A method and device for detecting the softness and hardness of a flame, an electronic device and a medium

Abstract

The present application provides a kind of method, device, electronic equipment and medium for detecting the softness and hardness of flame, the method comprises: by obtaining the original image of the flame in the converter steelmaking, and the original image is analyzed and processed, the softness and hardness of the flame in the converter steelmaking is identified, the softness and hardness of the flame in the converter steelmaking process can be quantitatively evaluated in real time and efficiently, and the evaluation result is returned to the automatic control equipment, to realize automatic steelmaking. Through the monitoring of the softness and hardness of the flame, the frequency of splashing in the middle of smelting can be effectively avoided, the cost is saved, the security risk is reduced, and the carbon content in the molten steel in the later stage of smelting can also be assisted to monitor, the terminal carbon hit rate is improved, the yield is improved and the smelting cycle is shortened.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of image processing, in particular to a method and device for detecting the softness and hardness of a flame, an electronic device and a medium. BACKGROUND

[0002] In the converter steelmaking process, due to the particularity of the converter structure, the state change of the molten steel in the furnace cannot be directly observed by the human eye or equipment, and can only be indirectly judged by observing the flame state or sound change at the furnace mouth. Among them, the softness and hardness of the flame is an important feature for judging the flame state in the smelting process: for example, an abnormally soft flame in the middle of smelting is an important feature of impending splashing; for example, in the late smelting, accurate control of the end point carbon content is of great significance to improve yield and production efficiency. The softness and hardness of the flame is also an important standard for evaluating the current carbon content in the molten steel.

[0003] As a key feature for judging the state of molten steel in the steelmaking process, the softness and hardness of the flame is currently based on the traditional way of human eye observation in the industrial converter steelmaking process. For example, an abnormally soft flame in the middle of smelting is an important feature of impending splashing, at which time the operator needs to accurately judge and timely operate the gun to avoid splashing, which has adverse effects on iron loss, furnace condition scouring, equipment damage, and worker safety. However, this working method is not only low in efficiency, but if the operator makes an incorrect judgment, it can even cause serious safety hazards, and manual observation errors are uncontrollable, and it is also difficult to accurately evaluate the carbon content in the current molten steel through the softness and hardness of the flame. SUMMARY

[0004] In view of the shortcomings of the above-mentioned related art, the present application provides a method and device for detecting the softness and hardness of a flame, an electronic device and a medium to solve the technical problem of detecting the softness and hardness of a flame.

[0005] The present application provides a method for detecting the softness and hardness of a flame, the method comprising: the method for detecting the softness and hardness of a flame comprises: acquiring a plurality of flame images containing a flame region in a converter steelmaking process, denoted as original flame images; converting the original flame images to generate a binary image; performing morphological connection on the binary image, and performing contour extraction on the binary image after morphological connection to obtain a region of interest contour; calculating the variance of a set of point coordinates in the region of interest contour, and determining the softness and hardness of the flame based on the variance.

[0006] In an embodiment of the present application, the original flame image is converted from a first color space to a second color space to obtain a pre-processed flame image; wherein the first color space is composed of red, green and blue, and the second color space is composed of hue, saturation and brightness; a hue upper limit value, a saturation upper limit value, a brightness upper limit value, a hue lower limit value, a saturation lower limit value and a brightness lower limit value are determined in the second color space based on a preset flame variation range, i.e. upper limit values and lower limit values of a mask are determined; the pre-processed flame image is compared with the upper limit values of the mask and the lower limit values of the mask respectively, and a binary image is generated according to a comparison result.

[0007] In an embodiment of the present application, the original flame image is converted from a first color space to a second color space to obtain a pre-processed flame image, including: if the hue of the pre-processed flame image is greater than a hue lower limit value in the second color space and less than a hue upper limit value in the second color space, the saturation is greater than a saturation lower limit value in the second color space and less than a saturation upper limit value in the second color space, and the brightness is greater than a brightness lower limit value in the second color space and less than a brightness upper limit value in the second color space, then the pre-processed flame image is determined as a white pixel value image; if the hue of the pre-processed flame image is less than or equal to the hue lower limit value in the second color space, the saturation is less than or equal to the saturation lower limit value in the second color space, and the brightness is less than or equal to the brightness lower limit value in the second color space, then the pre-processed flame image is determined as a black pixel value image; if the hue of the pre-processed flame image is greater than or equal to the hue upper limit value in the second color space, the saturation is greater than or equal to the saturation upper limit value in the second color space, and the brightness is greater than or equal to the brightness upper limit value in the second color space, then the pre-processed flame image is determined as a black pixel value image.

[0008] In an embodiment of the present application, the binary image is subjected to morphological connection, and the binary image after morphological connection is subjected to contour extraction to obtain a region of interest contour, including: a structure element is set; the structure element is subjected to a morphological dilation operation and then a morphological erosion operation, and the target region of the binary image is subjected to morphological connection; a closed region formed by horizontal coordinate values and vertical coordinate values of contour pixel points constituting a white part contour of the binary image is obtained, the closed region includes an inner contour and an outer contour; the inner contour of the closed region is removed, and the outer contour containing a contour with pixel points less than a first threshold value is removed to obtain the region of interest contour.

[0009] In an embodiment of the present application, the set point coordinates of the region of interest contour of the plurality of original flame images are extracted to form a set point coordinate group; the variance of the set point coordinate group is calculated; if the variance is less than a second threshold value and the coordinates of the set point of the region of interest contour corresponding to the current frame original flame image are greater than a third threshold value, a secondary judgment is made on the soft and hard degree of the flame, and a soft and hard degree recognition result of the flame is output based on the secondary judgment result; if the variance is greater than or equal to the second threshold value and the coordinates of the set point of the region of interest contour corresponding to the current frame original flame image are less than or equal to the third threshold value, the soft and hard degree recognition result of the flame is output.

[0010] In an embodiment of the present application, a monitoring area of a flame boundary is determined, and the monitoring area includes a left monitoring area and a right monitoring area; a weighted average value of the pixel point coordinates in the monitoring area contour is calculated to obtain a flame left boundary position coordinate and a flame right boundary position coordinate; and a soft and hard degree recognition result of the flame is output based on the flame left boundary position coordinate and the flame right boundary position coordinate.

