Coke oven column deformation monitoring method and system

Abstract

The invention provides a coke oven column deformation monitoring method and system, and relates to the technical field of coke oven measurement, and the coke oven column deformation monitoring method comprises the following steps: obtaining an image sequence of a marked area in the side surface of an oven column; gray processing is carried out on the image sequence to obtain a gray image sequence, and the gray image sequence comprises a plurality of gray images arranged according to a time sequence; selecting the same region as a target sub-region in each grayscale image, wherein the target sub-region has a plurality of adjacent sub-regions adjacent to the target sub-region; whether the furnace column is deformed or not is judged according to the gray value of the target sub-region in an initial gray image and the gray value of the target sub-region in a subsequent gray image, and the initial gray image is the first gray image in the gray image sequence. According to the invention, the deformation measurement precision of the coke oven column is improved, and the operation is simple.

Description

Technical Field

[0001] The invention relates to the technical field of coke oven measurement, and in particular to a method and system for monitoring the deformation of a coke oven column. Background Art

[0002] Coke oven columns are key components supporting coke oven production and are known as the "backbone" of the coke oven. Due to the complex and variability of on-site production conditions (such as long-term high-temperature baking) and maintenance of the coke oven, the performance of the columns is constantly changing. In addition, since the coke oven columns are composed of rigid components, under normal operating conditions, the columns deform under certain temperatures and pressures. Normally, the deformation can be restored by adjusting related facilities to ensure the normal operation of the coke oven columns. However, if it is not restored in time, it will often cause permanent deformation of the columns, thereby greatly weakening the protective function of the columns. Therefore, it is of great significance to timely grasp the deformation of the columns to facilitate adjustment and maintenance.

[0003] For coke oven columns, traditional displacement measurement methods include contact measurement, which is mainly simple manual measurement or mechanically assisted manual reading. The traditional method is relatively simple and stable to operate, but it is easily affected by the material of the object or the external environment, resulting in inaccurate measurement. It is also difficult to accurately determine whether the column is deformed. The non-contact three-dimensional displacement measurement method mainly uses optical principles for measurement. The main principle is to use laser to scan, aim, measure, and solve the three-dimensional coordinates of the object to calculate the three-dimensional displacement. This method greatly makes up for the shortcomings of the contact method, but the operation is more complicated. Summary of the invention

[0004] The present invention aims to solve the problem that, during the process of determining and measuring the deformation of a coke oven column, it is difficult to accurately determine whether the column is deformed, and the traditional deformation measurement is inaccurate and / or the operation is relatively complicated.

[0005] In order to solve the above problems, in a first aspect, the present invention provides a coke oven column deformation monitoring method, comprising: Acquire a sequence of images of the marked area in the side of the furnace column; Performing grayscale processing on the image sequence to obtain a grayscale image sequence; If it is determined that the furnace column is deformed, a target adjacent subregion is searched based on a plurality of adjacent subregions of the target subregion in the subsequent grayscale image, wherein the target adjacent subregion represents an adjacent region in the subsequent grayscale image whose grayscale correlation coefficient with the target subregion in the initial grayscale image is less than a preset coefficient threshold and is locally minimum, and the grayscale correlation coefficient is used to characterize the grayscale difference between the two regions; The actual offset of the furnace column is determined according to the pixel coordinate difference and pixel scale between each of the target adjacent sub-regions and the target sub-region.

[0006] Optionally, judging whether the furnace column is deformed according to the grayscale value of the target sub-region in the initial grayscale image and the grayscale value in the subsequent grayscale image includes: Determine the grayscale correlation coefficient between the target subregion in the initial grayscale image and the target subregion in the subsequent grayscale image according to the grayscale value of the target subregion in the initial grayscale image and the grayscale value of the target subregion in the subsequent grayscale image, obtain a first grayscale correlation coefficient value, and determine whether the first grayscale correlation coefficient value is greater than 0; If the first grayscale correlation coefficient value is greater than 0, it is determined that the furnace column is deformed, and if the first grayscale correlation coefficient value is equal to 0, it is determined that the furnace column is not deformed.

[0007] Optionally, if it is determined that the furnace column is deformed, searching for a target adjacent sub-region based on a plurality of adjacent sub-regions of the target sub-region in the subsequent grayscale image comprises: Based on the subsequent grayscale image, searching for the adjacent subregion with the smallest grayscale correlation coefficient with the target subregion in the initial grayscale image among the plurality of adjacent subregions spaced from the target subregion by a preset pixel step, and taking the adjacent subregion as the second subregion, and taking the target subregion in the subsequent grayscale image as the first subregion; A search step is performed to determine the Nth sub-region, where N is a positive integer greater than or equal to 3, and the search step includes: taking the direction from the N-2th sub-region to the N-1th sub-region as the search direction, taking the N-1th sub-region as the starting point, and determining the adjacent sub-region by a preset pixel step interval along the search direction, and determining whether the grayscale correlation coefficient between the determined adjacent sub-region and the target sub-region in the initial grayscale image is less than the grayscale correlation coefficient between the N-1th sub-region and the target sub-region in the initial grayscale image, and if so, taking the determined adjacent sub-region as the Nth sub-region; if not, based on The N-1th sub-region is rotated clockwise or counterclockwise with the direction from the N-2th sub-region to the N-1th sub-region as the rotation reference, and each rotation angle is increased by twice the preset angle compared with the original rotation angle, so as to determine the corresponding new search direction each time, and locate the adjacent sub-region along the new search direction at intervals of preset pixel steps until the grayscale correlation coefficient between the located adjacent sub-region and the target sub-region in the initial grayscale image is less than the grayscale correlation coefficient between the N-1th sub-region and the target sub-region in the initial grayscale image, and the located adjacent sub-region is used as the Nth sub-region; Determine whether the grayscale correlation coefficient between the Nth sub-region and the target sub-region in the initial grayscale image is less than a preset coefficient threshold; If yes, the Nth sub-region is determined to be the target adjacent sub-region; if no, N+1 is returned to execute the search step.

