, a target image area meeting a first preset condition is screened, and the target image area comprises at least one steel slag; according to the pixel points of the target image area, the superficial area, the hole number and the hole diameter of the steel slag in the target image area are counted; calculating the surface porosity of the steel slag according to the surface area, the hole number and the hole diameter of the steel slag; and under the condition that the surface porosity is not smaller than the critical porosity, determining that the volume stability of the steel slag is poor. According to the steel slag volume stability detection method based on the image recognition technology, the volume stability of the steel slag can be rapidly and accurately judged in batches, and the detection efficiency of the volume stability of the steel slag is effectively improved.

Description

Technical Field

[0001] The present application belongs to the field of image recognition technology, and specifically, relates to a method for detecting the volume stability of steel slag based on image recognition technology. Background Art

[0002] Steel slag is the main solid waste from the steel industry. Because of its wear resistance, anti-skid, high strength and good adhesion, if steel slag is used as highway aggregate, it can not only effectively reduce the dependence of engineering construction on stone, solve the problem of shortage and uneven distribution of stone resources, but also protect the ecological environment and reduce pollution.

[0003] In recent years, many studies have been conducted on the application of steel slag in road engineering, but it has not been widely promoted and applied. This is mainly due to the fact that steel slag with poor volume stability is easy to react with water during storage and aging, causing the steel slag to expand or even crack, affecting the quality of the project. Therefore, before using steel slag for engineering construction, it is necessary to finely screen the steel slag to screen out steel slag with poor volume stability. The existing technology usually uses the following methods to judge the volume stability of steel slag, such as measuring the content of free calcium oxide in steel slag, measuring the pressure steam expansion rate of steel slag, etc. However, the above method has certain shortcomings, and it is impossible to quickly and accurately judge the volume stability of steel slag in batches, and the detection efficiency is low. Summary of the invention

[0004] The purpose of this application is to provide a steel slag volume stability detection method based on image recognition technology, which can quickly and accurately judge the steel slag volume stability in batches, thereby effectively improving the detection efficiency of the steel slag volume stability.

[0005] In order to achieve the above-mentioned technical effect, the present application provides a steel slag volume stability detection method based on image recognition technology, including: obtaining a target image to be identified, and extracting color features of the target image; based on the color features, screening a target image area that meets a first preset condition, the target image area containing at least one steel slag; based on the pixel points of the target image area, counting the surface area, number of pores and pore diameter of the steel slag in the target image area; calculating the surface porosity of the steel slag based on the surface area, number of pores and pore diameter of the steel slag; and when the surface porosity is not less than the critical porosity, determining that the volume stability of the steel slag is poor.

[0006] The solution provided in this application is based on image recognition technology. By extracting the color features of the target image and screening the target image area based on preset conditions, the steel slag sample can be identified more accurately. Furthermore, by counting the surface area, number of pores and pore size of the steel slag, and calculating the surface porosity accordingly, the volume stability of the steel slag is quantitatively evaluated by the surface porosity. Compared with the detection methods of the prior art, this solution can significantly improve the detection efficiency. By automatically processing image data, a large number of steel slag samples can be quickly identified and analyzed without the need for professional and cumbersome experimental operations and long waiting times, thereby achieving a rapid judgment of the volume stability of the steel slag.

[0007] As an improvement of the technical solution of the present application, the surface porosity of the steel slag is calculated according to the surface area, number of openings and pore diameter of the steel slag, including: the surface porosity of the steel slag is calculated according to the following formula:

[0008]

[0009] Where V represents the surface porosity, S h Indicates the opening area, S n represents the surface area of ​​slag, α is the correction factor, and the opening area S h It is calculated from the diameter of a single opening and the number of openings.

[0010] The solution provided in this application provides a method for quantitatively evaluating the volume stability of steel slag by establishing the relationship between the surface porosity of steel slag and the surface area, number of openings and aperture of steel slag, and the relationship between the surface porosity of steel slag and volume stability. This method accurately identifies the opening conditions of steel slag particles based on image recognition, and then determines the volume stability of steel slag particles, reducing the interference of human factors and improving the objectivity and accuracy of the test results.