[0011] In an embodiment of the present application, the soft and hard degree recognition result of the flame is output based on the flame left boundary position coordinate and the flame right boundary position coordinate, which includes outputting the soft and hard degree recognition result of the flame according to the flame left boundary position coordinate, the flame right boundary position coordinate, an upper limit of the right change of the flame left boundary, an upper limit of the left change of the flame right boundary, and the width of the flame in a binary image.

[0012]

[0013] wherein, boundary left represents the coordinate of the flame left boundary position, boundary right represents the coordinate of the flame right boundary position, x l is the upper limit of the right change of the flame left boundary during smelting, x r is the upper limit of the left change of the flame right boundary during smelting, width is the width of the flame in a binary image, and res is the final result of the mapped soft and hard degree of the flame. The smaller the res is, the softer the flame is, and the larger the res is, the harder the flame is.

[0014] Beneficial effects: By obtaining the original image of the flame in the converter steelmaking and analyzing and processing the original image, the soft and hard degree of the converter steelmaking flame is recognized, the soft and hard degree of the flame in the converter steelmaking process can be quantitatively evaluated in real time and efficiently, the evaluation result is returned to the automatic control equipment, and automatic steelmaking is realized. Through monitoring the soft and hard degree of the flame, the frequency of splashing in the middle of smelting can be effectively avoided, the cost is saved, the security risk is reduced, the carbon content in the molten steel in the later stage of smelting can be also assisted to be monitored, the terminal carbon hit rate is improved, the yield is improved, and the smelting cycle is shortened.

[0015] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and do not limit this application. Attached Figure Description

[0016] Figure 1 This is a schematic diagram illustrating the implementation environment of a method for detecting the hardness of a flame, as shown in an exemplary embodiment of this application.

[0017] Figure 2 This is a flowchart illustrating a method for detecting the hardness of a flame, as shown in an exemplary embodiment of this application;

[0018] Figure 3 This is a flowchart illustrating a method for detecting the hardness of a flame, as shown in an exemplary embodiment of this application;

[0019] Figure 4 This is a schematic diagram illustrating the extracted flame outline as shown in an exemplary embodiment of this application;

[0020] Figure 5 This is a schematic diagram illustrating the extracted flame outline as shown in an exemplary embodiment of this application;

[0021] Figure 6 This is a schematic diagram illustrating the extracted flame outline as shown in an exemplary embodiment of this application;

[0022] Figure 7 This is a schematic diagram of a flame monitoring area shown in an exemplary embodiment of this application;

[0023] Figure 8 This is a block diagram illustrating the detection of the hardness of a flame in an exemplary embodiment of this application;

[0024] Figure 9 This is a schematic diagram illustrating the structure of an electronic device as shown in an exemplary embodiment of this application. Detailed Implementation

[0025] The following specific examples illustrate the implementation of this application. Those skilled in the art can easily understand other advantages and effects of this application from the content disclosed in this specification. This application can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of this application. It should be noted that, unless otherwise specified, the following embodiments and features in the embodiments can be combined with each other.

[0026] It should be noted that the drawings provided in the following embodiments only schematically illustrate the basic concept of the present application, and only the components related to the present application are shown in the drawings, not the number, shape and size of the components when actually implemented. The actual implementation of each component may be a random change, and the component layout pattern may be more complex.

[0027] It should be noted that in the present application, "first", "second", etc. are only for the differentiation of similar objects, and are not limited in order or sequence. The described "includes", "has", etc. means that the subject of the word covers the range in addition to the examples shown by the word, and is not exclusive.

[0028] It can be understood that the various numbers, step numbers, etc. in the present application are distinguished for convenience of description, and do not limit the scope of the present application. The size of the present application does not mean the order of execution, and the execution order of each process should be determined by its function and internal logic.

[0029] In the following description, a large number of details are discussed to provide a more thorough explanation of the embodiments of the present application, however, it is obvious to those skilled in the art that the embodiments of the present application can be implemented without these specific details, and in other embodiments, well-known structures and devices are shown in the form of block diagrams rather than in detail, to avoid making the embodiments of the present application difficult to understand.

[0030] The embodiments of the present application respectively propose a method for detecting the softness and hardness of a flame, a device for detecting the softness and hardness of a flame, an electronic device, a computer readable storage medium and a computer program product, which will be described in detail below.

[0031] First of all, it should be noted that computer vision technology (Computer Vision, abbreviated as CV) is a science that studies how to make machines "see". Further, it means that the camera and visual detection equipment replace the human eye to identify, track and measure the target machine vision, and further process the image to make the visual detection equipment processing more suitable for human eye observation or transmission to instrument detection. As a scientific discipline, computer vision researches related theories and technologies, trying to establish an artificial intelligence system that can obtain information from images or multidimensional data. Computer vision technology usually includes image processing, image recognition, image semantic understanding, image retrieval, OCR, video processing, video semantic understanding, video content / behavior recognition, three-dimensional object reconstruction, 3D technology, virtual reality, augmented reality, simultaneous localization and mapping, etc.

[0032] Please refer to Figure 1 ,Figure 1 is a schematic diagram of an implementation environment of a method for detecting the softness and hardness of a flame according to an example embodiment of the present application. As shown in Figure 1 , Figure 1 includes a converter 101, a visual detection device 102. In an example embodiment of the present application, a plurality of flame images of a flame region in a steelmaking process of the converter 101 are acquired by the visual detection device 102, denoted as original flame images, the original flame images are converted to generate binary images, the binary images are morphologically connected, the contours of the binary images after morphological connection are extracted to obtain a region of interest contour; the variance of a set of point coordinates of the region of interest contour is calculated, and the softness and hardness of the flame are determined based on the variance. In this embodiment, the visual detection device is an intelligent industrial camera, which generally consists of an image acquisition unit, an image processing unit, image processing software, network communication devices, etc. By acquiring the original images of the flame in the converter steelmaking process and analyzing and processing the original images, the softness and hardness of the flame in the converter steelmaking process can be identified, and the evaluation results can be returned to the automatic control device to realize automatic steelmaking. Through monitoring the softness and hardness of the flame, the frequency of splashing in the middle of the smelting process can be effectively avoided, the cost can be saved, the safety hidden danger can be reduced, and the carbon content in the molten steel in the later stage of smelting can also be assisted to monitor, the terminal carbon hit rate can be improved, the yield can be improved, and the smelting cycle can be shortened.