[0008] Optionally, based on the N-1th sub-region, the direction from the N-2th sub-region to the N-1th sub-region is used as a rotation reference, and the rotation is performed clockwise or counterclockwise, and each rotation angle is increased by twice the preset angle compared to the original rotation angle, so as to determine a new corresponding search direction each time, and locate the adjacent sub-region along the new search direction at intervals of preset pixel steps until the grayscale correlation coefficient between the located adjacent sub-region and the target sub-region in the initial grayscale image is less than the grayscale correlation coefficient between the N-1th sub-region and the target sub-region in the initial grayscale image, and the located adjacent sub-region is used as the Nth sub-region, including: taking the direction from the N-2th sub-region to the N-1th sub-region as a reference, rotating clockwise or counterclockwise by twice the preset angle, so as to determine a new corresponding search direction each time, and locating the adjacent sub-region along the new search direction at intervals of preset pixel steps, and judging whether the grayscale correlation coefficient between the located adjacent sub-region and the target sub-region in the initial grayscale image is less than the grayscale correlation coefficient between the N-1th sub-region and the target sub-region in the initial grayscale image. The grayscale correlation coefficient between the sub-region and the target sub-region in the initial grayscale image is determined based on this. If yes, the Nth sub-region is determined based on this. If not, the direction of the N-2th sub-region pointing to the N-1th sub-region is used as a reference, and the preset angle is rotated twice in the same direction as the original rotation direction to determine the new corresponding search direction each time. The neighboring sub-region is located at intervals of preset pixel steps along the new search direction, and it is determined whether the grayscale correlation coefficient between the located neighboring sub-region and the target sub-region in the initial grayscale image is less than the grayscale correlation coefficient between the N-1th sub-region and the target sub-region in the initial grayscale image. If yes, the Nth sub-region is determined based on this. If not, the grayscale correlation coefficient is rotated three times in the same direction as the original rotation direction, and so on, until the grayscale correlation coefficient between the located neighboring sub-region and the target sub-region in the initial grayscale image is less than the grayscale correlation coefficient between the N-1th sub-region and the target sub-region in the initial grayscale image, and the Nth sub-region is determined accordingly.

[0009] Optionally, searching, based on the subsequent grayscale image, among the plurality of adjacent subregions spaced from the target subregion by a preset pixel step, the adjacent subregion having the smallest grayscale correlation coefficient with the target subregion in the initial grayscale image comprises: Based on the subsequent grayscale image, based on multiple directions centered on the target sub-region and radially evenly expanded, the neighboring sub-regions that are spaced apart from the target sub-region by the preset pixel step are determined one by one, and among the determined neighboring sub-regions, the neighboring sub-region with the smallest grayscale correlation coefficient with the target sub-region in the initial grayscale image is determined.

[0010] Optionally, it also includes: If the target adjacent sub-region still cannot be found under the preset pixel step size, the preset pixel step size is gradually reduced, and the target adjacent sub-region is tried again to be found until the target adjacent sub-region is found.

[0011] Optionally, the grayscale correlation coefficient is determined according to the following formula: ; in, represents the grayscale correlation coefficient, Indicates the first pixel coordinate in the target sub-area The corresponding gray value, Indicates the coordinates of the second pixel in the adjacent sub-area Corresponding grayscale value, the target sub-region or the adjacent sub-region is a square, and M represents half of the side length of the target sub-region or the adjacent sub-region.

[0012] Optionally, determining the actual offset of the furnace column according to the pixel coordinate difference and pixel scale between each of the target adjacent sub-regions and the target sub-region comprises: The lateral offset is determined according to the ratio of the lateral pixel coordinate difference between the target adjacent sub-region and the target sub-region to the pixel scale, and the longitudinal offset is determined according to the ratio of the longitudinal pixel coordinate difference between the target adjacent sub-region and the target sub-region to the pixel scale, wherein the pixel coordinate difference includes the lateral pixel coordinate difference and the longitudinal pixel coordinate difference, and the actual offset includes the lateral offset and the longitudinal offset.

[0013] Optionally, before grayscale processing is performed on the image sequence, the method further includes: A filtering process is performed on the image sequence.

[0014] The present invention obtains an image sequence of a marked area on the side of a furnace column, performs grayscale processing on the image sequence, and obtains a grayscale image sequence. After selecting the same area as a target sub-area in each of the grayscale images, it is judged whether the furnace column is deformed according to the grayscale value of the target sub-area in the initial grayscale image and the grayscale value in the subsequent grayscale image. Since the difference in grayscale values ​​can reflect the changes in the shape and structure of an object, it is possible to more accurately judge whether the furnace column is deformed by means of the correlation between the grayscale values ​​of the target sub-area in the initial grayscale image and the subsequent grayscale image. If it is determined that the furnace column is deformed, a target adjacent sub-area is searched based on a plurality of adjacent sub-areas of the target sub-area in the subsequent grayscale image, and the target adjacent sub-area represents an adjacent sub-area in the subsequent grayscale image whose grayscale correlation coefficient with the target sub-area in the initial grayscale image is less than a preset coefficient threshold and is locally minimum, that is, from the target sub-area to the target adjacent sub-area The domain has experienced a deformation variable that first increases and then decreases, and the target adjacent sub-region is the location of the minimum position after deformation (that is, the target adjacent sub-region reflects the location of the farthest position of the initial target sub-region after deformation). Therefore, according to the pixel coordinate difference and pixel scale between each of the target adjacent sub-regions and the target sub-region, the actual offset of the furnace column corresponding to the image sequence can be calculated, and the calculated actual offset can more accurately reflect the maximum deformation of the target sub-region at a certain moment (actually reflects the deformation of the furnace column), thereby improving the measurement accuracy. In addition, since the image sequence is sorted in chronological order, the actual offset under the series of target adjacent sub-regions can reflect the deformation of the target sub-region of the furnace column of the coke oven over time, and can include peak values ​​and valley values, comprehensively reflecting the deformation of the core area of ​​the coke oven, and realizing all-weather real-time monitoring of the coke oven column. Compared with traditional measurements, the operation of the present invention is relatively simple and saves manpower.

[0015] In a second aspect, the present invention further provides a coke oven column deformation monitoring system, which uses the coke oven column deformation monitoring method as described in any one of the above items, comprising: An acquisition module, used for acquiring an image sequence of a marked area in a side surface of a furnace column; A pre-processing module, used for performing grayscale processing on the image sequence to obtain a grayscale image sequence, wherein the grayscale image sequence includes a plurality of grayscale images arranged in chronological order; A selection module, used for selecting a same area in each of the grayscale images as a target sub-area, wherein the target sub-area has a plurality of adjacent sub-areas adjacent to the target sub-area; A deformation judgment module, used for judging whether the furnace column is deformed according to the grayscale value of the target sub-area in the initial grayscale image and the grayscale value in the subsequent grayscale image, wherein the initial grayscale image is the first grayscale image in the grayscale image sequence, and the subsequent grayscale image is any grayscale image after the first grayscale image in the grayscale image sequence; a target sub-region determination module, configured to search for a target adjacent sub-region based on a plurality of adjacent sub-regions of the target sub-region in the subsequent grayscale image if it is determined that the furnace column is deformed, wherein the target adjacent sub-region represents an adjacent region in the subsequent grayscale image whose grayscale correlation coefficient with the target sub-region in the initial grayscale image is less than a preset coefficient threshold and is locally minimum, and the grayscale correlation coefficient is used to characterize the grayscale difference between the two regions; The offset determination module is used to determine the actual offset of the furnace column according to the pixel coordinate difference and pixel scale between each of the target adjacent sub-regions and the target sub-region.

[0016] In a third aspect, the present invention provides an electronic device, including a memory and a processor; The memory is used to store computer programs; The processor is used to implement the coke oven column deformation monitoring method as described in the first aspect when executing the computer program.