[0011] As an improvement of the technical solution of the present application, the acquisition of the target image to be identified includes: acquiring images of the steel slag to be tested from multiple directions to obtain at least one steel slag image; extracting key feature points in each of the steel slag images; aligning and merging each of the steel slag images according to the key feature points to obtain the target image.

[0012] As an improvement of the technical solution of the present application, the method of screening the target image area that meets the first preset condition based on the color feature includes: performing threshold segmentation on the target image in the HSV color space according to a preset color threshold range to obtain a binary image of the target image; identifying connected areas in the binary image; and screening the connected areas that meet the first preset condition as the target image area based on the shape characteristics of the connected areas, wherein the first preset condition includes but is not limited to a minimum area threshold, a maximum area threshold and a shape factor of the connected area.

[0013] As an improvement of the technical solution of the present application, counting the surface area, number of pores and aperture of the steel slag in the target image area according to the pixel points of the target image area includes: counting the surface area of ​​the steel slag in the target image area according to the number of pixel points of the target image area; counting the number of pores and aperture of the steel slag in the target image area according to the color characteristics of the pixel points of the target image area.

[0014] Based on the previous improvement, further, counting the surface area of ​​the slag in the target image area according to the number of pixels in the target image area includes: counting the number of pixels in the connected area in the target image area; determining the surface area of ​​the slag in the target image area according to the number of pixels, the resolution of the target image and the unit length.

[0015] On the basis of the above improvements, further, the number and diameter of openings of the slag in the target image area are counted according to the color characteristics of the pixel points in the target image area, including: based on a starting pixel point in the target image area, the adjacent pixel points of the starting pixel point are determined in each target direction; when the color characteristics of the adjacent pixel points meet the second preset condition, the adjacent pixel point is used as the starting pixel point, and the adjacent pixel points of the starting pixel point are determined in each target direction; when the color characteristics of the adjacent pixel points do not meet the second preset condition, the starting pixel point and all the adjacent pixel points that meet the second preset condition are determined as an opening; and the diameter of the opening is determined according to the starting pixel point and all the adjacent pixel points that meet the second preset condition.

[0016] Furthermore, the second preset condition includes: the brightness of the pixel is not greater than 9%.

[0017] The beneficial effects of this application are:

[0018] 1. The solution provided in this application is based on image recognition technology. By extracting the color features of the target image and screening the target image area based on preset conditions, the steel slag sample can be identified more accurately. Furthermore, by counting the surface area, number of pores and pore size of the steel slag, and calculating the surface porosity accordingly, the volume stability of the steel slag is quantitatively evaluated by the surface porosity. Compared with the detection methods of the prior art, this solution can significantly improve the detection efficiency. By automatically processing image data, a large number of steel slag samples can be quickly identified and analyzed without the need for professional and cumbersome experimental operations and long waiting times, thereby achieving a rapid judgment of the volume stability of the steel slag.

[0019] 2. The solution provided in this application provides a method for quantitatively evaluating the volume stability of steel slag by establishing the relationship between the surface porosity of steel slag and the surface area, number of openings and pore diameter of steel slag, and the relationship between the surface porosity of steel slag and volume stability. This method accurately identifies the opening conditions of steel slag particles based on image recognition, and then determines the volume stability of steel slag particles, reducing the interference of human factors and improving the objectivity and accuracy of the test results. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0021] Figure 1 It is a schematic diagram of a flow chart of a method for detecting the volume stability of steel slag based on image recognition technology in an embodiment of the present application;

[0022] Figure 2 This is a schematic diagram of the relationship between the surface porosity of steel slag and its water expansion rate. DETAILED DESCRIPTION

[0023] The embodiments of the technical solution of the present application will be described in detail below. The following embodiments are only used to more clearly illustrate the technical solution of the present application, and are therefore only used as examples, and cannot be used to limit the scope of protection of the present application. In addition, the technical features involved in each embodiment of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0024] The following is combined with Figure 1 and 2 , describe in detail the technical solutions provided by each embodiment of the present application.