[0033] Please refer to Figure 2 , Figure 2 is a flowchart of a method for detecting the softness and hardness of a flame according to an example embodiment of the present application. The method can be applied to the implementation environment shown in Figure 1 It should be understood that the method can also be applied to other example implementation environments and be specifically executed by devices in other implementation environments, and the present embodiment does not limit the implementation environment to which the method is applied.

[0034] As shown in Figure 2 , in an example embodiment, the method for detecting the softness and hardness of a flame at least includes steps S210 to S240, which are described in detail as follows:

[0035] Step S210, acquiring a plurality of flame images containing a flame region in a converter steelmaking process, denoted as original flame images.

[0036] In an example embodiment of the present application, a plurality of flame images containing a flame region in a converter steelmaking process are acquired by an intelligent industrial camera.

[0037] In an embodiment of the present application, before the flame image is acquired, the appropriate image acquisition device is selected according to the position of the intelligent industrial camera and the furnace mouth, so that the flame image of the furnace mouth can be completely imaged in the image, and the flame is located as much as possible in the central region of the image.

[0038] In step S220, the original flame image is converted to generate a binary image.

[0039] In step S230, morphological connection is performed on the binary image, and contour extraction is performed on the binary image after the morphological connection to obtain a contour of a region of interest.

[0040] In an embodiment of the present application, the purpose of the morphological operation is to extract component information in the image, which is usually important for expressing and depicting the shape of the image, and is usually the most essential shape feature used in image understanding. Contour extraction is performed on the processed binary image to obtain a contour of a region of interest. Through the morphological connection operation, small holes in the binary image can be connected to eliminate noise points, so that a clearer binary image is obtained to facilitate subsequent analysis.

[0041] In step S240, the variance of the set point coordinate group in the contour of the region of interest is calculated, and the soft and hard degrees of the flame are determined based on the variance.

[0042] In Figure 2 In the technical solution shown, the original image of the flame in the converter steelmaking is acquired, the original image is analyzed and processed, the soft and hard degrees of the converter steelmaking flame are identified, morphological operation processing is performed to obtain a contour region of interest, the variance of the highest point longitudinal coordinate group of the contour region of interest is calculated, and the soft and hard degrees of the flame in the converter steelmaking process are determined based on the variance. Through monitoring of the soft and hard degrees of the flame, the frequency of splashing in the middle stage of smelting can be effectively avoided, cost can be saved, safety hazards can be reduced, the carbon content in the molten steel in the late stage of smelting can be assisted to monitor, the terminal carbon hit rate can be improved, the yield can be improved, and the smelting cycle can be shortened.

[0043] In an embodiment of the present application, the process of converting the original flame image to generate a binary image includes converting the original flame image from a first color space to a second color space to obtain a pre-processed flame image; wherein the first color space is composed of red, green and blue, and the second color space is composed of hue, saturation and brightness; determining the upper limit value of hue, the upper limit value of saturation, the upper limit value of brightness, the lower limit value of hue, the lower limit value of saturation and the lower limit value of brightness in the second color space based on a preset flame variation range, i.e. determining the upper limit value and the lower limit value of the mask; comparing the pre-processed flame image with the upper limit value of the mask and the lower limit value of the mask respectively, and generating a binary image according to the comparison result. By pre-processing the original flame image to obtain a pre-processed flame image, a clearer binary image can be obtained, and the soft and hard degree of the flame can be more accurately determined.

[0044] In an embodiment of the present application, the original flame image is a BGR (blue, green, red) image, the flame region is extracted and a binary image is generated by converting the BGR image into HSV (hue, saturation, brightness) and setting a suitable HSV mask. In this embodiment, BGR is a color standard, which refers to blue B, green G and red R. All colors are obtained by changing and superimposing these three color channels. HSV is a color space created according to the intuitive characteristics of colors, also known as a hexagonal cone model. The parameters of colors in this model are: hue (H), saturation (S) and brightness (V). Hue H: measured by angle, the value range is 0°-360°, calculated in the counterclockwise direction from red, which is 0°, green is 120°, and blue is 240°. Their complementary colors are: yellow is 60°, cyan is 180°, and purple is 360°. Saturation S: saturation S represents the degree of color close to the spectrum color. A color can be regarded as the result of mixing a certain spectrum color with white. The greater the proportion of spectrum color, the higher the degree of color close to the spectrum color, and the higher the saturation of the color. The white light component of the spectrum color is 0, and the saturation reaches the highest. The value range is usually 0%-100%, and the larger the value, the more saturated the color. Brightness V: brightness represents the degree of color brightness. For light source color, the brightness value is related to the brightness of the light source; for object color, this value is related to the transmittance or reflectance of the object. The value range is usually 0% (black) to 100% (white).

[0045] In an embodiment of the present application, in digital image processing, a mask can be used to shield all or part of an image with selected images, patterns or objects to control the area or process of image processing. The image mask is mainly used for the following purposes: extracting a region of interest, multiplying a pre-made mask of the region of interest with the image to be processed to obtain a region of interest image, the image values in the region of interest remain unchanged, and the image values outside the region are all 0; shielding, using a mask to shield certain areas of the image so that they do not participate in the processing or calculation of processing parameters, or only the shielded area is processed or counted; structure feature extraction, using similarity variables or image matching methods to detect and extract similar structural features in the image.

[0046] In an embodiment of the present application, the first color space is a BGR color space, and the second color space is an HSV color space. The BGR color space image obtained from the visual detection device is converted to an HSV color space image, and an HSV mask close to the characteristic change range of the flame is defined for binary image generation. The format of the HSV mask is as follows:

[0047] thresh lower =[H lower ,S lower ,V lower ]

[0048] thresh upper =[H upper ,S upper ,V upper ]

[0049] wherein H lower represents the lower limit value of the hue of the HSV mask, S lower represents the lower limit value of the saturation of the HSV mask, V lower represents the lower limit value of the brightness of the HSV mask, H upper represents the upper limit value of the hue of the HSV mask, S upper represents the upper limit value of the saturation of the HSV mask, and V upper represents the upper limit value of the brightness of the HSV mask.