[0017] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the method for monitoring the deformation of a coke oven column as described in the first aspect is implemented.

[0018] The coke oven column deformation monitoring system, electronic device and computer-readable storage medium provided by the present invention have the same beneficial effects as the coke oven column deformation monitoring method relative to the prior art, which will not be repeated here. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 A schematic flow chart of a coke oven column deformation monitoring method according to an embodiment of the present invention is shown; Figure 2 A flow chart of searching for target adjacent sub-regions in an embodiment of the present invention is shown; Figure 3 A schematic structural diagram of a coke oven column deformation monitoring system according to an embodiment of the present invention is shown; Figure 4 A schematic diagram of the structure of an electronic device in an embodiment of the present invention is shown. DETAILED DESCRIPTION

[0020] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0021] It should be noted that, in the present invention, relational terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0022] In the description of this specification, the description with reference to the terms "embodiment", "one embodiment" and "one implementation" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or implementation are included in at least one embodiment or implementation of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or implementation. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or implementations in a suitable manner.

[0023] Reference Figure 1 As shown, an embodiment of the present invention provides a method for monitoring the deformation of a coke oven column.

[0024] The coke oven column deformation monitoring method comprises: S100: Acquire an image sequence of the marked area in the side of the furnace column.

[0025] Specifically, an area is selected on the side of the furnace column and marked. The marking can be done by spraying the area selected on the side of the furnace column. The area size can be set to 100mm×100mm. After marking, a camera is used to collect images of the marked area to obtain a series of images, that is, an image sequence. The image collection is carried out under normal working conditions of the coke oven, and the camera is at a fixed object distance position. In order to obtain higher measurement accuracy, the method for selecting the camera resolution is as follows: Resolution = (field of view height / accuracy) × (field of view width / accuracy) × λ. The measurement accuracy in this application is 1 mm. λ represents the pixel value contained in 1 precision size. In this application, the value of λ is 3 (indicating that 1 mm contains 3 pixel values).

[0026] S200: Performing grayscale processing on the image sequence to obtain a grayscale image sequence, wherein the grayscale image sequence includes a plurality of grayscale images arranged in chronological order.

[0027] Specifically, image grayscale conversion is the process of converting a color image into a grayscale image, that is, converting the image from the RGB three-channel color space to a single-channel grayscale space. Common grayscale conversion generally adopts the weighted average method, Gray = 0.299* R + 0.587 * G + 0.114 * B, where R, G, and B represent the pixel values ​​of the red, green, and blue channels of the image, respectively. Since the image sequence is shot in chronological order, the grayscale image sequence is also arranged in chronological order.

[0028] S300: Selecting a same region in each of the grayscale images as a target sub-region, wherein the target sub-region has a plurality of adjacent sub-regions.

[0029] Specifically, the target sub-region in each grayscale image is the same region. The purpose is to determine the deformation of the target sub-region and thus determine whether the furnace column is deformed. Exemplarily, a square region of 60×60 pixels is selected as the target sub-region P.

[0030] S400: Determine whether the furnace column is deformed according to the grayscale value of the target sub-area in the initial grayscale image and the grayscale value in the subsequent grayscale image, wherein the initial grayscale image is the first grayscale image in the grayscale image sequence, and the subsequent grayscale image is any grayscale image after the first grayscale image in the grayscale image sequence.

[0031] Specifically, the grayscale value is a numerical value that represents the brightness or intensity of each pixel in the image. In a grayscale image, 0 usually represents black, 255 represents white, and other values ​​represent different grayscales. When an object (such as a furnace column) is deformed, the illumination and reflection characteristics of its surface may change, causing the grayscale value of the area to change. Therefore, it is possible to determine whether the furnace column is deformed based on a certain correlation between the initial grayscale image and the subsequent grayscale image.

[0032] S500: If it is determined that the furnace column is deformed, a target adjacent sub-region is searched based on a plurality of adjacent sub-regions of the target sub-region in the subsequent grayscale image, wherein the target adjacent sub-region represents an adjacent region in the subsequent grayscale image whose grayscale correlation coefficient with the target sub-region in the initial grayscale image is less than a preset coefficient threshold and is locally minimum, and the grayscale correlation coefficient is used to characterize the grayscale difference between the two regions.

[0033] Specifically, the grayscale correlation coefficient can indicate the grayscale difference between a neighboring sub-region and the target sub-region at the initial moment (the target sub-region before deformation). If the deformation amount is larger, the grayscale correlation coefficient is larger. On the contrary, if the deformation amount is smaller, the grayscale correlation coefficient is smaller. The target sub-region is generally the core area of ​​the coke oven column to be monitored. When it is preliminarily determined that it has been deformed, based on the image sequence, with the target sub-region as the center, a neighboring sub-region with a grayscale correlation coefficient less than a preset coefficient threshold is determined in multiple neighboring sub-regions on the periphery of the target sub-region. This neighboring sub-region is the target neighboring sub-region. The target neighboring sub-region is used to characterize that the grayscale correlation coefficient between it and the target sub-region at the initial moment is small enough, and the grayscale correlation coefficient of this target neighboring sub-region is the local minimum. For example, there are multiple neighboring sub-regions in the periphery, and the target neighboring sub-region has the smallest grayscale correlation coefficient among the five neighboring sub-regions in the local range. The target neighboring sub-region indicates that it is very likely to be the farthest position after the deformation of the initial target sub-region, that is, from the target sub-region to the target neighboring sub-region, the deformation variable first increases and then decreases, and the target sub-region is the minimum deformation position (that is, the farthest deformation position). Therefore, based on the position of the target neighboring sub-region and the target sub-region, the deformation (or offset) of the target sub-region can be calculated.

[0034] S600: Determine the actual offset of the furnace column according to the pixel coordinate difference and pixel scale between the target adjacent sub-region and the target sub-region.

[0035] Specifically, according to the above description, after determining the target neighboring sub-region, since the actual distance can be determined through pixel coordinates and pixel scale, the offset actually corresponding to the pixel coordinate difference can be determined based on the pixel coordinate difference and pixel scale, that is, the actual offset, which is also the deformation offset.