[0025] See also Figure 1 , is a flow chart of a method for detecting the volume stability of steel slag based on image recognition technology provided by an embodiment of the present application, and the method may include the following steps:

[0026] S1: Acquire a target image to be identified and extract color features of the target image.

[0027] Specifically, the target image is an image obtained by photographing the steel slag that needs to be tested for volume stability. In order to ensure that the steel slag features in the target image are clear, in one implementation, the step of obtaining the target image to be identified includes the following sub-steps:

[0028] S11: collecting images of the steel slag to be tested from multiple directions to obtain at least one steel slag image;

[0029] S12: extracting key feature points from each of the slag images;

[0030] S13: Aligning and merging the slag images according to the key feature points to obtain the target image.

[0031] The above-mentioned multiple directions include i directions, i=1, 2, 3, 4, 5, 6, i.e., front, back, left, right, top, and bottom. Since the morphology of slag in various directions may be different, it is easy for slag images at different angles to overlap when aligning and merging the slag images, so that the surface area of ​​slag in the slag image is larger than the actual surface area. Exemplarily, this embodiment can use ORB feature point detection and matching algorithm to perform ORB feature point detection on at least two slag images respectively, extract key feature points in each slag image, and obtain a feature point set of each slag image, wherein the key feature points include corner points. Then, a binary feature descriptor is generated by comparing randomly selected pixel pairs around the key feature points. Since the binary feature descriptor has rotation invariance, a matching algorithm can be used to match the feature descriptors of at least two images to obtain matching point pairs. The matching point pairs are screened according to the matching quality, and the wrong matching point pairs are removed. Multiple slag images are aligned and merged to obtain a target image with all the features of the slag to be tested. This embodiment uses the key feature points extracted from multi-directional slag images for image alignment, which can ensure that slag images taken at different angles can be accurately aligned when merged, avoiding image overlap and surface area erroneous increase due to multi-directional differences in slag morphology. The merged target image can more comprehensively display the three-dimensional morphology of slag, reduce morphological loss caused by shooting angle restrictions, and improve the accuracy of subsequent recognition.

[0032] It should be noted that changes in lighting conditions will affect color perception. In order to avoid the loss of characteristics of slag particles due to external factors such as lighting, the recommended light intensity when collecting images is 500 lux to 900 lux.

[0033] The color feature includes at least one of hue, saturation and brightness. The target image captured by the image acquisition device is generally in RGB color space. In order to extract the color feature of the target image, the target image needs to be converted to HSV color space.

[0034] S2: According to the color feature, screening a target image area that meets a first preset condition, wherein the target image area contains at least one steel slag.

[0035] In one implementation, in order to efficiently and accurately identify the target image area containing steel slag and reduce the interference of non-steel slag materials, this step includes the following sub-steps:

[0036] S21: performing threshold segmentation on the target image in the HSV color space according to a preset color threshold range to obtain a binary image of the target image;

[0037] Specifically, the color threshold range is that the pixel brightness is not higher than 40%. According to the statistical results of slag brightness under light intensity of 500lux to 900lux, the brightness of slag openings generally does not exceed 9%, and the surface brightness of dense slag does not exceed 40%. Threshold segmentation of the target image in the HSV color space can distinguish between pixels belonging to the slag color and pixels not belonging to the slag color. The binary image obtained after threshold segmentation only includes the slag marked as the foreground and the slag marked as the background.

[0038] S22: Identify connected regions in the binary image;

[0039] Each object in the image can be regarded as a connected region. This step further determines the steel slag to be detected in the image by identifying the connected region. Exemplarily, this step can use a commonly used connected domain analysis algorithm for identification, such as depth-first search DFS or breadth-first search BFS.

[0040] S23: According to the shape features of the connected areas, a connected area meeting a first preset condition is selected as a target image area, where the first preset condition includes but is not limited to a minimum area threshold, a maximum area threshold and a shape factor of the connected area.