[0050] In an embodiment of the present application, the original flame image is converted from a first color space to a second color space to obtain a pre-processed flame image, including: if the hue of the pre-processed flame image is greater than a lower limit value of the hue in the second color space and less than an upper limit value of the hue in the second color space, the saturation is greater than a lower limit value of the saturation in the second color space and less than an upper limit value of the saturation in the second color space, the brightness is greater than a lower limit value of the brightness in the second color space and less than an upper limit value of the brightness in the second color space, the pre-processed flame image is determined as a white pixel value image; if the hue of the pre-processed flame image is less than or equal to the lower limit value of the hue in the second color space, the saturation is less than or equal to the lower limit value of the saturation in the second color space, and the brightness is less than or equal to the lower limit value of the brightness in the second color space, the pre-processed flame image is determined as a black pixel value image; if the hue of the pre-processed flame image is greater than or equal to the upper limit value of the hue in the second color space, the saturation is greater than or equal to the upper limit value of the saturation in the second color space, and the brightness is greater than or equal to the upper limit value of the brightness in the second color space, the pre-processed flame image is determined as a black pixel value image. The pre-processed flame image is compared with the upper limit value and the lower limit value of the mask, so that a clearer binary image is obtained, and the soft and hard degrees of the flame are more accurately determined.

[0051] In an example of the present application, the HSV region close to the characteristic change range of the flame is defined as an HSV mask; the pre-processed flame image is compared with the upper limit value and the lower limit value of the HSV mask, and the black and white pixel values are determined according to the HSV mask, and the formula for determining the black and white pixel values is as follows:

[0052]

[0053] wherein imgbinary(x, y) is the generated binary image of the flame region, 255 is the white pixel value, 0 is the black pixel value, thresh upper and thresh lower is the HSV mask, and imgHSV(i, j) is the HSV color space image.

[0054] In an example of the present application, the binary image after morphological connection is subjected to contour extraction to obtain a region of interest contour, including: setting a structure element; performing morphological dilation operation on the structure element, and then performing morphological erosion operation, and performing morphological connection on the target region of the binary image; obtaining a closed region formed by the horizontal coordinate value and the vertical coordinate value of the contour pixel point in the contour of the white part of the binary image, the closed region including an inner contour and an outer contour; removing the inner contour of the closed region, and removing the contour containing less than a first threshold value of pixel points in the outer contour to obtain the region of interest contour. In this embodiment, the first threshold value is determined according to the actual image size. By performing contour extraction on the processed binary image, removing the invalid internal contour, and the external contour with less than a preset number of pixels, the internal fireworks of the flame can be removed, and the spark interference outside the flame region can be removed. According to the extracted contour, the softness and hardness of the flame are analyzed, so that more accurate softness and hardness analysis results of the flame can be obtained, the monitoring of the softness and hardness of the flame can be realized, the frequency of splashing in the middle of smelting can be effectively avoided, the cost can be saved, and the security risk can be reduced.

[0055] In an embodiment of the present application, the erosion operation is one of the most basic operations in morphological operation, which can eliminate the boundary points of the image, shrink the image inward along the boundary, or remove the part smaller than the specified structure element. Erosion is used to "shrink" or "thinning" the foreground in the binary image, thereby achieving the functions of removing noise, element segmentation, etc. The dilation operation is another basic operation in morphological operation. The dilation operation and the erosion operation have opposite effects. The dilation operation can expand the boundary of the image. The dilation operation merges the pixel points in the background that contact the foreground object into the foreground object, thereby expanding the boundary points of the image outward.

[0056] In an embodiment of the present application, the softness and hardness of the flame are determined based on the variance of the set point coordinate group in the region of interest contour by calculating the variance of the set point coordinate group in the region of interest contour, including: extracting the set point coordinates of the region of interest contour of the multiple frames of original flame images to form a set point coordinate group; calculating the variance of the set point coordinate group; if the variance is less than a second threshold value and the coordinate of the set point of the region of interest contour corresponding to the current frame of flame image is greater than a third threshold value, performing secondary judgment on the softness and hardness of the flame, and outputting a softness and hardness recognition result of the flame based on the secondary judgment result; if the variance is greater than or equal to the second threshold value and the coordinate of the set point of the region of interest contour corresponding to the current frame of original flame image is less than or equal to the third threshold value, outputting the softness and hardness recognition result of the flame. Since the variance can reflect the dispersion degree of the measured value, the height range of the flame is analyzed based on the variance, the analysis result of the softness and hardness of the flame is obtained, and the monitoring of the softness and hardness of the flame can effectively avoid the frequency of splashing in the middle of smelting, save the cost, and reduce the security risk.

[0057] In an embodiment of the present application, the set point coordinate is the highest point longitudinal coordinate, the highest point longitudinal coordinates of the contours of the regions of interest of the plurality of original flame images are extracted to form a longitudinal coordinate group, the variance of the longitudinal coordinate group is calculated, and the following formula is used to calculate the variance of the coordinate group formed by the highest point longitudinal coordinates of the contours of the regions of interest in the continuous image frames before the current frame:

[0058]

[0059]

[0060] wherein S 2 represents the variance of the coordinate group formed by the highest point longitudinal coordinates of the contours of the regions of interest in the continuous image frames before the current frame, m represents the number of the continuous image frames, n represents the frame number of the current image, h represents the highest point longitudinal coordinate of the contour obtained from the contour of the region of interest extracted in the foregoing, flag represents the result of the preliminary judgment, and thresh represents the threshold value for judging the variance. In this embodiment, the second threshold value is thresh, and the third threshold value is 0. When the variance is less than thresh and the highest point longitudinal coordinate of the contour of the region of interest corresponding to the current flame image is greater than 0, a secondary judgment of the softness and hardness of the flame is required. When the variance is greater than or equal to thresh or the current longitudinal coordinate is less than or equal to 0, the softness and hardness of the flame res = 1 is directly output. Since the variance can reflect the dispersion degree of the values, the height range of the flame is analyzed by the variance, the analysis result of the softness and hardness of the flame is obtained based on the height range of the flame, and the monitoring of the softness and hardness of the flame can effectively avoid the frequency of splashing in the middle period of smelting, save the cost, and reduce the safety hidden danger.