[0036] When this embodiment is applied in practice, an image sequence of the marked area on the side of the furnace column is obtained, and grayscale processing is performed on the image sequence to obtain a grayscale image sequence. After the same area is selected as the target sub-area in each of the grayscale images, it is judged whether the furnace column is deformed according to the grayscale value of the target sub-area in the initial grayscale image and the grayscale value in the subsequent grayscale image. Since the difference in grayscale values ​​can reflect the changes in the shape and structure of an object, it is possible to more accurately judge whether the furnace column is deformed with the help of the correlation between the grayscale values ​​of the target sub-area in the initial grayscale image and the subsequent grayscale image. If it is determined that the furnace column is deformed, a target adjacent sub-area is searched based on multiple adjacent sub-areas of the target sub-area in the subsequent grayscale image, and the target adjacent sub-area represents an adjacent sub-area in the subsequent grayscale image whose grayscale correlation coefficient with the target sub-area in the initial grayscale image is less than a preset coefficient threshold and is locally minimum, that is, from the target sub-area to the target The adjacent sub-region experiences a deformation amount that first increases and then decreases, and the target adjacent sub-region is the location of the minimum position after deformation (that is, the target adjacent sub-region reflects the farthest position of the initial target sub-region after deformation). Therefore, according to the pixel coordinate difference and pixel scale between each of the target adjacent sub-regions and the target sub-region, the actual offset of the furnace column corresponding to the image sequence can be calculated, and the calculated actual offset can more accurately reflect the maximum deformation of the target sub-region at a certain moment (actually reflects the deformation of the furnace column), thereby improving the measurement accuracy. In addition, since the image sequence is sorted in chronological order, the actual offset under the series of target adjacent sub-regions can reflect the deformation of the target sub-region of the furnace column of the coke oven over time, and can include peak values ​​and valley values, comprehensively reflecting the deformation of the core area of ​​the coke oven, and realizing all-weather real-time monitoring of the coke oven column. Compared with traditional measurements, the operation of the present invention is relatively simple and saves manpower.

[0037] The present invention can monitor specific furnace columns during the coke oven production process, and can not only monitor whether the furnace columns are deformed, but also monitor their displacement. It is suitable for high temperature conditions in coke oven production, and uses a non-contact, all-weather measurement method with the characteristics of high measurement accuracy and simple operation.

[0038] As an optional embodiment of the present invention, judging whether the furnace column is deformed according to the grayscale value of the target sub-region in the initial grayscale image and the grayscale value in the subsequent grayscale image includes: According to the grayscale value of the target sub-region in the initial grayscale image and the grayscale value in the subsequent grayscale image, the grayscale correlation coefficient between the target sub-region in the initial grayscale image and the target sub-region in the subsequent grayscale image is determined to obtain a first grayscale correlation coefficient value, and it is determined whether the first grayscale correlation coefficient value is greater than 0.

[0039] If the first grayscale correlation coefficient value is greater than 0, it is determined that the furnace column is deformed, and if the first grayscale correlation coefficient value is equal to 0, it is determined that the furnace column is not deformed.

[0040] Specifically, if at the same center position (the position of the target sub-region), the grayscale correlation coefficient of the target sub-region at the initial moment and the subsequent moment is equal to 0, it means that the grayscale of the pixel is almost unchanged, indicating that the center position has not deformed, that is, the target sub-region has not deformed. Conversely, when the grayscale correlation coefficient of the target sub-region at the initial moment and the subsequent moment is greater than 0, it means that the grayscale difference of the pixel is large, and it is determined that the furnace column has been deformed.

[0041] like Figure 2 As shown, as an optional embodiment of the present invention, if it is determined that the furnace column is deformed, searching for the target adjacent sub-region based on the multiple adjacent sub-regions of the target sub-region in the subsequent grayscale image includes: Based on the subsequent grayscale image, among the multiple adjacent sub-regions that are spaced from the target sub-region by a preset pixel step, search for the adjacent sub-region with the smallest grayscale correlation coefficient with the target sub-region in the initial grayscale image, and use it as the second sub-region, and use the target sub-region in the subsequent grayscale image as the first sub-region.

[0042] Specifically, in the first step, for each subsequent grayscale image, the neighboring subregion with the smallest grayscale correlation coefficient around the target subregion is first determined. For the convenience of subsequent description, the target subregion is taken as the first subregion, and the neighboring subregion with the smallest grayscale correlation coefficient around the determined target subregion is taken as the second subregion. The second subregion is generally determined based on the first subregion (target subregion) as the center and multiple neighboring subregions uniformly distributed on the outer layer. The grayscale correlation coefficients of multiple outer neighboring subregions (and the target subregion in the initial grayscale image) are compared, and the smallest one is selected as the second subregion. At this time, the direction from the first region to the second region is determined, and the search continues along this direction. Exemplarily, in Figure 2 In , P is the target sub-region (the first sub-region), and the region corresponding to 1 (region 1) is the second sub-region.

[0043] A search step is performed to determine the Nth sub-region, where N is a positive integer greater than or equal to 3, and the search step includes: taking the direction from the N-2th sub-region to the N-1th sub-region as the search direction, taking the N-1th sub-region as the starting point, and determining the adjacent sub-region along the search direction at intervals of a preset pixel step size, and determining whether the grayscale correlation coefficient between the determined adjacent sub-region and the target sub-region in the initial grayscale image is less than the grayscale correlation coefficient between the N-1th sub-region and the target sub-region in the initial grayscale image; if so, taking the determined adjacent sub-region as the Nth sub-region. If it is not less than, based on the N-1th sub-region, take the direction from the N-2th sub-region to the N-1th sub-region as the rotation reference, rotate clockwise or counterclockwise, and the rotation angle each time is doubled by the preset angle than the original rotation angle, so as to determine the corresponding new search direction each time, and locate the adjacent sub-region at intervals of preset pixel step along the new search direction until the grayscale correlation coefficient between the located adjacent sub-region and the target sub-region in the initial grayscale image is less than the grayscale correlation coefficient between the N-1th sub-region and the target sub-region in the initial grayscale image, and the located adjacent sub-region is used as the Nth sub-region.

[0044] Specifically, if the position of the Nth sub-region is described by moving one preset pixel step each time, the Nth sub-region determined next time is determined and calculated based on the N-1th sub-region and the N-2th sub-region, and the grayscale correlation coefficient of the N-1th sub-region (and the grayscale correlation coefficient between the target sub-region in the initial grayscale image, hereinafter referred to as the grayscale correlation coefficient) is used to determine whether the determined adjacent sub-region is used as the Nth sub-region. If the grayscale correlation coefficient of the N-1th sub-region is small, the determined adjacent sub-region can be used as the Nth sub-region. N sub-regions, and the N-2 sub-region is used to determine the search direction of the N sub-region together with the N-1 sub-region. It preferentially selects the direction from the N-2 sub-region to the N-1 sub-region. Then, after searching a neighboring sub-region after a preset step length, the grayscale correlation coefficient of the neighboring sub-region is determined. The grayscale correlation coefficient of the N sub-region to be searched next time should be smaller than that of the N-1 sub-region. Otherwise, the neighboring sub-region searched cannot be used as the N sub-region. When it is not less than, it is rotated by a preset angle based on the original direction. For example, Figure 2 When the grayscale correlation coefficient C of the adjacent area corresponding to 1'' is found P1'' Grayscale correlation coefficient C of the adjacent area corresponding to 1' P1' , that is, C P1'' >C P1', then return to the 1' area, and determine 8 directions based on the 1' area, and then go clockwise to the adjacent direction of the direction from 1' to 1'' (this is also in line with the characteristic that the deformation of the target sub-area on the side of the furnace column is generally directional), that is, from 1' to the direction corresponding to 2'', and calculate C P2'' Is it less than C? P1'， If yes, it is considered that the search of the Nth sub-region is successful. Otherwise, the direction is continued to rotate, and then the direction corresponding to 1' points to 3'' is rotated, and so on. In this way, the Nth sub-region is determined.