[0041] Specifically, after identifying all connected areas, it is necessary to screen the target image areas that can be used for volume stability detection according to the preset shape feature conditions. Among them, the minimum area threshold in the first preset condition is used to exclude too small areas caused by noise or mis-segmentation, which are usually not valid slag areas. The maximum area threshold is used to prevent excessively large non-slag areas from being misjudged as target areas, especially when the image contains large non-slag objects with similar colors. The shape factor is a measure that describes the shape of a connected area and can be obtained by calculating the ratio of the perimeter of the area to the area. By setting an appropriate shape factor threshold, slag areas with more regular shapes or that meet specific requirements can be further screened out. The specific values ​​of the above minimum area threshold, maximum area threshold and shape factor can be set according to actual application requirements, and this embodiment does not impose any restrictions on this.

[0042] S3: Counting the surface area, number of openings and pore diameter of the slag in the target image area according to the pixel points in the target image area.

[0043] Specifically, according to the statistical results of slag brightness under light intensity of 500lux~900lux, the brightness of slag openings generally does not exceed 9%, and the surface brightness of dense slag is between 10% and 40%. Therefore, based on the brightness difference between slag openings and slag surface, key parameters such as surface area, number of openings and aperture of slag can be effectively counted according to the pixel points in the target image area.

[0044] S4: Calculate the surface porosity of the steel slag according to the surface area, number of pores and pore diameter of the steel slag.

[0045] In one implementation, the surface porosity of the steel slag can be calculated by the following formula (1):

[0046]

[0047] Where V represents the surface porosity, S h Indicates the opening area, S n represents the surface area of ​​slag, α is the correction factor, and the opening area S h It is calculated from the diameter of a single opening and the number of openings.

[0048] Specifically, the surface porosity of steel slag is the ratio of the pore volume to the particle volume, which includes the slag surface and the internal voids of the steel slag. In the image recognition process, only the pore size and the number of openings on the steel slag surface are considered, and a preliminary relationship for the surface porosity of steel slag is established, that is, the ratio of the opening area of ​​steel slag to the surface area of ​​steel slag. Since there is a deviation between the surface porosity of steel slag calculated by this relationship and the actual surface porosity of steel slag, a correction coefficient α is introduced according to the proportional relationship between the surface area and volume of multiple groups of steel slag to reduce the deviation between the surface porosity of steel slag calculated by the above formula (1) and the actual surface porosity of steel slag. The correction coefficient α takes a value of 0.6 to 0.8.

[0049] S5: When the surface porosity is not less than the critical porosity, it is determined that the volume stability of the steel slag is poor.

[0050] Specifically, the above critical porosity is set to 6.0%.

[0051] The internal reason for the poor volume stability of steel slag is that free calcium oxide and free magnesium oxide come into contact with water and undergo hydration reaction, causing the volume of steel slag to expand by 0.98 to 1.28 times. The external reason for the poor volume stability of steel slag is the porous structure of steel slag, whose rich pore size and number of pores provide the necessary channels for the hydration of steel slag. The decisive factor for the difficulty of the hydration reaction of steel slag is the surface porosity of steel slag. Therefore, the surface porosity of steel slag is an important factor affecting the water expansion rate of steel slag.

[0052] In order to determine the effect of the surface porosity of steel slag on its water expansion rate, the inventors of the present application selected a certain amount of steel slag, calculated the surface porosity according to step S4, and calculated its water expansion rate after heat-stifling the steel slag.

[0053] Figure 2 The relationship between the surface porosity of steel slag and the water expansion rate measured after the hot stuffing process is shown in the figure. Figure 2 It can be seen that the water expansion rate of steel slag gradually increases with the increase of surface porosity. When the surface porosity of steel slag is less than 6.0%, the steel slag is dense, the pore size is small and the number of pores is small. The hydration channels of free calcium oxide and free magnesium oxide are few, and it is not easy to carry out hydration reaction. Therefore, the water expansion rate of steel slag is low, and the volume stability of steel slag is good; when the surface porosity of steel slag is greater than or equal to 6.0%, the surface of steel slag is porous, the pore size is large and the number of pores is large. The hydration reaction of free calcium oxide and free magnesium oxide is easy to proceed. The water expansion rate of steel slag is high, and the volume stability of steel slag is poor. The existing specifications require that when steel slag is used as highway aggregate, its water expansion rate limit is 1.0% to 3.0%. Therefore, according to Figure 2 The results show that in order to ensure that the water expansion rate of steel slag does not exceed 1.0%,