[0061] In an embodiment of the present application, the softness and hardness of the flame recognition result is output based on the secondary judgment result, which comprises determining the monitoring region of the flame boundary, the monitoring range comprising a left monitoring region and a right monitoring region; performing weighted average value calculation on the pixel points in the contour of the monitoring region to obtain the flame left boundary position coordinate and the flame right boundary position coordinate; and outputting the softness and hardness of the flame recognition result based on the flame left boundary position coordinate and the flame right boundary position coordinate. By normalizing the left boundary position coordinate and the flame right boundary position coordinate, the softness and hardness of the current flame can be quantified, the softness and hardness of the flame result can be more accurately obtained, the frequency of splashing in the middle period of smelting can be effectively avoided, the cost can be saved, the safety hidden danger can be reduced, the carbon content in the molten steel in the later period of smelting can be also monitored, the end-point carbon hit rate can be improved, and the yield can be improved and the smelting period can be shortened.

[0062] In one embodiment of this application, the left boundary position coordinate is the abscissa of the left boundary position, and the right boundary position coordinate is the abscissa of the right boundary position. The monitoring area of ​​the flame boundary is determined by the following method: based on the preset position of the currently deployed converter mouth area in the image, the preset boundary range of the flame boundary during normal smelting is delineated, and two abscissas are formed. The range of the monitoring area is as follows:

[0063] X left ∈[0,x l Equation (4)

[0064]

[0065] Among them, X left X represents the range of changes in the horizontal coordinate of the monitoring area on the left. right x represents the range of changes in the x-coordinate of the monitoring area on the right. l x represents the upper limit of the change in the left boundary of the flame to the right during normal smelting. r The upper limit of the change of the right boundary of the flame to the left during normal smelting, and width is the width of the flame in the binary image.

[0066] In the above embodiment, based on equation (3), when the judgment result output by equation (3) is true, the monitoring area of ​​the flame boundary is determined, and the monitoring range includes the left monitoring area and the right monitoring area; the weighted average value of the pixels in the contour of the monitoring area is calculated to obtain the abscissa of the left boundary position and the abscissa of the right boundary position of the flame; the average value of the abscissa of all pixels in the contour within the monitoring area is taken as the centroid abscissa of the current contour, and then a weighted average value is calculated to obtain the coordinates of the left and right boundary positions, as follows:

[0067]

[0068]

[0069] Where a and b are the number of flame outlines in the left and right monitoring areas, respectively, and leftnum idx `rightnum` represents the number of pixels included in the `idx`-th contour within the left monitoring region. idx leftnum represents the number of pixels included in the idx-th contour within the right monitoring region. idx For the left monitoring area, the id x The centroid x-coordinate of the contour, rightx idx Let x be the centroid x-coordinate of the idx-th contour within the right monitoring region, and boundary. left The x-coordinate of the calculated left boundary position of the flame is called boundary. right The x-coordinate of the calculated right boundary position of the flame.

[0070] In an embodiment of the present application, the flame soft and hard degree recognition result is output based on the flame left boundary position coordinate and the flame right boundary position coordinate, including outputting the flame soft and hard degree recognition result according to the flame left boundary position coordinate, the flame right boundary position coordinate, the upper limit of the flame left boundary changing to the right, the upper limit of the flame right boundary changing to the left, and the width of the flame in the binary image, and the flame soft and hard degree recognition result has:

[0071]

[0072] wherein boundary left represents the coordinate of the flame left boundary position, boundary right represents the coordinate of the flame right boundary position, x l is the upper limit of the flame left boundary changing to the right during smelting, x r is the upper limit of the flame right boundary changing to the left during smelting, width is the width of the flame in the binary image, and res is the final result of the mapped flame soft and hard degree. The smaller the res is, the softer the current flame is, and the larger the res is, the harder the current flame is. By normalizing the left boundary position coordinate and the flame right boundary position coordinate, the soft and hard degree of the current flame can be quantified, so that the flame soft and hard degree result can be obtained more accurately, the frequency of splashing in the middle period of smelting can be effectively avoided, the cost can be saved, the security risks can be reduced, the carbon content in the molten steel in the later period of smelting can be monitored, the terminal carbon hit rate can be improved, the yield can be improved, and the smelting period can be shortened.

[0073] In an embodiment of the present application, the flame left boundary position coordinate is the horizontal coordinate of the flame left boundary position, and the flame right boundary position coordinate is the right coordinate of the flame right boundary position. The flame soft and hard degree recognition result is output according to the horizontal coordinate of the flame left boundary position, the horizontal coordinate of the flame right boundary position, the upper limit of the flame left boundary changing to the right, the upper limit of the flame right boundary changing to the left, and the width of the flame in the binary image. By normalizing the left boundary position coordinate and the flame right boundary position coordinate, the soft and hard degree of the current flame can be quantified, so that the flame soft and hard degree result can be obtained more accurately, the frequency of splashing in the middle period of smelting can be effectively avoided, the cost can be saved, the security risks can be reduced, the carbon content in the molten steel in the later period of smelting can be monitored, the terminal carbon hit rate can be improved, the yield can be improved, and the smelting period can be shortened.

[0074] In an embodiment of the present application, the setting of the flame soft and hard monitoring area is related to the monitoring range of the flame soft and hard degree. The flame soft and hard monitoring area can be directly set according to experience, or the flame change range during normal smelting can be counted to obtain a more accurate range.

[0075] In an embodiment of the present application, please refer to Figure 3 , Figure 3A flowchart of the method for detecting the softness and hardness of the flame shown in the embodiments of the present application specifically includes the following steps:

[0076] In step S310, a plurality of flame images containing a flame area in the converter steelmaking process are collected in real time, denoted as original flame images.

[0077] In step S320, the original flame images are converted to generate binary images.

[0078] In step S330, the binary images are morphologically connected, and the contours of the connected binary images are extracted to obtain the contours of the regions of interest.

[0079] In step S340, the setpoint vertical coordinates of the contours of the regions of interest of the plurality of original flame images are extracted to form a setpoint vertical coordinate group.

[0080] In step S350, the variance of the setpoint vertical coordinate group is calculated.

[0081] In step S360, if the variance is greater than or equal to a second threshold value and the vertical coordinate of the highest point of the contour of the region of interest corresponding to the current frame original flame image is less than or equal to a third threshold value, the softness and hardness recognition result of the flame is output.

[0082] In step S370, if the variance is less than the second threshold value and the vertical coordinate of the highest point of the contour of the region of interest corresponding to the current frame original flame image is greater than the third threshold value, the monitoring area of the flame boundary is determined, and the monitoring range includes a left monitoring area and a right monitoring area.