[0045] Whether to continue searching the N+1th sub-region depends on the size relationship between the Nth sub-region and the preset coefficient threshold, which will be described in the next step.

[0046] In simple terms, the grayscale correlation coefficient of the neighboring sub-region searched next time should be smaller than the grayscale correlation coefficient of the neighboring sub-region searched last time, so that the requirement of the target neighboring sub-region being the local minimum can be met.

[0047] And after determining the adjacent sub-region (at least starting from the second sub-region), it is necessary to determine whether the corresponding grayscale correlation coefficient is less than or equal to the preset coefficient threshold. If not, continue to determine the adjacent sub-region of the next layer and calculate the preset coefficient threshold.

[0048] It is determined whether the grayscale correlation coefficient between the Nth sub-region and the target sub-region in the initial grayscale image is less than a preset coefficient threshold.

[0049] Specifically, the conditions that the target neighboring sub-region needs to meet are: it satisfies the condition that the grayscale correlation coefficient is relatively reduced (regarded as a local minimum) and is less than or equal to the preset coefficient threshold. In simple terms, the target neighboring sub-region also needs to satisfy the condition that its grayscale correlation coefficient is small enough (close to 0). At this time, the target neighboring sub-region is considered to be the position where the target sub-region is first deformed and then approaches the position where the deformation is recovered (that is, the farthest deformation position).

[0050] If yes, then the Nth sub-region is determined to be the target adjacent sub-region. If no, then N+1 is returned to perform the search step.

[0051] Specifically, when the Nth sub-region satisfies the grayscale correlation coefficient being less than the preset coefficient threshold, it is considered that the target adjacent sub-region is found. If not, the search step is returned to continue searching for the N+1th sub-region. The steps for searching the N+1th sub-region are the same as those for searching the Nth sub-region, except that N becomes N+1.

[0052] exist Figure 2In the figure, the preset angle is selected as 45 degrees, that is, for a certain N sub-region, there are 8 defined directions around it, and these 8 directions are evenly distributed radially with the N sub-region as the center.

[0053] In summary, the search direction and strategy are adjusted according to the results of each search, which improves the flexibility and effectiveness of the search and avoids possible local optimal solutions. By introducing preset angles and grayscale correlation coefficient thresholds, it can adapt to different conditions and changing image features. This method clarifies the iterative relationship and judgment criteria from the N-2th sub-region to the N-1th sub-region, and then to the Nth sub-region, forming a complete search mechanism, gradually narrowing the search range (from the 1st region to the Nth region), and using the results of the previous search to guide the next search, reducing unnecessary calculations and repeated search costs, which can ensure more accurate identification and tracking of target adjacent sub-regions.

[0054] like Figure 2As shown, as an optional embodiment of the present invention, based on the N-1th sub-region, the direction from the N-2th sub-region to the N-1th sub-region is used as a rotation reference, and the rotation is performed clockwise or counterclockwise, and each rotation angle is doubled by the preset angle compared to the original rotation angle, so as to determine the corresponding new search direction each time, and locate the adjacent sub-region along the new search direction at intervals of preset pixel steps until the grayscale correlation coefficient between the located adjacent sub-region and the target sub-region in the initial grayscale image is less than the grayscale correlation coefficient between the N-1th sub-region and the target sub-region in the initial grayscale image. Grayscale correlation coefficient, and taking the located adjacent sub-region as the Nth sub-region includes: taking the direction from the N-2th sub-region to the N-1th sub-region as a reference, rotating by one times the preset angle clockwise or counterclockwise to determine the corresponding new search direction each time, locating the adjacent sub-region along the new search direction with a preset pixel step interval, judging whether the grayscale correlation coefficient between the located adjacent sub-region and the target sub-region in the initial grayscale image is less than the grayscale correlation coefficient between the N-1th sub-region and the target sub-region in the initial grayscale image, and if so, determining the Nth sub-region accordingly. If not, then taking the direction from the N-2 sub-region to the N-1 sub-region as a reference, rotate twice the preset angle in the same direction as the original rotation direction to determine the corresponding new search direction each time, and locate the adjacent sub-region at intervals of a preset pixel step along the new search direction, and determine whether the grayscale correlation coefficient between the located adjacent sub-region and the target sub-region in the initial grayscale image is less than the grayscale correlation coefficient between the N-1 sub-region and the target sub-region in the initial grayscale image. If so, determine the Nth sub-region accordingly; if not, rotate three times the preset angle in the same direction as the original rotation direction, and so on, until the grayscale correlation coefficient between the located adjacent sub-region and the target sub-region in the initial grayscale image is less than the grayscale correlation coefficient between the N-1 sub-region and the target sub-region in the initial grayscale image, and thereby determine the Nth sub-region.

[0055] Specifically, when searching the Nth sub-region, if the grayscale correlation coefficient of the adjacent sub-region searched next time does not decrease relative to the grayscale correlation coefficient of the N-1th sub-region, it is necessary to adjust the search direction, and when adjusting the search direction, the N-1th sub-region is taken as the center, the direction of the N-2 sub-region pointing to the N-1th sub-region is taken as the reference, and it is rotated clockwise or counterclockwise by one times the preset angle. After the first rotation, if the grayscale correlation coefficient of the adjacent sub-region is not found to be smaller than the grayscale correlation coefficient of the N-1th sub-region, the N-1th sub-region is taken as the center, the direction of the N-2 sub-region pointing to the N-1 sub-region is taken as the reference, and it is rotated clockwise or counterclockwise by 2 times the preset angle. If it is not found after rotating 45 degrees clockwise, it is necessary to rotate 90 degrees to continue searching until the grayscale correlation coefficient of the adjacent sub-region is found to be smaller than the grayscale correlation coefficient of the N-1th sub-region. At this time, the searched adjacent sub-region is the Nth sub-region.

[0056] Combining the above two embodiments Figure 2 For example, refer to Figure 2 , in the first search C P1 minimum, and it is greater than the preset coefficient thresholds of all neighboring sub-regions in the other 7 directions. In the second search, move 1 preset pixel step along the direction of region 1 to continue searching, and then reach 1' (if you continue along this direction, the outer layers are 1'', 1''', 1'''', and so on. For the direction of region 2, the outer layers are 2'', 2''', 2'''', and so on). In the next search, if C P1 >C P1' > preset coefficient threshold, move the preset pixel step along this direction to continue searching, and if C is determined later P1'' >C P1' At this time, based on the 1' sub-region, jump to the adjacent sub-region 2'' of 1'' (rotate 45 degrees clockwise), and detect C P1' >C P2'' , then continue to determine C P2'' <preset coefficient threshold, at this time, 2'' is the target adjacent sub-region, and the distance between each two adjacent searches is the preset pixel step size.