[0054] The solution provided in this application provides a method for quantitatively evaluating the volume stability of steel slag by establishing the relationship between the surface porosity of steel slag and the surface area, number of openings and aperture of steel slag, and the relationship between the surface porosity of steel slag and volume stability. This method accurately identifies the opening conditions of steel slag particles based on image recognition, and then determines the volume stability of steel slag particles, reducing the interference of human factors and improving the objectivity and accuracy of the test results.

[0055] In order to ensure the accuracy of the porosity of the steel slag surface, in one implementation, the counting of the surface area, the number of pores and the pore diameter of the steel slag in the target image area according to the pixel points of the target image area includes the following sub-steps:

[0056] S31: counting the surface area of ​​the slag in the target image area according to the number of pixels in the target image area;

[0057] In one implementation, this step includes: counting the number of pixels in the connected area within the target image area; and determining the surface area of ​​the slag within the target image area based on the number of pixels, the resolution of the target image and the unit length.

[0058] Specifically, the resolution refers to the width and height of the image in pixels, and the unit length is the actual physical length represented by each pixel, usually in millimeters, centimeters or meters. Based on the number of pixels, the resolution of the target image and the unit length, the actual physical surface area of ​​the slag in the target image area can be calculated.

[0059] S32: Counting the number and diameter of openings of the slag in the target image area according to the color features of the pixels in the target image area.

[0060] In one implementation, this step includes the following sub-steps:

[0061] S321: Based on a starting pixel point in the target image area, determine the neighboring pixel points of the starting pixel point in each target direction;

[0062] S322: When the color feature of the adjacent pixel point meets the second preset condition, use the adjacent pixel point as the starting pixel point, and determine the adjacent pixel points of the starting pixel point in each target direction;

[0063] S323: If the color characteristics of the adjacent pixel points do not satisfy the second preset condition, determine the starting pixel point and all adjacent pixel points satisfying the second preset condition as an opening;

[0064] S324: Determine the aperture of the opening according to the starting pixel point and all adjacent pixel points that meet a second preset condition.

[0065] Specifically, the second preset condition is that the pixel brightness is not greater than 9%. The starting pixel point is a random pixel point in the target image area, and a recursive search is performed with the starting pixel point. When the color characteristics of the adjacent pixel points fall within the range of the second preset condition, it is considered that the adjacent pixel points and the starting pixel point belong to the same opening area. When an adjacent pixel point that does not meet the second preset condition is encountered during the recursive search, it means that the boundary of the current opening area has been searched. The number of openings in the slag in the target image area is thereby determined. It should be noted that after step S324, the number of openings and the aperture of the slag need to be converted into actual physical dimensions based on the resolution of the target image.

[0066] In order to verify the accuracy of the numerical values ​​of the surface area, number of openings and aperture of steel slag in the above embodiment, the inventor of the present application set up a comparative experiment and used 50 steel slags of different conditions as experimental samples. The size range of steel slag was controlled at 10-15 mm. The surface area, number of openings and aperture of each steel slag in the experimental sample were calculated by artificial observation of the microstructure of steel slag and image recognition. The recognition results are shown in Table 1.

[0067] Table 1 Comparison results of the accuracy of various data of manual recognition and image recognition experimental samples

[0068] name Manual Identification Image Recognition Quantity of steel slag (pieces) 50 50 Time(s) 15780 6 Slag surface area accuracy (%) 78 98 Slag hole number accuracy (%) 87 100 Slag aperture accuracy (%) 34 98

[0069] The results in Table 1 above can verify that the method based on image recognition in the embodiment of the present application, that is, according to the pixel points of the target image area, the surface area, number of openings and aperture of the steel slag in the target image area are counted, compared with the recognition method of the prior art, it has higher accuracy and recognition speed. Therefore, the scheme provided in the present application is based on image recognition technology, which can more accurately identify the steel slag sample by extracting the color features of the target image and screening the target image area based on preset conditions. Furthermore, by counting the surface area, number of openings and aperture of the steel slag, and calculating the surface porosity accordingly, the volume stability of the steel slag is quantitatively evaluated by the surface porosity. Compared with the detection method of the prior art, the present scheme can quickly identify and analyze a large number of steel slag samples by automatically processing image data, without the need for professional and cumbersome experimental operations and long waiting, thereby realizing a rapid judgment of the volume stability of the steel slag, which can significantly improve the detection efficiency.