[0083] In step S380, the weighted average value of the horizontal coordinates of the pixel points in the monitoring area contour is calculated to obtain the horizontal coordinates of the left and right boundary positions of the flame.

[0084] In step S390, the softness and hardness recognition result of the flame is output based on the horizontal coordinates of the left and right boundary positions of the flame, i.e., the change of the flame boundary is mapped to the softness and hardness of the flame and the recognition result 0-1 is output.

[0085] In the technical solution shown in Figure 3 In the technical solution shown in

[0086] In an embodiment of the present application, the first color space is a BGR color space, the second color space is an HSV color space, a plurality of flame images containing a flame region in a converter steelmaking process are obtained, denoted as original flame images, the original flame images are in the BGR color space, the original flame images are processed, that is, the original flame images are preprocessed, the original flame images are converted from the BGR color space to the HSV color space, denoted as a preprocessed image, by setting a suitable HSV mask, the upper and lower limit values of the mask are determined, that is, the hue upper limit value, the saturation upper limit value, the brightness upper limit value, the hue lower limit value, the saturation lower limit value and the brightness lower limit value in the second color space based on the preset flame variation range, that is, the upper limit value and the lower limit value of the mask are determined, the flame region is extracted and a binary image is generated, that is, by comparing the preprocessed image with the upper and lower limit values of the HSV mask, a binary image is obtained. By morphological connection operation, the set structure element is subjected to morphological dilation operation, and then morphological erosion operation is performed, so that small holes in the binary image are connected to achieve the purpose of eliminating noise points. The closed region formed by the horizontal coordinate values and the vertical coordinate values of the contour pixels constituting the white part contour of the binary image is obtained, the closed region includes an inner contour and an outer contour; the inner contour of the closed region is removed, the contour containing less than a first threshold value of pixel points in the outer contour is removed, and a region of interest contour is obtained, which can remove the internal smoke of the flame or the spark interference outside the flame region, and obtain the final flame region contour as shown in Figure 4 In this embodiment, the set point vertical coordinates of the regions of interest contours of the plurality of original flame images are extracted to form a set point vertical coordinate group; the variance of the set point vertical coordinate group is calculated; if the variance is greater than or equal to a second threshold value and the vertical coordinate of the set point of the region of interest contour corresponding to the current frame of the original flame image is less than or equal to a third threshold value, a flame soft and hard degree recognition result is output, in this embodiment, the set point is the highest point, the variance of the vertical coordinate group of the highest points of the continuous frame flame regions of interest contours is calculated, the height change of the continuous frame flame regions of interest contours is analyzed through the variance, if the flame image exists as Figure 5 and Figure 6 When these two states occur, the flame soft and hard degree is directly determined as 1. If the variance is less than the second threshold value and the vertical coordinate of the set point of the region of interest contour corresponding to the current frame of the original flame image is greater than the third threshold value, the flame soft and hard degree is judged again, and a flame soft and hard degree recognition result is output based on the second judgment result, that is, the height change of the continuous frame flame regions of interest contours is analyzed through the variance, when the variance is less than the second threshold value and the vertical coordinate of the current frame of the original flame image is greater than the third threshold value, that is, the height change of the continuous frame flame regions of interest contours is maintained within a certain range, the flame presents a state as shown in Figure 4 At this time, the flame soft and hard degree is judged again. In this embodiment, a monitoring region of the flame boundary is determined, for example, the monitoring region of the flame boundary isFigure 7 The white diagonal slash area represents the flame softness and hardness detection area, which includes a left monitoring area and a right monitoring area. The weighted average value of the pixel horizontal coordinates in the monitoring area contour is calculated to obtain the left boundary position horizontal coordinate and the right boundary position horizontal coordinate of the flame, i.e. the horizontal coordinate of the centroid of all contours in the monitoring area is calculated, and the centroid positions in the left and right areas are respectively weighted and averaged according to the number of pixels in each contour, and finally two left and right horizontal coordinates describing the flame boundary position are obtained. Finally, according to the flame boundary change range defined by the monitoring area, the distance between the above two horizontal coordinates is normalized, and a value ranging from 0 to 1 is output to quantitatively evaluate the softness and hardness of the current flame. By obtaining the original image of the flame in the converter steelmaking process and analyzing and processing the original image, the softness and hardness of the converter steelmaking flame are identified, which can quantitatively evaluate the softness and hardness of the flame in the converter steelmaking process in real time and efficiently. The evaluation result is returned to the automatic control equipment to realize automatic steelmaking. Through the monitoring of the softness and hardness of the flame, the frequency of splashing in the middle of the smelting process can be effectively avoided, the cost can be saved, the safety hidden danger can be reduced, and the carbon content in the molten steel in the later stage of smelting can also be assisted to monitor, the terminal carbon hit rate can be improved, the yield can be improved, and the smelting cycle can be shortened.

[0087] Please refer to Figure 8 , Figure 8 is a device block diagram for detecting the softness and hardness of the flame according to an example embodiment of the present application. The device can be applied to the implementation environment shown in Figure 1 , and is specifically configured in the computer equipment 103. The device can also be applied to other example implementation environments and is specifically configured in other equipment, and the implementation environment to which the device is applied is not limited in the present embodiment.

[0088] As Figure 8As shown, the exemplary flame softness and hardness detection device includes an acquisition module 801, a generation module 802, an extraction module 803, and a determination module 804. The acquisition module 801 is configured to acquire a plurality of flame images containing a flame area in a converter steelmaking process, denoted as original flame images. The generation module 802 is configured to convert the original flame images to generate binary images. The extraction module 803 is configured to perform morphological connection on the binary images, and perform contour extraction on the binary images after morphological connection to obtain a region of interest contour. The determination module 804 is configured to calculate the variance of a set of point coordinates in the region of interest contour, and determine the softness and hardness of the flame based on the variance. By taking the original image of the flame in the converter steelmaking process, and analyzing and processing the original image, the softness and hardness of the flame in the converter steelmaking process can be identified, the softness and hardness of the flame in the converter steelmaking process can be quantitatively evaluated in real time and efficiently, and the evaluation result is returned to the automatic control equipment to realize automatic steelmaking. Through monitoring the softness and hardness of the flame, the frequency of splashing in the middle of smelting can be effectively avoided, the cost can be saved, the safety hidden danger can be reduced, the carbon content in the molten steel in the later stage of smelting can be assisted to be monitored, the terminal carbon hit rate can be improved, the yield can be improved, and the smelting cycle can be shortened.