[0057] It should be noted that if the target adjacent sub-region cannot be found after searching for one circle (i.e. rotating 360 degrees), it means that the preset pixel step size may be too large, causing the target adjacent sub-region to be skipped. For the solution in this case, please refer to the next embodiment.

[0058] The search direction and strategy are adjusted according to the result of each search. The deformation of the furnace column has a certain directionality, and the angle of rotation counterclockwise or clockwise increases successively, which improves the flexibility and effectiveness of the search and ensures that the local optimal solution is found as soon as possible.

[0059] As an optional embodiment of the present invention, based on the subsequent grayscale image, searching for the adjacent subregion having the smallest grayscale correlation coefficient with the target subregion in the initial grayscale image among the plurality of adjacent subregions spaced from the target subregion by a preset pixel step length comprises: Based on the subsequent grayscale image, based on multiple directions centered on the target sub-region and radially evenly expanded, the neighboring sub-regions that are spaced apart from the target sub-region by the preset pixel step are determined one by one, and among the determined neighboring sub-regions, the neighboring sub-region with the smallest grayscale correlation coefficient with the target sub-region in the initial grayscale image is determined.

[0060] Specifically, Figure 2 As shown, the target sub-region is P. With P as the center, there are 8 radial directions, and the interval between two adjacent directions is 45 degrees. Along these 8 directions, the interval is a preset step size, and 8 adjacent sub-regions can be determined respectively, that is, 1, 2...8. At this time, C is calculated P1 Minimum, that is, C P1 Less than C Pi (i=2...8), thereby determining the second sub-area mentioned above.

[0061] As an optional embodiment of the present invention, it also includes: If the target adjacent sub-region still cannot be found under the preset pixel step size, the preset pixel step size is gradually reduced, and the target adjacent sub-region is tried again to be found until the target adjacent sub-region is found.

[0062] Specifically, when the local minimum of the adjacent sub-region cannot be found under a given preset pixel step, it is very likely that the preset pixel step is too long. In this case, shorten the preset pixel step, and then follow the above steps in sequence. For example, the initial pixel step is set to 3, and the set pixel step k is reduced. Based on the previous two embodiments, the search for the Nth sub-region continues, and the calculation is repeated until k is reduced to 1, and the target adjacent sub-region can be searched.

[0063] As an optional embodiment of the present invention, the grayscale correlation coefficient is determined according to the following formula: .

[0064] in, represents the grayscale correlation coefficient, Indicates the first pixel coordinate in the target sub-area The corresponding gray value, Indicates the coordinates of the second pixel in the adjacent sub-area Corresponding grayscale value, the target sub-region or the adjacent sub-region is a square, and M represents half of the side length of the target sub-region or the adjacent sub-region.

[0065] Specifically, when calculating the grayscale correlation coefficient, since the target sub-region and the adjacent sub-region are both square in shape, the above formula covers the calculation of the grayscale of the pixel coordinates in the two-dimensional plane of the target sub-region or the adjacent sub-region. The larger the grayscale correlation coefficient, the larger the grayscale difference, which means the larger the deformation of the target sub-region. Conversely, the smaller the grayscale correlation coefficient, the smaller the grayscale difference, which means the smaller the deformation of the target sub-region. The grayscale value of each two-dimensional pixel coordinate is determined based on automatic identification after grayscale processing. The formula represents the square of the grayscale matrix difference of the pixel coordinate point.

[0066] Through the above formula, the grayscale correlation coefficient between any adjacent sub-region and the target sub-region can be determined, and then used to determine which region has the smallest grayscale correlation coefficient with the target sub-region, and the grayscale correlation coefficient of this sub-region should satisfy the "minimum" condition, indicating that it is very likely to be the farthest position of the target sub-region after deformation.

[0067] As an optional embodiment of the present invention, the determining the actual offset of the furnace column according to the pixel coordinate difference and the pixel scale between each of the target adjacent sub-regions and the target sub-region comprises: The lateral offset is determined according to the ratio of the lateral pixel coordinate difference between the target adjacent sub-region and the target sub-region to the pixel scale, and the longitudinal offset is determined according to the ratio of the longitudinal pixel coordinate difference between the target adjacent sub-region and the target sub-region to the pixel scale, wherein the pixel coordinate difference includes the lateral pixel coordinate difference and the longitudinal pixel coordinate difference, and the actual offset includes the lateral offset and the longitudinal offset.

[0068] Specifically, the horizontal pixel difference and the vertical pixel difference between the target sub-region and the target adjacent sub-region are calculated, and the actual offset is calculated as: Yi=[horizontal pixel coordinate difference / λ, vertical pixel coordinate difference / λ], where λ represents the pixel value contained in 1 precision size. For example, in this application, every 3 pixels can represent 1 mm.

[0069] The horizontal pixel coordinate difference represents the difference between the horizontal coordinates, and the vertical pixel coordinate difference represents the coordinate difference between the vertical coordinates. The horizontal and vertical coordinates of the corresponding target sub-region and the target adjacent sub-region are respectively subtracted, and then the absolute value is taken, or the absolute value of the offset is finally taken.

[0070] The following example illustrates the calculation. For example, the horizontal pixel coordinate difference between the target adjacent sub-region and the target sub-region is 5, the vertical pixel coordinate difference is 4, and λ is 3, then Yi=[1.67, 1.33]. In actual applications, both the horizontal pixel offset and the vertical pixel offset are taken as absolute values.

[0071] As an optional embodiment of the present invention, before grayscale processing is performed on the image sequence, the method further includes: A filtering process is performed on the image sequence.

[0072] Specifically, the filtering process includes selecting a suitable filter, such as a Gaussian filter, a median filter or a mean filter, according to the characteristics of the image to be processed. Mean filtering can reduce image noise by calculating the average value of surrounding pixels and is suitable for removing random noise. Gaussian filtering: Using a Gaussian function to perform weighted averaging on an image can more effectively retain image edge information.

[0073] In summary, through the above filtering process, the subsequent grayscale processing can be performed more effectively and the data of grayscale processing can be reduced, thereby improving the final image quality and processing effect.

[0074] like Figure 3 As shown, the present invention further provides a coke oven column deformation monitoring system 200, which uses the coke oven column deformation monitoring method described in the above embodiment, including: The acquisition module 210 is used to acquire an image sequence of the marked area on the side of the furnace column.

[0075] The pre-processing module 220 is used to perform grayscale processing on the image sequence to obtain a grayscale image sequence, wherein the grayscale image sequence includes a plurality of grayscale images arranged in chronological order.

[0076] The selection module 230 is used to select a same area in each of the grayscale images as a target sub-area, and the target sub-area has a plurality of adjacent sub-areas.