[0070] 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 it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein by equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for detecting the volume stability of steel slag based on image recognition technology, characterized in that: include: Acquire a target image to be identified, and extract color features of the target image; According to the color feature, a target image region meeting a first preset condition is screened, wherein the target image region contains at least one steel slag; According to the pixel points of the target image area, the surface area, the number of openings and the aperture of the slag in the target image area are counted; Calculating the surface porosity of the steel slag according to the surface area, number of pores and pore diameter of the steel slag; When the surface porosity is not less than the critical porosity, it is determined that the volume stability of the steel slag is poor.

2. The method for detecting the volume stability of steel slag based on image recognition technology according to claim 1, characterized in that: Calculating the surface porosity of the steel slag according to the surface area, pore number and pore diameter of the steel slag includes: calculating the surface porosity of the steel slag according to the following formula: Where V represents the surface porosity, S h Indicates the opening area, S n represents the surface area of ​​slag, α is the correction factor, and the opening area S h It is calculated from the diameter of a single opening and the number of openings.

3. The method for detecting the volume stability of steel slag based on image recognition technology according to claim 1, characterized in that: The obtaining of the target image to be identified comprises: Collect images of the steel slag to be tested from multiple directions to obtain at least one steel slag image; Extracting key feature points from each of the slag images; The slag images are aligned and merged according to the key feature points to obtain the target image.

4. The method for detecting the volume stability of steel slag based on image recognition technology according to claim 1, characterized in that: The screening of the target image area meeting the first preset condition according to the color feature comprises: According to a preset color threshold range, threshold segmentation is performed on the target image in the HSV color space to obtain a binary image of the target image; identifying connected regions in the binary image; According to the shape features of the connected areas, connected areas that meet first preset conditions are screened as target image areas, where the first preset conditions include but are not limited to a minimum area threshold, a maximum area threshold and a shape factor of the connected areas.

5. The method for detecting the volume stability of steel slag based on image recognition technology according to claim 1, characterized in that: The counting of the surface area, the number of openings and the aperture of the slag in the target image region according to the pixel points in the target image region comprises: Counting the surface area of ​​the slag in the target image area according to the number of pixels in the target image area; According to the color features of the pixels in the target image area, the number and diameter of the pores of the slag in the target image area are counted.

6. The method for detecting the volume stability of steel slag based on image recognition technology according to claim 5, characterized in that: The counting of the surface area of ​​the slag in the target image area according to the number of pixels in the target image area comprises: Counting the number of pixels in the connected area within the target image area; The surface area of ​​the slag in the target image region is determined according to the number of pixels, the resolution of the target image and the unit length.

7. The method for detecting the volume stability of steel slag based on image recognition technology according to claim 5, characterized in that: The counting of the number and diameter of openings of the slag in the target image area according to the color features of the pixels in the target image area includes: Based on a starting pixel point in the target image area, determining adjacent pixel points of the starting pixel point in each target direction; When the color characteristics of the adjacent pixel points satisfy a second preset condition, the adjacent pixel points are used as the starting pixel points, and adjacent pixel points of the starting pixel points are determined in each target direction; When the color characteristics of the adjacent pixels do not satisfy the second preset condition, the starting pixel and all adjacent pixels satisfying the second preset condition are determined as one opening; The aperture of the opening is determined according to the starting pixel point and all adjacent pixel points that meet a second preset condition.

8. The method for detecting the volume stability of steel slag based on image recognition technology according to claim 7, characterized in that: The second preset condition includes: the brightness of the pixel is not greater than 9%.

Images