[0089] In an exemplary embodiment of the present application, the generation module 802 is configured to convert the original flame image from a first color space to a second color space to obtain a preprocessed flame image. The first color space is composed of red, green, and blue, and the second color space is composed of hue, saturation, and brightness. Based on a preset flame variation range, the upper limit value and the lower limit value of the mask are determined in the second color space, i.e., the upper limit value and the lower limit value of the mask are determined. The preprocessed flame image is compared with the upper limit value and the lower limit value of the mask respectively, and a binary image is generated according to the comparison result.

[0090] In an exemplary embodiment of the present application, the generation module 802 is configured to determine the preprocessed flame image as a white pixel value image if the hue of the preprocessed flame image is greater than the lower limit value of the hue in the second color space and less than the upper limit value of the hue in the second color space, the saturation is greater than the lower limit value of the saturation in the second color space and less than the upper limit value of the saturation in the second color space, and the brightness is greater than the lower limit value of the brightness in the second color space and less than the upper limit value of the brightness in the second color space. The preprocessed flame image is determined as a black pixel value image if the hue of the preprocessed flame image is less than or equal to the lower limit value of the hue in the second color space, the saturation is less than or equal to the lower limit value of the saturation in the second color space, and the brightness is less than or equal to the lower limit value of the brightness in the second color space. The preprocessed flame image is determined as a black pixel value image if the hue of the preprocessed flame image is greater than or equal to the upper limit value of the hue in the second color space, the saturation is greater than or equal to the upper limit value of the saturation in the second color space, and the brightness is greater than or equal to the upper limit value of the brightness in the second color space.

[0091] In an example embodiment of the present application, the extraction module 803 is configured to set a structure element; perform a morphological dilation operation on the structure element, and then perform a morphological erosion operation; perform a morphological connection on a target region of a binary image; obtain a closed region formed by horizontal coordinate values and vertical coordinate values of contour pixel points in a contour of a white part of the binary image, the closed region including an inner contour and an outer contour; remove the inner contour of the closed region, and remove contours in the outer contour that contain less than a first threshold of pixel points, to obtain a contour of a region of interest.

[0092] In an example embodiment of the present application, the extraction module 803 is configured to extract setpoint coordinates of a contour of a region of interest of a plurality of original flame images to form a setpoint coordinate group; calculate a variance of the setpoint coordinate group; if the variance is less than a second threshold and a vertical coordinate of a setpoint of a contour of a region of interest corresponding to a current original flame image is greater than a third threshold, then perform a secondary judgment on a soft and hard degree of the flame, and output a soft and hard degree recognition result of the flame based on a result of the secondary judgment; and if the variance is greater than or equal to the second threshold and the vertical coordinate of the setpoint of the contour of the region of interest corresponding to the current original flame image is less than or equal to the third threshold, then output the soft and hard degree recognition result of the flame.

[0093] In an example embodiment of the present application, the determination module 804 is configured to determine a monitoring region of a flame boundary, the monitoring region including a left monitoring region and a right monitoring region; perform a weighted average calculation on pixel point coordinates within a contour of the monitoring region to obtain a left boundary position coordinate of the flame and a right boundary position coordinate of the flame; and output the soft and hard degree recognition result of the flame based on the left boundary position coordinate of the flame and the right boundary position coordinate of the flame.

[0094] In an example embodiment of the present application, the determination module 804 is configured to output the soft and hard degree recognition result of the flame according to the left boundary position coordinate of the flame, the right boundary position coordinate of the flame, an upper limit of a right change of a left boundary of the flame, an upper limit of a left change of a right boundary of the flame, and a width of the flame in a binary image, and the soft and hard degree recognition result of the flame has:

[0095]

[0096] wherein, boundary left represents a coordinate of a left boundary position of the flame, boundary right represents a coordinate of a right boundary position of the flame, x l is an upper limit of a right change of a left boundary of the flame during smelting, x r is an upper limit of a left change of a right boundary of the flame during smelting, width is a width of the flame in a binary image, and res is a final result of the mapped soft and hard degree of the flame, wherein the smaller the res is, the softer the flame is, and the larger the res is, the harder the flame is.

[0097] The embodiment further provides an electronic device, comprising: one or more processors; a storage device configured to store one or more programs, which, when executed by the one or more processors, cause the electronic device to implement the method for detecting softness and hardness of a flame provided in any of the above embodiments.

[0098] Referring to Figure 8 , Figure 8 is a structural schematic diagram of an electronic device according to an example embodiment of the present application. It should be noted that Figure 8 The electronic device 800 shown is only an example, and should not impose any limitation on the functions and use range of the embodiments of the present application.

[0099] As Figure 9 shown, the electronic device 900 comprises a processor 901, a memory 902 and a communication bus 903; the communication bus 903 is configured to connect the processor 901 and the memory 902; the processor 901 is configured to execute a computer program stored in the memory 902, so as to implement the method of one or more of the above embodiments.

[0100] The embodiment further provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor of a computer, so that the computer executes the method for detecting softness and hardness of a flame as described above. The computer readable storage medium can be included in the electronic device described in the above embodiments, or can exist separately and not be assembled into the electronic device. The embodiment further provides a computer program product or a computer program, which comprises computer instructions stored in a computer readable storage medium. The processor of the computer device reads the computer instructions from the computer readable storage medium, and the processor executes the computer instructions, so that the computer device executes the method for detecting softness and hardness of a flame provided in any of the above embodiments.

[0101] The electronic device provided by the embodiment comprises a processor, a memory, a transceiver and a communication interface, the memory and the communication interface are connected with the processor and the transceiver and complete communication between each other, the memory is configured to store a computer program, the communication interface is configured to communicate, and the processor and the transceiver are configured to run the computer program, so that the electronic device executes each step of the above method.

[0102] In the embodiment, the memory can include a random access memory (RAM) and can also include a non-volatile memory, for example, at least one disk memory.

[0103] The processor described above can be a general processor, including a central processing unit (CPU), a network processor (NP), etc.; can also be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic device, a discrete gate or transistor logic device, a discrete hardware component.