[0077] The deformation judgment module 240 is used to judge whether the furnace column is deformed according to the grayscale value of the target sub-area in the initial grayscale image and the grayscale value in the subsequent grayscale image, wherein the initial grayscale image is the first grayscale image in the grayscale image sequence, and the subsequent grayscale image is any grayscale image after the first grayscale image in the grayscale image sequence.

[0078] The target sub-region determination module 250 is used to search for the target adjacent sub-region based on multiple adjacent sub-regions of the target sub-region in the subsequent grayscale image if it is determined that the furnace column is deformed, wherein the target adjacent sub-region represents an adjacent region in the subsequent grayscale image whose grayscale correlation coefficient with the target sub-region in the initial grayscale image is less than a preset coefficient threshold and is locally minimum, and the grayscale correlation coefficient is used to characterize the grayscale difference between the two regions.

[0079] The offset determination module 260 is used to determine the actual offset of the furnace column according to the pixel coordinate difference and pixel scale between each of the target adjacent sub-regions and the target sub-region.

[0080] The specific implementation of this embodiment can refer to the corresponding implementation method mentioned above and will not be described again here.

[0081] like Figure 4 As shown, an electronic device 300 provided by an embodiment of the present invention includes a memory 310 and a processor 320; the memory 310 is used to store computer programs; the processor 320 is used to implement the coke oven column deformation monitoring method as described above when executing the computer program.

[0082] In other words, an electronic device 300 includes a memory 310 and a processor 320 coupled to the memory 310. The memory 310 is configured to store a computer program; the processor 320 is configured to perform the following operations when executing the computer program: Acquire a sequence of images of the marked area in the side of the furnace column.

[0083] Grayscale processing is performed on the image sequence to obtain a grayscale image sequence.

[0084] If it is determined that the furnace column is deformed, a target adjacent sub-region is searched based on multiple adjacent sub-regions of the target sub-region in the subsequent grayscale image, wherein the target adjacent sub-region represents an adjacent region in the subsequent grayscale image whose grayscale correlation coefficient with the target sub-region in the initial grayscale image is less than a preset coefficient threshold and is locally minimum, and the grayscale correlation coefficient is used to characterize the grayscale difference between the two regions.

[0085] The actual offset of the furnace column is determined according to the pixel coordinate difference and pixel scale between each of the target adjacent sub-regions and the target sub-region.

[0086] An embodiment of the present invention provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the above-mentioned coke oven column deformation monitoring method is implemented.

[0087] In other words, a non-volatile computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the processor performs the following operations: Acquire a sequence of images of the marked area in the side of the furnace column.

[0088] Grayscale processing is performed on the image sequence to obtain a grayscale image sequence.

[0089] If it is determined that the furnace column is deformed, a target adjacent sub-region is searched based on multiple adjacent sub-regions of the target sub-region in the subsequent grayscale image, wherein the target adjacent sub-region represents an adjacent region in the subsequent grayscale image whose grayscale correlation coefficient with the target sub-region in the initial grayscale image is less than a preset coefficient threshold and is locally minimum, and the grayscale correlation coefficient is used to characterize the grayscale difference between the two regions.

[0090] The actual offset of the furnace column is determined according to the pixel coordinate difference and pixel scale between each of the target adjacent sub-regions and the target sub-region.

[0091] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be completed by instructing the relevant hardware through a computer program, and the program can be stored in a non-volatile computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided by the present invention can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0092] The above description is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.

[0093] Although the present invention is disclosed as above, the protection scope of the present invention is not limited thereto. Those skilled in the art may make various changes and modifications without departing from the spirit and scope of the present invention, and these changes and modifications will fall within the protection scope of the present invention.

Claims

1. A method for monitoring the deformation of a coke oven column, characterized in that: include: Acquire a sequence of images of the marked area in the side of the furnace column; Performing grayscale processing on the image sequence to obtain a grayscale image sequence, wherein the grayscale image sequence includes a plurality of grayscale images arranged in chronological order; Selecting a same region in each of the grayscale images as a target subregion, wherein the target subregion has a plurality of adjacent subregions; Determining whether the furnace column is deformed according to the grayscale value of the target sub-region in the initial grayscale image and the grayscale value in the subsequent grayscale image, wherein the initial grayscale image is the first grayscale image in the grayscale image sequence, and the subsequent grayscale image is any grayscale image after the first grayscale image in the grayscale image sequence; If it is determined that the furnace column is deformed, a target adjacent subregion is searched based on a plurality of adjacent subregions of the target subregion in the subsequent grayscale image, wherein the target adjacent subregion represents an adjacent region in the subsequent grayscale image whose grayscale correlation coefficient with the target subregion in the initial grayscale image is less than a preset coefficient threshold and is locally minimum, and the grayscale correlation coefficient is used to characterize the grayscale difference between the two regions; The actual offset of the furnace column is determined according to the pixel coordinate difference and the pixel scale between the target adjacent sub-region and the target sub-region.

2. The coke oven column deformation monitoring method according to claim 1, characterized in that: The step of judging whether the furnace column is deformed according to the grayscale value of the target sub-region in the initial grayscale image and the grayscale value in the subsequent grayscale image comprises: Determine the grayscale correlation coefficient between the target subregion in the initial grayscale image and the target subregion in the subsequent grayscale image according to the grayscale value of the target subregion in the initial grayscale image and the grayscale value of the target subregion in the subsequent grayscale image, obtain a first grayscale correlation coefficient value, and determine whether the first grayscale correlation coefficient value is greater than 0; If the first grayscale correlation coefficient value is greater than 0, it is determined that the furnace column is deformed, and if the first grayscale correlation coefficient value is equal to 0, it is determined that the furnace column is not deformed.

3. The coke oven column deformation monitoring method according to claim 1, characterized in that: If it is determined that the furnace column is deformed, searching for a target adjacent sub-region based on a plurality of adjacent sub-regions of the target sub-region in the subsequent grayscale image comprises: Based on the subsequent grayscale image, searching for the adjacent subregion with the smallest grayscale correlation coefficient with the target subregion in the initial grayscale image among the plurality of adjacent subregions spaced from the target subregion by a preset pixel step, and taking the adjacent subregion as the second subregion, and taking the target subregion in the subsequent grayscale image as the first subregion; A search step is performed to determine the Nth sub-region, where N is a positive integer greater than or equal to 3, and the search step includes: taking the direction from the N-2th sub-region to the N-1th sub-region as the search direction, taking the N-1th sub-region as the starting point, and determining the adjacent sub-region by a preset pixel step interval along the search direction, and determining whether the grayscale correlation coefficient between the determined adjacent sub-region and the target sub-region in the initial grayscale image is less than the grayscale correlation coefficient between the N-1th sub-region and the target sub-region in the initial grayscale image, and if so, taking the determined adjacent sub-region as the Nth sub-region; if not, based on The N-1th sub-region is rotated clockwise or counterclockwise with the direction from the N-2th sub-region to the N-1th sub-region as the rotation reference, and each rotation angle is increased by twice the preset angle compared with the original rotation angle, so as to determine the corresponding new search direction each time, and locate the adjacent sub-region along the new search direction at intervals of preset pixel steps until the grayscale correlation coefficient between the located adjacent sub-region and the target sub-region in the initial grayscale image is less than the grayscale correlation coefficient between the N-1th sub-region and the target sub-region in the initial grayscale image, and the located adjacent sub-region is used as the Nth sub-region; Determine whether the grayscale correlation coefficient between the Nth sub-region and the target sub-region in the initial grayscale image is less than a preset coefficient threshold; If yes, the Nth sub-region is determined to be the target adjacent sub-region; if no, N+1 is returned to execute the search step.