[0104] The computer readable storage medium in the embodiment can be understood by those skilled in the art that all or part of the steps of the above-mentioned method embodiments can be completed by a computer program related hardware. The foregoing computer program can be stored in a computer readable storage medium. When the program is executed, the steps of the above-mentioned method embodiments are executed; and the foregoing storage medium includes ROM (read only memory), RAM (random access memory), magnetic disk or optical disk and various storage media that can store program codes.

[0105] The above-mentioned embodiments only exemplarily illustrate the principles and effects of the present application, and are not used to limit the present application. Any person skilled in the art can modify or change the above-mentioned embodiments without departing from the spirit and scope of the present application. Therefore, all equivalent modifications or changes made by those skilled in the art without departing from the spirit and technical idea of the present application should be covered by the claims of the present application.

Claims

1. A method for detecting the hardness of a flame, characterized in that, The method for detecting the hardness or softness of a flame includes: Multiple frames of flame images containing the flame region are acquired during the converter steelmaking process and recorded as the original flame images; The original flame image is converted to generate a binary image; Morphological concatenation is performed on the binary image, and contour extraction is performed on the morphologically concatenated binary image to obtain the contour of the region of interest. Calculate the variance of the set of coordinates of designated points in the region of interest contour, and determine the hardness of the flame based on the variance. Calculating the variance of the set of ordinates of designated points in the region of interest contour and determining the hardness of the flame based on the variance includes: extracting the coordinates of designated points in the region of interest contour from multiple frames of original flame images to form a set of designated point coordinates; calculating the variance of the set of designated point coordinates; if the variance is less than a second threshold and the coordinates of the designated points in the region of interest contour corresponding to the current frame of original flame image are greater than a third threshold, then a second determination of the hardness of the flame is performed, and a flame hardness identification result is output based on the second determination result; if the variance is greater than or equal to the second threshold and the coordinates of the designated points in the region of interest contour corresponding to the current frame of original flame image are less than or equal to the third threshold, then a flame hardness identification result is output; outputting the flame hardness identification result based on the second determination result includes: The monitoring area for the flame boundary is determined, including the left and right monitoring areas. A weighted average of the pixel coordinates within the monitoring area contour is calculated to obtain the left and right boundary coordinates of the flame. Based on these coordinates, the flame hardness / softness level is identified. This identification includes: Based on the coordinates of the left and right edges of the flame, the upper limit of the change of the left edge to the right, the upper limit of the change of the right edge to the left, and the width of the flame in the binary image, the flame hardness / softness recognition result is output, which includes: in, The coordinates represent the position of the left boundary of the flame. The coordinates represent the position of the right boundary of the flame. This represents the upper limit of the change in the left boundary of the flame to the right during smelting. This represents the upper limit of the leftward change of the right boundary of the flame during smelting. The width of the flame in the binary image. The final result is the mapped hardness or softness of the flame. The smaller the value, the softer the flame. The larger the value, the harder the flame.

2. The method for detecting the hardness of a flame according to claim 1, characterized in that, The process of converting the original flame image to generate a binary image includes: The original flame image is converted from a first color space to a second color space to obtain a preprocessed flame image; wherein the first color space consists of red, green and blue, and the second color space consists of hue, saturation and brightness; Based on the preset flame variation range, the upper limit value of hue, the upper limit value of saturation, the upper limit value of brightness, the lower limit value of hue, the lower limit value of saturation, and the lower limit value of brightness are determined in the second color space, that is, the upper limit value and the lower limit value of the mask are determined. The preprocessed flame image is compared with the upper limit value and the lower limit value of the mask, and a binary image is generated based on the comparison results.

3. The method for detecting the hardness of a flame according to claim 2, characterized in that, The original flame image is converted from a first color space to a second color space to obtain a preprocessed flame image, including: If the hue of the preprocessed flame image is greater than the lower limit of the hue value in the second color space and less than the upper limit of the hue value in the second color space, the saturation is greater than the lower limit of the saturation value in the second color space and less than the upper limit of the saturation value in the second color space, and the brightness is greater than the lower limit of the brightness value in the second color space and less than the upper limit of the brightness value in the second color space, then the preprocessed flame image is determined to be a white pixel value image. If the hue of the preprocessed flame image is less than or equal to the lower limit of the hue in the second color space, the saturation is less than or equal to the lower limit of the saturation in the second color space, and the brightness is less than or equal to the lower limit of the brightness in the second color space, then the preprocessed flame image will be determined as a black pixel value image. If the hue of the preprocessed flame image is greater than or equal to the upper limit of the hue value in the second color space, the saturation is greater than or equal to the upper limit of the saturation value in the second color space, and the brightness is greater than or equal to the upper limit of the brightness value in the second color space, then the preprocessed flame image will be determined as a black pixel value image.

4. The method for detecting the hardness of a flame according to any one of claims 1 to 3, characterized in that, The binary image is morphologically concatenated, and the concatenated binary image is then contour extracted to obtain the region of interest contour, including: Set up the structural element; The structuring element is subjected to morphological dilation and then morphological erosion to perform morphological connection on the target region of the binary image. Obtain the closed region formed by the x-coordinate and y-coordinate values ​​of the contour pixels in the white part of the binary image, wherein the closed region includes an inner contour and an outer contour. Remove the inner contour of the closed region and remove the outer contour containing fewer than a first threshold pixels to obtain the region of interest contour.

5. A device for detecting the hardness of a flame, characterized in that, The method for detecting the hardness of a flame as described in any one of claims 1-4, wherein the apparatus for detecting the hardness of a flame comprises: The acquisition module is used to acquire multiple frames of flame images containing the flame region during the converter steelmaking process, which are denoted as the original flame images; The generation module is used to convert the original flame image to generate a binary image; The extraction module is used to perform morphological concatenation on the binary image, and to extract the contour of the morphologically concatenated binary image to obtain the contour of the region of interest. The determination module is used to calculate the variance of the coordinate group of set points in the outline of the region of interest, and determine the hardness or softness of the flame based on the variance.

6. An electronic device, characterized in that, The electronic device includes: One or more processors; A storage device for storing one or more programs, which, when executed by the one or more processors, cause the electronic device to implement the method for detecting the hardness or softness of a flame as described in any one of claims 1 to 4.

7. A medium, characterized in that, It stores a computer program that, when executed by the computer's processor, causes the computer to perform the method for detecting the hardness of a flame as described in any one of claims 1 to 4.

Images