4. The coke oven column deformation monitoring method according to claim 3, characterized in that: Based on the N-1th sub-region, the direction from the N-2th sub-region to the N-1th sub-region is used as a rotation reference, and the rotation is performed clockwise or counterclockwise, and each rotation angle is doubled by the preset angle compared to the original rotation angle, so as to determine a new search direction corresponding to each time, and the neighboring sub-region is located at intervals of a preset pixel step along the new search direction until the grayscale correlation coefficient between the located neighboring sub-region and the target sub-region in the initial grayscale image is less than the grayscale correlation coefficient between the N-1th sub-region and the target sub-region in the initial grayscale image, and the located neighboring sub-region is used as the Nth sub-region, including: Taking the direction from the N-2 sub-region to the N-1 sub-region as a reference, rotate clockwise or counterclockwise by one times the preset angle to determine the new search direction, and locate the adjacent sub-region at intervals of preset pixel steps along the new search direction, and determine whether the grayscale correlation coefficient between the located adjacent sub-region and the target sub-region in the initial grayscale image is less than the grayscale correlation coefficient between the N-1 sub-region and the target sub-region in the initial grayscale image. If so, determine the Nth sub-region based on this; if not, taking the direction from the N-2 sub-region to the N-1 sub-region as a reference, rotate clockwise or counterclockwise by twice the preset angle to determine the new search direction, and locate the adjacent sub-region at intervals of preset pixel steps along the new search direction. A pixel step is preset to locate the adjacent sub-region, and it is determined whether the grayscale correlation coefficient between the located adjacent sub-region and the target sub-region in the initial grayscale image is less than the grayscale correlation coefficient between the N-1th sub-region and the target sub-region in the initial grayscale image. If so, the Nth sub-region is determined accordingly. If not, the direction from the N-2th sub-region to the N-1th sub-region is used as a reference, and the region is rotated clockwise or counterclockwise by three times the preset angle, and so on, until the grayscale correlation coefficient between the located adjacent sub-region and the target sub-region in the initial grayscale image is less than the grayscale correlation coefficient between the N-1th sub-region and the target sub-region in the initial grayscale image, and the Nth sub-region is determined accordingly.

5. The coke oven column deformation monitoring method according to claim 3 or 4, characterized in that: The step of searching, based on the subsequent grayscale image, among the plurality of adjacent subregions spaced from the target subregion by a preset pixel step, the adjacent subregion having the smallest grayscale correlation coefficient with the target subregion in the initial grayscale image comprises: Based on the subsequent grayscale image, based on multiple directions centered on the target sub-region and radially evenly expanded, the neighboring sub-regions that are spaced apart from the target sub-region by the preset pixel step are determined one by one, and among the determined neighboring sub-regions, the neighboring sub-region with the smallest grayscale correlation coefficient with the target sub-region in the initial grayscale image is determined.

6. The coke oven column deformation monitoring method according to claim 3 or 4, characterized in that: Also includes: If the target adjacent sub-region still cannot be found under the preset pixel step size, the preset pixel step size is gradually reduced, and the target adjacent sub-region is tried again to be found until the target adjacent sub-region is found.

7. The method for monitoring the deformation of a coke oven column according to any one of claims 1 to 4, characterized in that: The grayscale correlation coefficient is determined according to the following formula: ； in, represents the grayscale correlation coefficient, Indicates the first pixel coordinate in the target sub-area The corresponding gray value, Indicates the coordinates of the second pixel in the adjacent sub-area Corresponding grayscale value, the target sub-region or the adjacent sub-region is a square, and M represents half of the side length of the target sub-region or the adjacent sub-region.

8. The method for monitoring the deformation of a coke oven column according to any one of claims 1 to 4, characterized in that: Determining the actual offset of the furnace column according to the pixel coordinate difference and pixel scale between each of the target adjacent sub-regions and the target sub-region comprises: The lateral offset is determined according to the ratio of the lateral pixel coordinate difference between the target adjacent sub-region and the target sub-region to the pixel scale, and the longitudinal offset is determined according to the ratio of the longitudinal pixel coordinate difference between the target adjacent sub-region and the target sub-region to the pixel scale, wherein the pixel coordinate difference includes the lateral pixel coordinate difference and the longitudinal pixel coordinate difference, and the actual offset includes the lateral offset and the longitudinal offset.

9. The coke oven column deformation monitoring method according to claim 1, characterized in that: Before the grayscale processing is performed on the image sequence, the method further comprises: A filtering process is performed on the image sequence.

10. A coke oven column deformation monitoring system, characterized in that: The method for monitoring the deformation of a coke oven column according to any one of claims 1 to 9 comprises: An acquisition module, used for acquiring an image sequence of a marked area in a side surface of a furnace column; A pre-processing module, used for performing grayscale processing on the image sequence to obtain a grayscale image sequence, wherein the grayscale image sequence includes a plurality of grayscale images arranged in chronological order; A selection module, used for selecting a same area in each of the grayscale images as a target sub-area, wherein the target sub-area has a plurality of adjacent sub-areas adjacent to the target sub-area; A deformation judgment module, used for judging whether the furnace column is deformed according to the grayscale value of the target sub-area in the initial grayscale image and the grayscale value in the subsequent grayscale image, wherein the initial grayscale image is the first grayscale image in the grayscale image sequence, and the subsequent grayscale image is any grayscale image after the first grayscale image in the grayscale image sequence; a target sub-region determination module, configured to search for a target adjacent sub-region based on a plurality of adjacent sub-regions of the target sub-region in the subsequent grayscale image if it is determined that the furnace column is deformed, wherein the target adjacent sub-region represents an adjacent region in the subsequent grayscale image whose grayscale correlation coefficient with the target sub-region in the initial grayscale image is less than a preset coefficient threshold and is locally minimum, and the grayscale correlation coefficient is used to characterize the grayscale difference between the two regions; The offset determination module is used to determine the actual offset of the furnace column according to the pixel coordinate difference and pixel scale between each of the target adjacent sub-regions and the target sub-region.

Images